[scikit-learn] branch master created (now 80fc9e7)

Andreas Tille tille at debian.org
Wed Dec 28 12:57:52 UTC 2016


This is an automated email from the git hooks/post-receive script.

tille pushed a change to branch master
in repository scikit-learn.

        at  80fc9e7   Build-Depends: python-pytest, python3-pytest

This branch includes the following new commits:

       new  92fbeab   Initial repository layout
       new  619da34   Initial commit for learn, a set of algorithms and tools for machine learing in python
       new  fbb3c51   Adding oldfaithful to datasets
       new  def073b   Adding Iris data from UCI ML db
       new  2cc9f2b   Adding pendigits from UCI ML DB
       new  9e79310   Add __init__ file to iris dataset module
       new  7288b73   Add heart data
       new  7635286   Add German credit dataset
       new  3f5b1f2   Add __init__ file to german and heart datasets
       new  552535a   Split german data between features and label
       new  a889cae   Add learn.preprocessing, for data pre processing
       new  e824c3f   Add a basic readme to learn.datasets package
       new  076938f   Flatten data in Iris dataset
       new  290018f   Add a misc subpackage in datasets for tools useful to several datasets
       new  898e32f   Flatten oldfaithful
       new  cd076af   flatten german and pendigits
       new  2e66160   Add a scaler object + tests.
       new  4b049f1   Add function to handle Nan data for scaling.
       new  9cd2040   First code commit.
       new  db45b2f   From: fred.mailhot <fred.mailhot at cb17146a-f446-4be1-a4f7-bd7c5bb65646>
       new  06e6ed2   Adding data dir with data files.
       new  b7459be   - Got least-squares optimization working - Stuck more info into my "test" function to   confirm that leastsq is doing something good - Began incorporating simplex, bfgs and ncg   (added explicit objective fxn for this to work) - discovered dot() and transpose() fxn for arrays... :)   some refactoring of matrix multiplication avoiding   conversions calls to matrix() and array()
       new  1901d26   Implementing error/gradient functions more carefully. Borrowing heavily from Netlab.
       new  394dd1f   Extensive refactoring of mlp (still convergence probs with some optimization algs). Started work on srn, should commit a good deal more tonight (want to have a working srn with at least one training alg before the weekend).
       new  a164501   Minor changes to srn.py. Plans for future dev included in docstrings & comments.
       new  e6f86d6   New start with RBF net. Will finish this Sunday.
       new  3e6e770   Heavily refactored MLP. Only leastsq used for optimization.
       new  633fbcc   Refactoring SRN based on MLP. Simple 1-step backprop only.
       new  3875829   data files for testing
       new  8a55615   more data files for testing
       new  7c48b54   Heavy refactoring. These are more or less complete now. Small additions remain to make the interface more obvious for users.
       new  7bcd7fc   Updates/refactorings for mlp and srn, addition of rbf.
       new  a81331e   Removed two data files.
       new  6bd1582   Updated __init__
       new  47f46e6   Easier way to select random sample of data for RBF centers.
       new  ff25635   HTML files for {mlp/rbf/srn}.py generated by pydoc uploaded.
       new  75b6c2b   Add axis arguments to functions that changed in NumPy 1.0b2
       new  3de66ff   initial checkin of ga library.
       new  effb5e5   much clean up of module paths after files were moved.
       new  8ec6973   initial attempt at including Numeric as a sub-package
       new  a48083f   attempt at fixing the chaos created by my last checkin.
       new  f545bc6   fixed several package related issues with from x import *
       new  2a1f48e   added from fastumath import * to all sub-modules in scipy (besides Numeric)
       new  7b201b4   Fixed a time.time function replacement error that Prabhu found.
       new  03f8b71   adding setup_xxx.py files for the new setup.py infrastructure.
       new  2a365d2   Fixed import of stats module in ga.
       new  6e77163   Moved shelve to io.  Cleaned up __init__ files.
       new  7db9945   Changed references to fastumath to scipy_base.fastumath
       new  c3005a2   Changed == None to is None
       new  3ee61d0   Converted module to use 'import' directly, instead of exec 'import...'
       new  e361f17   Carried out major unification of xxx/setup_xxx.py files. Discussion: some modules contain setup.py files that repeat the functionality of the corresponding setup_xxx.py files. Are there any objections if setup.py and setup_xxx.py will be merged into setup_xxx.py and setup.py files will be removed from CVS?
       new  ab97ba3   1) Finished applying ppimport hooks to scipy. New scipy import timings on PII 400MHz, Debian with postponed import hooks enabled:
       new  bb4a10c   * Made changes to bring ga up to speed with latest stats module. * fixed remove_NaN function to use a more scipyish solution.
       new  153e48d   Replaced execfile statements with simplified import statements in order to avoid problems with Windows installer tools.
       new  2b55e32   New import hooks are applied to cow,ga,sparse. Clean up.
       new  6123125   Fixed import problem with importing __pre___init__
       new  9515166   Converted usage of whrandom to random.
       new  609841c   More changes to get rid of whrandom usage.
       new  5b515cc   Fixed issue 216
       new  bbf65a1   Converted scipy to use numpy. Several tests fail but are easily fixed.
       new  0b3f121   Fixed old imports (ticket #36)
       new  c017b1f   Run reindent.py (script distributed with Python) over the source to remove extraneous whitespace
       new  6169fc6   remove unused imports
       new  430ed93   remove duplicate definition of __float__
       new  1549b35   Add axis arguments to functions that changed in NumPy 1.0b2
       new  e3ed0fd   Initial commit David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-06-08 17:59:58 +0900 (Thu, 08 Jun 2006)
       new  ba57268   Put files into a package David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-12 13:52:02 +0900 (Wed, 12 Jul 2006)
       new  d74b2e8   Package uses distutil David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-13 15:20:08 +0900 (Thu, 13 Jul 2006)
       new  3d83d80   Merge with main branch David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-13 19:35:51 +0900 (Thu, 13 Jul 2006)
       new  ac1c6a7   Add Changelog, put version to 0.3 David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-13 19:51:04 +0900 (Thu, 13 Jul 2006)
       new  d2fbad4   Refactoring of EM into classes David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-14 16:26:45 +0900 (Fri, 14 Jul 2006)
       new  0ab64ed   Add scripts for benchmarking, minor corrections David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-07-14 19:28:13 +0900 (Fri, 14 Jul 2006)
       new  33ad4b1   Push to version 0.4.2 David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-04 19:48:19 +0900 (Fri, 04 Aug 2006)
       new  a4a1e58   Version 0.5, add custom confidence ellipsoids David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-07 18:46:57 +0900 (Mon, 07 Aug 2006)
       new  5b70fa2   Adapt to numpy 1.0b3SVN David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-17 18:48:51 +0900 (Thu, 17 Aug 2006)
       new  18e7b0e   correct bug with GMM.init_method David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-24 19:44:42 +0900 (Thu, 24 Aug 2006)
       new  5c17a83   revert to custom kmean David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-24 20:12:27 +0900 (Thu, 24 Aug 2006)
       new  a2dd902   Various cleanup, include ctypes version of diag gden; update profile function to include ctypes version David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-08-28 22:03:28 +0900 (Mon, 28 Aug 2006)
       new  95f3f83   - Add a plot1d method to GM class
       new  fb65255   - Improve plot1d method to be more useful
       new  2d0a200   MSG David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-10-03 18:29:50 +0900 (Tue, 03 Oct 2006)
       new  c5ca73f   Bump to 0.5.3 David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-10-03 18:34:42 +0900 (Tue, 03 Oct 2006)
       new  02648cb   Add preliminary online functions (does not work yet) David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-10-06 20:03:59 +0900 (Fri, 06 Oct 2006)
       new  8ccb807   Change of layout for inclusion in scipy David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-10-12 21:19:17 +0900 (Thu, 12 Oct 2006)
       new  b2c0a76   Change of layout for scipy (2) David Cournapeau <david at ar.media.kyoto-u.ac.jp> | 2006-10-12 21:21:33 +0900 (Thu, 12 Oct 2006)
       new  231fef1   Last pyem layout cleanup to finalize inclusion into scipy trunk
       new  fb4ef88   Add svm and pyem package in the comments of setup.py; correct bugs in the online EM script
       new  b6015fb   * Convert the online_em.py script to a class, OnGMM * Add tests for OnGMM * Small bugfixes in _c_densitites (convert general exception to ImportException when wrong ctypes version found), setup (remove the name attribute to avoid warning when installing)
       new  1b1682c   * pyem: added GM.bic function to compute Bayesian Information Criterion for automatic model selection + various docstrings fixes
       new  84dabaa   * pyem: trivial change of API for OGMM
       new  6788c70   * sandbox.pyem: fix a bug which prevents full covariance from being used + tests to avoid this issue from being unnoticed
       new  ca1fa13   * bump to 0.5.6 * various cosmetic changes
       new  b0ecea1   Various bug fixes (see Changelog)
       new  07ed219   Add densities2.py for preliminary axis support and add specialized class for online EM in 1d for a 10-100x speed incread
       new  90ada96   Add a script to profile online em
       new  5bf5c97   Move documentation files in doc repository. * all examples lie in doc/examples. * A Makefile to generate the pdf from the rest files and python examples is now included.
       new  119c2ed   Put examples directory at the top level, so that they be easily imported for tests.
       new  2298f9c   Remove kmean as scipy.cluster.vq.kmeans2 does everything we need now
       new  72e9a55   More benchmarking for basic operations in row vs col
       new  672d1e7   Refactor 1d computation for plotting
       new  6bdc230   Add some profiling scripts to compare likelihood computation with matlab.
       new  179dde0   Add faithful data in data.
       new  b0f6ea4   Add function to plot density contours in GM.
       new  410679d   Add proper license in data, correct typo (double copyright)
       new  3335c5a   Polish contour functions, so that choosing the dimension of projection works.
       new  ac448f4   Fail nicely when call wrong plot function (plot1d for multinomial, plot for 1d models).
       new  542b00d   Add example of pdf estimation with EM
       new  1a85ab2   Add data as a proper submodule
       new  aeffb9e   Clean up densities.py code, set docstrings to rest
       new  0f2bc9c   Set def arguments to immutable to avoid nasty side effect.
       new  b4b1527   Heavy liftup of the code + docstrings.
       new  fe9f15b   Add special initialization method for mixture models for testing purpose.
       new  e71b89c   * Correct bogus GM._get_va which caused bogus isodensity plot + test * Support for plain matrix in GM.check_state
       new  e921989   Add a plotting example.
       new  bead4de   Remote outdated test script.
       new  b32b132   Add basic tests for EM, 1d, 2d, full and diag mode
       new  51c2aa1   Reenable tests I forgot to uncomment in gmm_em tests
       new  b5405eb   Convert check calls to test calls in tests, for future convertion to setuptools
       new  3fcf234   Add tests for pdf computation in log domain (1st step for logsumexp trick support)
       new  be7f78d   Add logsumexp function + tests. Not used in the code yet, though
       new  577db67   Add function to compute log responsabilities with logsumexp.
       new  57bd65f   Fix importing datasets in pyem/test
       new  3686215   More fix for broken tests in gmm_em
       new  c28a855   Trivial fix for typo in pyem tests.
       new  c14e5e0   Refactor update step for EM (split diag and full case in subfunction)
       new  f72cefa   add pendigits data from UCI machine learning website. Will be useful for testing regularization.
       new  b765fa6   Add pendigits as a subpackage of data for distutils.
       new  b888eed   Change named arg varmode to mode for consistency, and add an argument for number of contours in density on grid
       new  7ebfeb9   Update docstring of logsumexp.
       new  9fe7ff7   Add (crude) regularized EM
       new  c3835d0   Add support for EM in log domain + tests
       new  f58bbc8   Add utils function to easily load pendigits data
       new  c4dd621   Add iris data from UCI ML database
       new  1409be5   Remove datasets (include them in sckits.learn instead)
       new  9e7e963   Add curry class to misc tools
       new  b071a5d   Remove data dir from subpackages in setup
       new  c3ce1f6   Clean up some examples, and add an example for regularized EM
       new  86ca759   Forgot to add the file implementing regularized example...
       new  c311be4   Use log pdf when possible in plot functions
       new  b6c2fd7   Clean up code for 1d plotting.
       new  ad6b37f   Clean up for densities.py
       new  bd1e6b2   More clean up
       new  9d53710   Add discriminant analysis example
       new  4bcf71f   Update doc
       new  92270d9   Add forgotten picture necessary for generating the doc
       new  11e3e1b   Significantly improve speed of gauss_den
       new  ac88c10   More speed improvements for full matrices case
       new  3ffe3c2   pylibsvm SoC project getting under way.
       new  a05ad03   libsvm 2.82 sources.
       new  d0552f5   Basic classification, regression and one-class with new API.
       new  a0a116f   Fixed tests.
       new  fdd5b76   Started on documentation.
       new  d8810b5   Separated out kernels. Added facility for custom kernels.
       new  66fb7fb   Fixed RBF kernel.
       new  7b5799b   New design to hide libsvm data structure from the users.
       new  79b1fa2   minor
       new  e570e2f   Build a proper libsvm DLL on Windows.
       new  77e1e19   Code integrated into NumPy itself.
       new  4eb9bce   Datasets.
       new  b50628a   Basic tests for datasets.
       new  f0075bc   Minor changes due to code moving to Numpy.
       new  f363dc6   Fix kernels.
       new  5cc746e   Training of regression model works.
       new  1537b28   Predict stuff. Added some docs.
       new  4f8d73d   Classification test and some minor fixes.
       new  1bbaba7   Regression test. Workaround for possible ctypes bug. Probability output for classification models.
       new  f6ee8ee   Workaround for bug in ctypes 0.9.9.6.
       new  b04f33a   Extend one-class estimation a bit.
       new  8143c8b   Cross-validation for classification and regression.
       new  b7dd4f0   SConstruct file to build shared library until numpy.distutils gets sorted out.
       new  1a3b4fd   Tweak so that test passes with GCC.
       new  5def97e   First attempt at a setup.py.
       new  ece0e89   Howto for getting the test suite to run.
       new  48878f6   Minor updates.
       new  eb68f21   Let ctypes handle references to the contents of the svm_problem.
       new  b4e59b7   'randn' is no longer in top level namespace. Call N.random.randn.
       new  b46d08e   Simplify libsvm wrapper and make tests run without having to set PYTHONPATH.
       new  ccbf125   Dataset for making precomputed kernels.
       new  b0c7b74   Precompute with any kernel.
       new  3ec80d4   Working on support for precomputed kernels.
       new  4b06142   Reformat code to conform to PEP 8.
       new  b505041   Dataset for dealing with precomputed kernels.
       new  6d2ebc9   Precomputed model training.
       new  4f1c12f   More cleanups to conform to PEP 8.
       new  b871fa4   Factored out prediction logic so that we can support precomputed and custom kernels.
       new  ef76a6f   minor
       new  414eb0f   Prediction refactoring in progress.
       new  55bb742   Predict without calling libsvm.
       new  9179e3b   Code cleanup. Disabled tests that are failing due to NumPy defect.
       new  71d9f90   More tests.
       new  d791ff4   Collapse predictor classes to yield a nicer API.
       new  95ade90   Fix libsvm prediction with precomputed kernels.
       new  1b475cc   Make sure custom kernels are only used with precomputed datasets.
       new  7c5cd77   Test for classification with precomputed kernel.
       new  660b778   Don't train probability models by default.
       new  21ffc14   Use .ctypes.data_as instead of ctypes.cast.
       new  6f3c2b6   Use release build linker options.
       new  01ce0e3   Extend classification tests.
       new  5dc3dc6   Enable tests that exposed NumPy defect.
       new  b130007   Improve one-class tests.
       new  9bef429   Extend tests to check nu-SVR.
       new  44fb893   Classification in Python in preparation for some optimization.
       new  d2b39db   Model compaction.
       new  a49aa3a   Move zipping inside the datasets.
       new  57638ed   Benchmark with large test dataset.
       new  a2514e3   Refactoring kernels to work better in sparse and dense cases.
       new  fd9ab5d   Fix a bug in ellipses of confidence computation for mixture + adapt to new datasets layout in examples
       new  7f7d1e5   [PATCH 049/166] test
       new  cb2e537   undo fix
       new  3aa0828   resyncing with scipy version
       new  b1e868c   Move former scipy.sandbox modules into a separate package machine
       new  88dd6f3   Modify setup to that svm and pyem are built from their setup
       new  70f6234   Enable scikits.learn.utils test through setuptools --test-module mechanism.
       new  1366836   Adapt tests + module import in pyem such as they work with setuptools test command
       new  1678723   Add test and test_suite function to svm module
       new  c48bfe5   Ignore junk files in the whole learn tree, make machine a proper subpackage by adding an __init__.py file
       new  931406f   Make svm/examples a package so that its files are copied when installing
       new  e7a2dfa   Prototype of attributes and class selection, based on introspection
       new  44a1c48   Do not detect subpackages using find_packages from setuptools, but uses hiearchical setup.py ala scipy instead. Should solve #24
       new  5792b54   Set correct name for datasets packages (data->datasets) in the setup, and trivial changes in datasets info message
       new  91b83e4   Changing pyem to em
       new  ef0e920   Change examples and tests following the change of package name pyem->em
       new  e458dbc   Fixed online em tests
       new  bc1f4b5   Split pendigits into training and testing datasets, and convert return value of load to the package conventions.
       new  89e377b   Remove pyc junk.
       new  febe21b   Adapt examples using pendigits to the new format.
       new  141297f   Put back changes which somewhat were not commited before for setup in em sub package
       new  8162903   First version of arffreader. Seems to read correctly most headers. Data not parsed yet
       new  7af8d29   Non working arff reader
       new  d5b5935   Working arffreader
       new  9065f50   Printing a small summary for arffreader
       new  0eb4296   More thoughts on arffreader
       new  cf7d2b6   Handling missing values (?)
       new  99bc40c   Simplify Kernal Api names
       new  8bb99f8   Finish api names simplification: no more Kernel suffix, no more LibSvm prefix
       new  b56690d   Add partial evaluation tool to replace functools module for python < 2.5
       new  bb07195   Make common a module
       new  37bfa7e   Add common as a module to install in setup
       new  d5944a7   arffreader can now parse all numeric data of UCI :)
       new  a05ec04   Catch error while parsing arff file, and raise ParseError for expected problems
       new  f7126a4   Cleaning a bit the arffreader module
       new  6e52abe   Add a small java program to parse arff using weka (meant to be used for testing our arff implementation)
       new  fe9bbcc   Abstract away all meta data returned in user functions into a MetaData class
       new  759f9c8   Add scripts for quick and dirty test from java
       new  1b20d36   Adding dataset proposal
       new  309f465   Added misc to setup.py
       new  a82765a   added a data set with missing entries.
       new  a4a13db   Added the narr module to load NCEP NARR data sets (reanalysis of around 190 variables). Depends on pydap.
       new  46c619e   First manifold_learning_toolkit commit : not everything may work
       new  ea98ad4   Small update for the dimensionality reduction folder
       new  fe226ff   Small argument mistake
       new  52c89b0   Some bug fixes, some name changes, ...
       new  0203715   Remove fake pyem package.
       new  e846b09   Remove the arff reader: an improved version is available in scipy.io, under the name loadarff.
       new  f3d88c8   Added some missing files
       new  a2a7473   Added a simple compression example
       new  2b5d2b1   Fixed some additional arguments and modules issues
       new  4a9cc47   Remove pyem from setup.py
       new  16f7762   Changed neighbo(o)rs name
       new  1ae5026   Fixed some extensions bugs
       new  7812eeb   Disable manifold learning, because it does not build.
       new  9be9fc7   Updates
       new  9db3bb6   Setup.py should be more robust now
       new  9200b4f   Updated the classes according to the online tutorial
       new  34a2a8a   Small correction
       new  6277087   refs #60: for gcc < 4.1, SFINAE seems to be bogus, so iteration copies are deactivated by a macro.
       new  4f88104   Rename NumpyTestCase to TestCase.
       new  2f77054   Remove path madness, which is deprecated anyway.
       new  d40909b   Do not use set_package_path anymore in machine.em tests.
       new  8200553   Fix typos in tests.
       new  f00ae30   Remove deprecated test runned for em machine.
       new  0346cd4   Add online_gmm_em method.
       new  6bd7120   Add compute factor.
       new  70def2e   Start working on em2, an backward incompatible, more scalable rewrite of em.
       new  3a89497   Start working on new GM class.
       new  ca6a571   Be consistent between w and weights.
       new  3c603cd   Add changelog.
       new  630b576   Add a README for em2.
       new  96a990c   Trailing spaces.
       new  d09869e   Add logsumexp wrapper around cython implementation.
       new  f58c20a   Implement logsumexp in cython.
       new  7c3f4aa   implement pdf method for GM
       new  bca18a6   Implement pdf for GM.
       new  bada733   Do not execute example statements at import time.
       new  77016cb   Implement sample method for GM.
       new  7762b60   Postpone learn import.
       new  cc22ebb   Add a GMM class.
       new  c9c862a   Add test code.
       new  9ff9132   Start experimenting with accumulate API for EM.
       new  421c06a   Add kmeans-based init for w, mu, va.
       new  5f5b380   Start EM imp.
       new  ec79c06   1e-15 too stringent.
       new  649dead   Remove GMM class.
       new  527df19   Use new classes for new EM.
       new  efd49f2   Whitespace.
       new  9a740ce   Add em2 to learn scikits.
       new  3dd6ea5   Implement one component likelihood function.
       new  8a235ac   Add (buggy) cython implementation for one comp likelihood function.
       new  b9eb69b   Add tests for logsumexp and mnormalik.
       new  154d234   Add basic test for EM.
       new  ad5990a   Update README.
       new  ae436f9   Remote hint parameter from EM ctor.
       new  18bd1e0   Fix utils subpackage __init__; used old numpy testing framework.
       new  183671f   Raise exception in normalik, since it is buggy.
       new  83f306e   Fix import.
       new  5abce80   Add test for logresp.
       new  32c2415   Revert "Raise exception in normalik, since it is buggy."
       new  0fad29c   Make a frame per frame inline logsumexp to reuse it.
       new  8924c97   Use a frame/frame inline quadform to reuse it.
       new  4114f07   More tests for logresp (python version).
       new  5baa817   Start working on cython version of logresp.
       new  efd31fb   Split tests for logresp python vs cython.
       new  8784f21   cython logresp now works.
       new  78ae432   Rename methods of the Scaler class.
       new  c77d320   Do not install em2 by default, not mature enough yet.
       new  61bc9aa   Moved from old openopt to the new optimization scikit
       new  45a4032   Deleted Boost dependency (should OK now)
       new  70663e9   BUG: port em tests to new numpy >= 1.2 infrastructure.
       new  bf790ab   Added README file.
       new  85c4c9b   Use nosetest for testing.
       new  f2c190a   Disable test for this dataset
       new  33fd417   Add .gitignore
       new  100202b   affread is now part of scipy.io.arff
       new  8e5e7ed   Add __init__.py file to the test module.
       new  89140d8   Fix import paths in manifold_learning.*
       new  9cd44a0   Remove old testing framework
       new  cdf45a1   Web page sphinx infrastructure.
       new  1c53013   Update README file.
       new  93735fe   Fix: Add std namespace in order to use std::isnan.
       new  8a2de05   Build extensions inplace.
       new  c2a527b   Update README
       new  abc71c9   Added COPYING file with the license terms.
       new  78e424b   MISC: Cosmetic changes in the rst formatting of the README
       new  e6fb4b2   Update README.
       new  a7f3ab2   Use unittest.TestCase instead of numpy.testing.TestCase.
       new  85ae94f   Added funding orgs to web/funding.rst
       new  564651c   Add warning to web page.
       new  215868f   Add example to Neighbors.kneighbors in machine.manifold_learning.neighbors
       new  ac81348   Revert commit 229 (inclusion of namespace std in ModifiedGeneralClustering.h)
       new  362fe46   Initial import of rst docs.
       new  1b462c7   Docstring fixes.
       new  43238e5   Fix typo in class name.
       new  61cd53e   Remove old testing framework.
       new  4f9eed2   Update README.
       new  a53469a   Update distutils info.
       new  de98912   Move examples to a common place.
       new  75e331d   Remove deprecated path hack.
       new  9165acf   Enabling test_speed.
       new  fad3396   Neighbors refactoring, predict method added.
       new  878afb2   Update examples and move them to a common place.
       new  bf01dba   Update README. Specify dependencies more clearly.
       new  4378441   Remove old testing framework.
       new  916a939   Remove rendundant file extension in numpy.ctypeslib.load_library().
       new  f6d5ada   Fix doctests in Neighbors.
       new  f9c2d4f   Move (and update) documentation for em module.
       new  81be8f7   Sphinx extensions.
       new  558cca0   Add ReST docs.
       new  2c08d5a   Delete module manifold_learning.regression.neighbors.utilities
       new  f035b0f   Fix functions that depend on Neighbors.
       new  5506171   Add AUTHORS file
       new  3a505a7   Update web page.
       new  00ee445   Change location of header files.
       new  2d539db   Fix typos in setup.py.
       new  501bdaa   Add needed files from numpy.testing.
       new  6b82d6d   Remove genetic algorithms.
       new  734a7f7   Remove references to deleted classes.
       new  4315b69   Cosmetic cleanup of setup.py.
       new  8802ddc   Use KDTree from scipy.spatial in Nearest Neighbors.
       new  0851752   Add tests for Nearest Neighbor Algorithm.
       new  4c60fb5   Update KNN examples.
       new  ccccfd0   Add a MANIFEST.in file.
       new  bddd02a   Remove C header file for neighbors.
       new  93f28a0   Fix relative imports in manifold_learning/regression/tests.
       new  9dd5d28   Fix bug in barycenters.
       new  f11ec15   Fix imports of scikits.optimization.
       new  a515b51   API changes for Neighbors broke some tests.
       new  179921f   Fix bug in knn example.
       new  6f97956   Deleted MANIFEST.in
       new  c15e8ed   Fix typo in web page.
       new  ac02741   Clean some setup.py files.
       new  f209999   Deleted common/myfunctools.py.
       new  8ab0e8f   Refactor test_neighbors.
       new  e6989ef   0.1 beta release
       new  da93d78   Remove references to deleted submodule common/
       new  bb252af   Added docstring to cost_function module.
       new  a58432e   Fix path in examples/projection.py.
       new  b91f84b   Fix classification example in machine.svm.
       new  9309f53   Remove sciktis.optimization as explicit dependece from setup.py.
       new  04b256b   Remove em2 module.
       new  7171d37   Remove tests that depend on weka.
       new  fca4eb0   Update README.
       new  d1e8269   Fix import of libsvm in tests.
       new  dcd50b4   Fix: some distutils cleaning and import fixing.
       new  2a1b83d   Building the modules inplace broke installation of binary modules.
       new  979c04c   Fix parzen method in Neighbors.
       new  8ff9fc8   0.1 Release
       new  b517458   Update README with Sourceforge download section.
       new  1a373fe   Remove machine prefix from namespace.
       new  24efb57   Fix import paths that where broken in last commit.
       new  04595f7   Fix typo in numpydoc (sphinx extension).
       new  d1177d8   Updated docs to reflect the new namespace structure.
       new  c7424f6   Some disutils cleanup and bug fixing.
       new  1fc649f   Update neighbors module.
       new  a2358b2   Delete old neighbors module.
       new  6ffc89e   Fix typo in setup.py
       new  c07d790   Start of 0.2 development cycle.
       new  244dcc5   Update setup.py to reflect neighbors new location.
       new  677c6fb   test: added my name to AUTHORS
       new  09f649d   Remove redundant imports of cost_function
       new  c2713a3   Do not compile modules that depend on boost by default.
       new  9e1a5b7   New faithful dataset.
       new  f1e5c48   Datasets refactoring.
       new  a499830   Import dap.client only locally.
       new  d489d0f   Remove redundant info from project's setup.py
       new  a35fe1c   Fix typo in README
       new  3336b99   Delete web page, we will use the doc as webpage.
       new  46adad5   New bindings for libsvm.
       new  fcc11d2   Neighbors module refactoring.
       new  d122751   first attempt to handle standardization of data in SVM
       new  1a1658a   Plotting sample for svm.
       new  9ae3b4f   Delete old examples.
       new  80c58df   Add support for latex equations in sphinx.
       new  7b611ca   Update contribute section.
       new  4c1f024   Add sourceforge download area to install info.
       new  62e482e   Preliminary rst doc for svm.
       new  af0f229   Fix: do not scale by default in SVM.
       new  40eefbc   added unisvaraiet selection. untested
       new  897074f   First commit of bayes methods
       new  7ab28cd   bootstrapping lasso implementation with coordinate descent
       new  2d9b9af   Tests should not be executable (or nose will complain)
       new  f43e97d   Add convergenve threshold
       new  3398eab   adding myself to authors list
       new  abe6351   Doc and corrections for ridge_regression
       new  032c14d   some progress on feature selection: Ftest OK
       new  0dc79e2   Clean bayes code.
       new  43e9e18   add the log-likelihood to the ridge regression function.
       new  2b0e079   first trial with svm. Does not work yet
       new  6e0358a   adding missing init file
       new  4c29de1   remove empty dir
       new  6b937e9   Add samples_generator folder for simulated data
       new  8b6b770f6 Move samples_generator to datasets
       new  bb683ef   Bayes classifier object.
       new  04fee4e   bug fix in Lasso with coordinate descent
       new  f04fd1f   renaming lasso.py to cd.py
       new  d287ddc   renaming in cd.py
       new  b07e7c2   cosmit + module docstring
       new  87105f2   ENH: Implement the feature selection the simplest way.
       new  68291e5   first test for coordinate descent lasso
       new  fae9cab   adding support for elastic-net with coodinate descent
       new  8e2649a   adding test for ElasticNet
       new  750f06e   Fix linear and add test for ridge
       new  95a8d47   make sample_generator a package
       new  55b0f03   cosmit
       new  60d0456   add comment in test
       new  984f563   cosmit + docstrings
       new  cb133fa   adding implementation of ridge regression + test
       new  1265dc1   docstring for enet_dual_gab
       new  c59b10a   DOC: Add an example of feature selection.
       new  2694fe4   Put fast_logdet into utils.
       new  60ddea6   added the tests for univariate selection
       new  010d7b5   bayesian regression with ARD priors
       new  fd06977   adding paths for Lasso and ElasticNet + test
       new  1d5c6a7   a small fix in univariate selection
       new  f15ab79   test that the L1 and ElasticNet prior actually induce sparsity
       new  714a5ef   addind Lasso + ElasticNet example
       new  58e3af8   cosmit
       new  4cda9c0   cosmit + docstring
       new  638d6bf   Fix ARD and add the log-likelihood
       new  35643d2   Fix import bug
       new  6458d08   cleaning Lasso + ElasticNet example
       new  9e0bc0c   cosmit
       new  a7cb941   draft of documentation for GLM
       new  225196c   minor fix of imports and plot
       new  c5fd678   temptative commit--deletion of feature-select messed up things
       new  0c25ea6   Bayes return objects
       new  6e27673   did the svn move manually
       new  541f65b   added a test to compare with stat_model
       new  888e49b   use npy_intp instead of int
       new  e7d86c7   add util method for Leave One Out crossvalidation
       new  e8d3632   add split method in LOO
       new  2dce140   adding missing svn keywords
       new  fb5bfea   Correct API and nose tests for Linear Regression (Noprior, Ridge and ARD)
       new  b2f4fc8   ENH/API: Rewrite the leave one out to use a generator.
       new  19724be   svn 64 bit fixes
       new  d3f27bd   revert unintended changes
       new  b7f18cc   DOC: Beautify the lasso coordinate example and documentation.
       new  e9a5506   Some docs and examples - move bayes.bayes to bayes.regression
       new  473bf33   updated cross-validation
       new  87150fc   cosmit on bayes
       new  81828fb   cosmit on crossval
       new  3adc669   Add nonlinear samples generator
       new  c2dd8e7   Cosmit
       new  ae6f559   work in progress on callback system during fitting (not used yet)
       new  7c75370   Type changes in libsvm bindings.
       new  1ebb534   minor fix
       new  e8a6772   Remove empty directories
       new  50da323   leveraging the new callback feature for GLM to implement early stopping based on duality gab
       new  78f2392   typo
       new  e5550f7   typo
       new  116c7be   give default values for alpha and beta and check that duality gap never increases for early stopping check
       new  5854321   Implement copy of parameters from svm (python) object to libsvm.
       new  365ce02   Refactor plot_svm examples.
       new  2e8b74d   Fix cython extension to work on 64 bit
       new  5082699   Change __repr__ for classes
       new  650679a   Cosmit : add some __repr__ for classifiers classes. Perhaps a more standard API should be implemented for this... This is a temporary fix.
       new  196251b   Enhacement: Performance improvements for libsvm_helper.c
       new  c6a2621   Flat is better than nested
       new  b1e7e1d   Add Olivier Grisel to the AUTHORS file.
       new  03954da   Rename of SVM classes.
       new  6a1f4be   Remove unused imports in svm's benchmarks
       new  ef9bbfe   Update examples with new API changes.
       new  677d231   Remove redundant info in libsvm_helper.c
       new  66c4b5d   Update docstrings in svm module.
       new  567f766   Remove unused imports from cython files in glm/
       new  a19b92b   Add the Lemma of Matrix inversion
       new  6bc18b1   Refactoring of the glm module.
       new  8de51b7   Refactoring of the bayes module.
       new  d549a60   Add and update documentation for linear regression
       new  9afe339   Implement new dataset.
       new  3aab715   Fix performance issues in bayesian ridge regression.
       new  35e6a87   Update examples to reflect new dataset
       new  dd998ff   Move Bayesian Ridge Regression to module linreg.
       new  486dc51   Move remaining code in bayes to linreg.
       new  524b80d   Remove empty directories
       new  6f7c396   Remove empty directories
       new  1b754e0   Remove empty directories (yes, another one)
       new  7df0e83   Remove empty directory ...
       new  3ee4403   Remove empty directories (the last one, I hope)
       new  a7ada5b   Flat is better than nested.
       new  f1cac6f   Remove empty directories
       new  97aeb8d   Refactor svm model.
       new  e7cb8e8   Add tests for linreg.Ridge.
       new  f26dc0d   Add some doc for the svm module.
       new  b806b12   SVM refactoring
       new  d5a0905   Remove nonexistant directories.
       new  35e12c2   - bug fix in stop criterion of Elastic-Net - rewriting callback system to handle objective value computation and density. It  is designed to be usable by other iteration procedures such as proximal based optimization schemes.
       new  7f4bb30   revert commit linreg -> glm (GLM works for models where the predicted value is a linear funcition of the data such as least squares, lasso or logistic regression)
       new  6ffed3e   moving GLM cython code to src folder
       new  54551c0   renaming cd.py to coordinate_descent.py
       new  c7d85c5   renaming cd.py to coordinate_descent.py in example
       new  41158da   Welcome Vincent.
       new  ad9d5a9   Documentation for the svm module.
       new  8b53b61   Cosmetic fixes.
       new  2c6e37e   update Iris dataset
       new  75842de   Fix memory leaks.
       new  529305e   tests for svm
       new  0dd4fdc   svm
       new  ce49941   Fix segfaults in the case of SVR, OneClass.
       new  59e8ccf   Some dosctring and cosmetic changes to module svm
       new  abca47f   cosmetic changes to svm module.
       new  3faffbb   Add svm example.
       new  d973250   speed improvement in coordinate descent methods
       new  538917c   Fix bug when target vector dtype != int.
       new  96ed447   Add neighbors example
       new  0b3bee7   neighbors squash
       new  d01f64d   Dataset refactoring.
       new  706b8bc   Cometic commit, svm module
       new  bb5ce9f   Beautify the svn example (colors where ugly!)
       new  7dc5f07   Documentation update.
       new  3ef4274   Add another svm example.
       new  eebb4dd   Put images in a directory in doc.
       new  b3f4268   Documentation update.
       new  f7b9686   More doc for the svm module.
       new  f0ce49f   Fix spelling errors in svm doc.
       new  fae7bf3   Mark Ann module as deprecated.
       new  2704a3f   Fix memory leaks in the svm bindings.
       new  82e5b47   Update benchmarks
       new  fc4cf73   Update Neighbors module.
       new  444aabf   Remove unused functions from svm module.
       new  98d0a18   Add module glm to parent's setup.py
       new  17d2127   Fix failing tests.
       new  9be51b4   Fix distutils issues.
       new  66c7a14   Cosmetic changes
       new  ab036da   Fix some docstring failures in crossval
       new  0327126   Fix some docstring failures in svm
       new  9b55904   Comment failing doctests.
       new  538bad1   Remove failing doctests.
       new  b137987   Fix doctest in neighbors module.
       new  1a82f5b   Fix doctests.
       new  8fe9b70   Update feature_selection.
       new  6186117   fix doctest issue in crossval (issue #34)
       new  73dc673   Add small script to run all the examples.
       new  29361a3   Add missing directories to distutils
       new  fcea442   Do not run examples from tests in module em.
       new  aadac42   remove iris as python package (it is now just a data directory)
       new  8d6bd0b   Removed test for em/tests/test_densities.py
       new  07262da   ooops, I deleted too much in last commit.
       new  0e1faaf   0.2-beta release.
       new  a5979a5   Add .mailmap file for nice git log messages :-)
       new  9927ab0   Fix image location in doc.
       new  a40d325   0.2 release
       new  50c47e3   Start 0.3 development cycle
       new  24473d3   Fix a bug when using an estimator in UnivSelection. x[self.support_] -> x[:,self.support_]
       new  8ad16aa   Add a function for crossval : leave-one-label-out
       new  574f2b6   fix doctest in crossval
       new  9cbb1c9   MISC: Safety commit
       new  db67917   Fix import path in BallTree benchmarks
       new  3133469   Add another balltree benchmark, this time with plot interface.
       new  2a5eee8   Add a test for OneClassSVN.
       new  8919fa8   BUG: fix for ticket 39: selecting 100% should now select 100%.
       new  929bcd8   MISC: Set svn ignores.
       new  8060c02   DOC: Improve the feature selection example.
       new  f77d9b7   adding UnivSelect object in univ_selection to by pass the use of predictor object. The UnivSelect implements fit + reduce rather than fit + predict
       new  c0a5810   fix univ_selection test
       new  ff91a6d   Fix docstring for svm.SVC
       new  4dd2ea8   Added svnauthors file
       new  c018036   adding leave_p_out in crossval.py
       new  dcda191   DOC: minor spelling mistakes (just to check upload rights and my login)
       new  38782ac   adding myself to svnauthors
       new  25b4b81   Remove mailmap file
       new  e1a519d   Add feature_selection to setup.py
       new  89bf5d1   Remove empty directories
       new  99dda61   allowing Lasso CD + Elastic-Net CD to pickle
       new  28d6893   DOC: extending README with information about GIT Mirror
       new  8632266   DOC: minor changes in README in regard to use of git-svn
       new  e76786e   NF: initial Debian packaging skeleton
       new  866a79f   BF: little fixes around
       new  b21b077   more thorow cleaning, excluding manifold from unittest, unittest before moving libraries
       new  30446f7   various fixes thanks to lintian
       new  c19fe9b   First ever package by me to adhere to dkg-source 3.0 (quilt) format.
       new  b5e024c   Remove old datasets.
       new  1d3da21   ENH: remove scikits/__init__.py upon installation
       new  69d1b08   adding LDA implementation with test + example
       new  98ed29f   adding "proba_predict" method in LDA (should be done for SVC too)
       new  1d9e0aa   BF: adding build-dep on python-setuptools
       new  d305742   BUG: removing evil tab
       new  5bef3e8   BUG: removing add_subpackage for deprecated datasets
       new  d52dd78   BF: remove generated datasets/__config__.py upon clean:
       new  aa50e11   BF: jquery link is relevant for -doc
       new  735b499   BF: adding copyright on sphinxext
       new  a9f7b23   working on dh version of packaging
       new  37b6c95   adding documentation to be installed
       new  d795f6a   might to be complete dh setup
       new  3f21e63   adjustments to debian/control -- cython, descriptions
       new  cf5db96   BF: avoiding nested examples directory
       new  118629b   BF: add scipy to build-dep due to unittests
       new  ad7a81f   RF: build Cython code during build time
       new  a4cbe5d   RF: removing Cythoned C code
       new  918a6b2   RF: removing obsolete .pxd files
       new  dafaa9a   BUG: correct import within test_densities.py
       new  7752c90   RF: use assertEqual to compare str(e) instead of a print (could simply be assertRaises)
       new  d754995   updating changelog for the release
       new  2ae3497   BF: pruning forgotten cdbs build-dep
       new  a5087aa   Remove empty directories
       new  0c969d1   Revert "RF: build Cython code during build time"
       new  4a44ada   Revert "RF: removing Cythoned C code"
       new  e6f424e   Preliminary sphinx example generation
       new  5e692c7   Add directory for auto-generated examples.
       new  e6d2075   More on auto-generated examples
       new  1f8984c   Temporary hack to generate the examples.
       new  c0695e8   Generate examples dinamically.
       new  38864c8   Update documentation.
       new  f82b562   Cosmetic changes to gen_rst
       new  d7cdd0c   Fix typos in examples.
       new  c3d5d96   Add Gaussian Naive Bayes classifier
       new  6fa7e06   fix docstring test in Neighbors class
       new  374c800   cosmit in naive_bayes.py
       new  46d72d1   cosmit in lda renaming method proba_predict to predict_proba for consistency with GNB
       new  1a9a8e4   DOC: Small changes to the documentation building to be able to references examples and to be able to download the source code of the examples.
       new  331516b   DOC: Sprucing up the documentation.
       new  173611d   BUG/DOC Correcting bugs in the documentation generation code.
       new  1df1839   Workaround for proxy problems.
       new  fb816b3   DOC: Adding plotting to feature selection, in 2 times to work around proxy problems (lousy youth hotel in Rotterdam :>)
       new  f81cc2a   DOC: howto upload generated doc to sf.
       new  49148f3   Update README
       new  e64bd21   MISC: Cosmetic commit: lines longer than 80 chars.
       new  2993bec   Preliminary Logistic regression classification using liblinear.
       new  1bd4045   Implent L1-penalized Logistic Regression.
       new  44a5f99   Some bug fixes for liblinear bindings.
       new  6793924   BUG: More bug fixing around liblinear bindings.
       new  6f8a9d8   ENH: Manage special case of 2 classes in LogisticRegression.
       new  1c2aa64   ENH: Refactor the datasets to add the digits standard dataset.
       new  2e49fbe   MISC: Cosmit (cosmetic commit): s/label/labels, because it is a list.
       new  dde6f2a   BUG: Proper description for digit dataset.
       new  6abff58   Preliminary libsvm support of probability predict.
       new  13418b8   Preliminary support for probability estimate in SVM.
       new  422de5f   Turn off probability estimates by default.
       new  c0eb7aa   Implement a test for SVC probability estimates.
       new  75f0b44   Bug fixing in SVM predict probabilities.
       new  1b85122   Use liblinear in SVM module when kernel is linear.
       new  41ceead   BUG: Fix bugs in liblinear bindings.
       new  3f59299   BUG: Fix the digits datasets: wrong data loading.
       new  661af63   MISC: Remove the checked-in generated documentation files from the version control.
       new  c02855e   BUG: Correct bug in digits dataset loading introduced by previous checking.
       new  7e8c34e   correct typo
       new  181e3de   BUG: Make sure the docs build even if the directories do not exist.
       new  d7c07ab   ENH/DOC: Add an example doing classification on digits.
       new  c0fee47   Remove liblinear-related code from SVC.
       new  6c0e3aa   DOC: Update docs README to use rsync and try to be careful about the permissions.
       new  4605098   Implement LibSVC class for liblinear bindings.
       new  f968487   Implement probability predict in Logistic and LinearSVC
       new  5d06eff   Implement probability predict in the case of l1-logistic regression.
       new  46823a3   DOC: Specify git repo in development section.
       new  bb50c1a   Remove empty directories
       new  8303ec0   Fix compatibity issues of bindings in 64bit.
       new  03c55c4   Always use np.float64 for compatibility.
       new  0bd5615   ENH: Make it possible for SVM.predict to work on a single sample.
       new  d09d617   DOC: Improve the example docstrings.
       new  4e7ef58   DOC: Work on the front page.
       new  6d1c9f6   uniforming the access to the primal variables (the "w") with a "coef_" property in the lasso and enet coordinate methods.
       new  18d20ab   SVM : renaming coef_ to dual_coef_ in SVM and uniformisation of coef_ property between SVC with linear kernel and LinearSVC
       new  6f6060b   Logistic : adding properties to get access to coef_ and intercept_ coef_ is now a vector of size n_features (the intercept is left out)
       new  ecb8672   Improve front page
       new  2d43c4d   DOC: Some work on the beginning of the doc.
       new  b1c363a   s/n_features/n_samples/ in digits example
       new  90eb5b7   add some print statements to monitor lasso/enet progress and tweaks parameters to make it faster to run
       new  7cb4c28   Use syste-wide BLAS libraries if available.
       new  1559654   Update svm benchmarks
       new  2df46b2   improve computation of coef_ in SVC
       new  6cd3ab0   DOC/MISC: correct docstring on digits loading code.
       new  7c2434c   adding doc to LogisticRegression class
       new  4827353   update in doc for LogisticRegression
       new  1e45cc1   introducing intercept_ in SVM (from libsvm) intercept_ == - rho_ for consistency with liblinear outputs
       new  a00f908   adding example to view posterior class probabilities when using LogisticRegression or SVC
       new  a9a5003   adding example that illustrates that L1 penalty leads to sparse estimates of coef_
       new  47110f0   DOC: Adapt the getting started tutorial to complete beginners.
       new  7b45bd9   DOC: Moving the em examples. Doing this in two times, to avoid what looks like bugs in the svn sourceforge serve.
       new  e5c769d   DOC: Moving the em examples, step 2.
       new  1896240   DOC: Rework of the docs and examples: multiple commits because commits with 'mv' fail.
       new  fe36e69   DOC: Moving SVM examples in a separate folder, step 1.
       new  03c7d92   DOC: Moving examples in a separate directory, final commit (hopefully)
       new  564495a   adding docstring to LinearSVC and test to make sure the coef_ and intercept_ returned by SVC and LinearSVC are the same.
       new  79f16a0   MISC: Typo.
       new  69905bb   Fix typo in setup.py
       new  1caffe8   Fix a bug in the computation of the log lokelihood
       new  5cb5512   Add coef_ to SVR.
       new  117faf6   adding predict_proba method in UnivSelection class
       new  449cc00   Refactoring on the svm module.
       new  71d58fb   ii
       new  466bfd2   Add more tests to test_svm.py
       new  477aa3e   Add the option to link against system-wide libsvm by editing site.cfg
       new  eb34e43   Remove unused inports from lasso_cd
       new  bd8f100   Add logos (in svg and bmp format)
       new  5e3a7e0   Remove examples from module glm.coordinate_descent.
       new  88ce154   FIX: Compatibility Visual Studio
       new  f3af695   FIX: Fix a bug that caused segfault when intercept=False on classes that call liblinear.
       new  5201738   Do not execute tests in degenerate cases.
       new  675e90e   0.3 Beta release
       new  e23ea16   Add download link to main web page.
       new  9d0fc56   ENH: add docstrings to module svm
       new  4125b93   Remove scikits.optimization as requirement.
       new  8a4bc2f   0.3 release
       new  a3cca44   Start of 0.4 development cycle.
       new  7198eba   Update information relative to git repo in README
       new  37003c2   Merge branches 'master' and 'master' of git://github.com/yarikoptic/scikit-learn
       new  6710890   Merge branches 'master' and 'master' of git://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  831d914   typo
       new  93b54e4   NF: adding a patch to site.cfg to use system libraries
       new  91f6d47   additional build-depends
       new  8d29334   build docs after modules
       new  b75858e   changelog entry
       new  8216797   depend on unversioned libboost-dev
       new  7a09a46   .gitignore: vim power! + generated __config__.py and build
       new  3392829   Typo in README.
       new  ba39488   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  f4401db   Refactor lasso coordinate descent code.
       new  6e21b1c   More work on coordinate descent
       new  e45a40d   more work
       new  7aa056a   more work.
       new  39ba311   more work
       new  38d9be6   Merge branches 'master' and 'master' of ssh://ogrisel@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  e98f230   rename README to README.rst so that it looks nice on github
       new  de9e866   finally working
       new  dc579a4   Update repository location in documentation.
       new  89921b2   more
       new  41f3256   More on updating the doc.
       new  d7fcc72   Add also a link to the repo via ssh
       new  b30abe7   ENH/API: Change the cross-validation utilities from generators to objects that expose an iterator. Also, move the module from utils.crossval to cross_val.
       new  0840398   MISC: Rename *_indexes to *_indices in cross_val.
       new  6557471   more on coordinate descent
       new  847a507   Merge branch 'coordinate_descent_2' of http://github.com/fseoane/scikit-learn
       new  d4c6f8c   Fix broken setup.py in datasets.
       new  ec364fe   Refactor lasso coordinate descent code.
       new  0189d36   Merge branch 'coordinate_descent_3' of http://github.com/fseoane/scikit-learn into HEAD
       new  60fd1f4   Add test directory to setup.py and update dependencies on README.
       new  e7cf7e7   changing interface of lasso + enet for consistency with GLMNET
       new  47eef9a   Updates and fixes for the svm documentation.
       new  72d51be   bug fix in coordinate descent for GLMNET
       new  88e9d3a   Refactor lasso coordinate descent code.
       new  c641642   include the cython generated C code for cd_fast
       new  1e68aac   changing interface of lasso + enet for consistency with GLMNET
       new  dbc2eaf   bug fix in coordinate descent for GLMNET
       new  d45285f   Merge branches 'master' and 'master' of ssh://ogrisel@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  6777523   Merge branch 'coordinate_descent_4' of http://github.com/fseoane/scikit-learn into coordinate_descent_4
       new  8c54eda   fixed tests
       new  4c700ce   PEP8 is beautiful
       new  05499ca   Update references to the svn in the docs.
       new  66016e8   Implement predict_margin in classes that wrap libsvm.
       new  60491b4   Remove references to svn repo.
       new  4a96ecf   Refactor lasso coordinate descent code.
       new  3c46620   include the cython generated C code for cd_fast
       new  fecc1dd   changing interface of lasso + enet for consistency with GLMNET
       new  6d37c66   bug fix in coordinate descent for GLMNET
       new  5c82f77   yet another attempt at early stopping with regularization path on LASSO + CD
       new  4576ff6   note for later
       new  ddbb875   ENH: documentation for coordinate_descent
       new  e783e38   merging from master
       new  414aa79   ENH to LassoPath: cleaner, shorter code and storing path is now optional
       new  07fa138   Merge branch 'master' of ssh://ogrisel@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  b23bf71   Merge branch 'master' into ticket-51-cd-early-stopping_2
       new  c081c4b   removing wrongully added conflict markers
       new  3dab5a1   make fit work on python tuples and lists
       new  8b95406   spare one local variable
       new  8a0df7a   oriented programming magic to factorize common logic between Lasso and Elastic Net
       new  f3627c5   some tests for the elastic net path-based solver
       new  0325be3   minor changes in descriptions
       new  d498e80   Setting both HOME and MPLCONFIGDIR while building documentation (slipped through fingers somehow) (Close: #580727).
       new  c7208c1   Merge branch 'debian' into debian-release0.3
       new  f4b1d21   Set HOME and MPLCONFIGDIR while unittesting as well (Closes: #580727)
       new  47890ac   Merge branch 'debian' into debian-release0.3
       new  b570e28   merge origin
       new  c5000c4   refactoring to avoid writing model_params in path object construction
       new  1be2188   removing unnecessary imports
       new  a676ee6   adding __str__ method in LinearModel. adding examples for lasso + elastic + path with crossval. adding computation of explained variance (ie. r^2). disabling early stopping in paths as it turns out on an example that it tends to stop the computation too early
       new  265c6ec   bug fix in elastic-net dual gap computation
       new  cc91603   fix docstring test in cross_val LeaveOneLabelOut
       new  5c7200e   adding __str__ method in LinearModel. adding examples for lasso + elastic + path with crossval. adding computation of explained variance (ie. r^2). disabling early stopping in paths as it turns out on an example that it tends to stop the computation too early
       new  020ff44   Add generated cython code for cd_fast.
       new  cb6a56f   Implement predict_margin in classes that wrap libsvm.
       new  aa73508   Add tests for predict_margin.
       new  591c2f6   BF: enforce little-endian byte order for loadmat call in test_gmm_em (Closes: #580879)
       new  6f42e64   Disabling test_gmm_em unittests while building (Closes: #580879). Actual reason for failing tests lies within NumPy -- see #581043, which should be fixed there, but should not prevent building of this package atm.
       new  b573fc0   Merge branch 'debian' into debian-release0.3
       new  57cdf8d   BUG: fix bug in tutorial
       new  f642719   Add support for precomputed kernel.
       new  2849920   BUG: fix bug in automatic doc generator.
       new  bf3e14a   Update doc.
       new  a777d4a   Update test after solved bug in elastic net path.
       new  4beb08c   adding possibility to recover the length ie the number of folds of a cross-validation object
       new  ad0ad0f   calculate likelihood of the GMM model in the log dimension
       new  0276741   API: Remove the obsolate attrselect.py module.
       new  6f38737   A bit of documentation on svm precomputed kernels.
       new  de88b27   Welcome Chris.
       new  4f8991c   Update docstrings in svm module.
       new  8143abb   Add a complete example of SVM classifier using a precomputed kernel.
       new  278a05f   Update examples.
       new  ea74be3   Give appropriate substitutes for modules not in python 2.5.
       new  2475c22   Updates for the doc.
       new  5ffb29c   More documentation from BallTree object into sphinx documentation system.
       new  4b08a2c   Add some notes on how to run the development version.
       new  6bf88a5   Remove svn_ignores
       new  354bd20   Welcome Angel Soler.
       new  a02233d   fix typo
       new  f0307ed   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  59df5be   ENH: adding possibility to compute confusion matrix
       new  2d23dc2   ENH: adding possibility to compute ROC curve
       new  a647fb4   ENH: adding possibility to compute precision-recall curve
       new  8ac9d24   fix name conflict issue with stdlib math module
       new  e152106   moving ROC etc. to metrics.py adding tests for metrics.py adding examples for ROC, AUC, precision recall, confusion matrix
       new  04333d0   Adding myself into AUTHORS and fixing up atypo
       new  f1c3915   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  f008ebe   cleaning up my mess...
       new  67338d0   Fixes in Ordinary Least Squares (glm module)
       new  20006d0   Refactoring: Squash samples_generator module into one file.
       new  18f2b9f   More docstrings for module glm.regression.
       new  0462355   Add an optional argument intercept to Ordinary Least Squares.
       new  1c82d45   Remove outdated script.
       new  d3ce932   Fix broken link in rst feature_selection.
       new  edbe976   adding references to wikipedia pages in metrics.py
       new  cc5d7de   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  06ef709   Remove unused parameters from objects that depend on libsvm and liblinear.
       new  37e7d69   Removed unused doc.
       new  738a2fb   DOC: more documentation for svm
       new  e9c71c9   Implement class weights in classes that query libsvm.
       new  6d5f102   Fix imports in glm test
       new  349b4bf   Cometic: Change name of weights parameter in Libsvm.fit
       new  73079dc   Simplify example svm_hyperplane
       new  6fcffc3   Add an exmple of svm using weighted classes
       new  1a52195   Rename plotted examples.
       new  5771ac4   DOC: fixes and enhacements for the docs.
       new  c20ad73   Add title to examples
       new  9ab8276   FIX: fixes in svm predict_proba
       new  46e17e2   updating ROC example and adding an ROC example with cross-validation
       new  0e911c8   BUG: fix in precision recall example
       new  662501b   BUG: fix in test metrics
       new  2f4844c   Cosmetic changes to the svm tests.
       new  990bad7   make the LogisticRegression fit return the classifier instance for chained calls
       new  8725a89   make the LinearSVC fit return the classifier instance for chained calls
       new  6797b3d   make loss and penalty notation casing consistent accross scikit.learn (LinearSVC and LogisticRegression)
       new  1d3754b   merge master heads
       new  158cda2   cleanup conflict marker leftover in test_svm.py docstring
       new  4fe5e40   FIX: make the window_size parameter taken into account in the BallTree implementation of KNN
       new  c77a7a7   Add intercept to classes Lasso and ElasticNet
       new  29b4faa   Cosmetic changes in SVM doc.
       new  d107671   Merge branch 'master' of /Volumes/DAVID/scikit-learn
       new  2f34950   BUG - LDA, separate out "priors" possibly estimated from the ones specified in constructor
       new  1386a22   extending LDA unittest to catch issue fixed with previous commit + 1 more (just in case)
       new  5517e09   pylint friendliness -- spaces before commas, use warnings, etc
       new  3d6e835   Initial import of Ron Weiss' gmm module.
       new  f326a8b   Add intercept to classes Lasso and ElasticNet
       new  360c4d7   Cosmetic changes in SVM doc.
       new  a7b77e6   Fix failing tests because of added intercept.
       new  3b211b5   Remove property _ndim (replaced with field ndim) in gmm
       new  db526d2   Rename train --> fit in gmm
       new  531e111   Welcome Ron Weiss.
       new  e34e541   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  71962a0   Some docs on contributing to the project.
       new  851e95c   API: changing the way the parameters of Lasso+E-Net are optimized
       new  d76ecb8   ENH : imroving documentation of lasso + enet paths function
       new  e25782a   API: changing the way the parameters of Lasso+E-Net are optimized
       new  03a5c15   ENH : imroving documentation of lasso + enet paths function
       new  8ac14a0   Doc and cosmetic changes in coordinate_descent.
       new  b3b1fee   Refactoring: squash glm module in a single file.
       new  b85d91c   Initial import of LAR algorithm.
       new  f873b18   Fix blas build for liblinear when no system-wide blas is found.
       new  229542d   LARS test clean-up
       new  1657eca   fix module path in pickled samples for manifold learning
       new  c1ce73c   Add .mailmap file.
       new  ea7c8ec   Fixes for LARS algrithm.
       new  28c57f7   Refactoring in glm module.
       new  d8fc705   Remove logging from gmm module.
       new  c45e7d9   Add disclaimer and some comments to gmm module.
       new  2a03d3b   Refactoring in gmm module.
       new  ab705b6   Update API docs.
       new  bf762ce   Update contribute doc.
       new  725a93b   Update doc in glm module.
       new  7e410ec   add note for IRC chan
       new  9a56bc1   Some fixes for LARS
       new  e4191a5   Removing code in __main__ for univ_selection
       new  9cbd1cd   Fixes for LARS algorithm.
       new  dea2b85   Remove duplicated files.
       new  b0dea39   Add compute values of lambdas in LeastAngle Path.
       new  a3e7959   Some fixes for LAR and added example.
       new  022873c   Cosmetic changes to LeastAngleRegression.
       new  9a08fe3   Cosmetic changes in LAR.
       new  cbe1a91   Cosmit
       new  8d35dd5   Enhance plot_classification_probability example.
       new  3af8048   Merge branch 'master' of ssh://gvaroquaux@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  6ce25e0   Fix import path in examples.
       new  035288d   typos
       new  3d36704   Beautify the SVM docs
       new  47dbd45   Merge branch 'master' of ssh://gvaroquaux@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  17f7be4   More beautify
       new  e95bc43   Add the diabetes dataset and modify the LAR example.
       new  01f728d   LAR segfault fixes and cleanup.
       new  5cb6042   Update benchmark to stay in sync with pymvpa developement.
       new  86168f6   Typo in glm.
       new  7705feb   ENH: new univariate feature selection code
       new  98138f6   cleaning plot_lar example
       new  c3a7e2d   API: renaming LassoPath to LassoCV and ElasticNetPath to ElasticNetCV
       new  8aea220   Fixes for lars.c to compile under Visual Studio.
       new  9c0408a   LeastAngleRegression fixes & refactoring.
       new  5e78778   Call gc explicitly in the benchmarks to avoid unnecessary noise.
       new  46f3faf   Initial import of shortcuts module.
       new  f84bce9   Cleanup in LAR example
       new  d3e537f   GMM refactoring.
       new  7ada25a   Example on plotting a gaussian mixture.
       new  d293c6f   Remove manifold module.
       new  3f8552d   Update tests (fix broken tests on 64 bits).
       new  6dddf51   Remove the shortcuts module.
       new  47ef169   Rename libsvm --> _libsvm
       new  3382ec2   Rename module BallTree -> ball_tree
       new  30d5a4d   Use np.allclose in gmm instead of custom functions.
       new  75dbbfa   better tests for LeastAngleRegression
       new  f155c32   New univariate feature selection.
       new  74ceb89   Remove outdated doc.
       new  3a4ecf2   Update doc with new gmm module.
       new  02d0d0e   Normalize in LAR fit
       new  8d1ed42   FIX a test.
       new  e3a75ec   test to reproduce issue 67 on LARS coef shape
       new  40dcb58   Merge branch 'master' into issue-67-LARS-shape
       new  dfa60cd   FIX: typo in setup.py
       new  a380be8   Add  support function for computing the selected features.
       new  03ea91a   Rename C extension liblinear -> _liblinear
       new  2740b6d   Rename minilearn -> _minilearn
       new  c328ee1   More doc on contribute.
       new  50d070d   Add reference to mailing list in main page.
       new  27d3900   Refactoring in the glm module.
       new  6fe856c   Add more tests for LeastAngleRegression.
       new  0ff244c   Small modification in the LARS algorithm.
       new  0529f37   FIX: intercept in ElasticNet
       new  784e1fe   Use default values in lasso example.
       new  af0a717   GLM refactoring: put explained_variance_ as a public field.
       new  d2f88ad   Some fixes to make it compile against MKL library.
       new  2bac216   tracking changes from master
       new  a7843b9   follow API change in LARS
       new  8936024   Updates on gmm example: cleanup.
       new  3f5f7c0   Update example gmm.
       new  cc42768   Fix broken links in doc.
       new  32f4c67   Add an example with probability distribution estimates using GMM.
       new  1a17e79   Remove examples from deprecated modules.
       new  60c047c   Remove duplicate module univariate_selection.
       new  6b05414   Remove old em module.
       new  59db7f0   Initialize with a fixed seed tests in feature_select
       new  dca0ff8   Remove redundant seed generator.
       new  69e22c7   Add docstring to UnivariateFilter.
       new  53daf5e   Comment out broken parts of the feature selection example that are broken.
       new  742db04   Add a __version__ attribute.
       new  54a4be5   Remove also auto-generated examples with make clean
       new  860f009   Rename examples that do not plot results.
       new  97a2ec2   Move glm examples to common directory.
       new  e356827   Remove benchmarking code from lasso example.
       new  a446edf   Add title to example.
       new  d85c81d   Remove code from roc plotting that was interfering with build system.
       new  65d06f8   0.4 release.
       new  9ded72a   Start of 0.5 development cycle.
       new  5b397fa   Re-enable code that was removed for the release
       new  8945d3d   Cleanup gmm example. Removed unused modules.
       new  23125ea   Merge branch 'master' into issue-67-LARS-shape
       new  db687ec   ENH : add LeavePLabelOut cross-validation generator
       new  f9a8a95   In LAR, normalize only non-zero columns.
       new  c3b3ea9   Add support in LAR for unnormalized columns.
       new  3a59f74   LAR: add a test for zero coefficients.
       new  8722352   Merge branch 'master' into issue-67-LARS-shape
       new  3eb4f5a   make sparse coding test pass
       new  b3887c5   ENH : adding support for mean-shift clustering with a flat kernel
       new  18e8913   Merge commit '0.4' into debian-release
       new  9493856   changelog: Fresh upstream release + under ND umbrella
       new  dc10cf7   adjusted patches/deb_use_system_libraries
       new  95c2aa9   no boost library used any longer (removed from Build-Depends)
       new  f8508e5   oops - fixing up patch for site.cfg
       new  31b2ebc   accounting for rename of README to README.rst
       new  9072aa5   Merge branch 'debian' into debian-release
       new  42cec21   Cosmetic changes in glm module.
       new  c2145fb   Add modules to top-level __init__.
       new  dc42419   more .gitignore
       new  4449afd   added hmm code from http://github.com/ronw/gm
       new  e25e75f   Rename ninter --> n_iter in the API guidelines.
       new  27aa259   removed logging, dependency on abc, and unnecessary imports
       new  c7a2791   added hmm unit tests
       new  8272465   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  3e7af84   started work on document classification: bag of wordsw extraction and hashed tfidf
       new  fb297e6   some tests for the text features extractor
       new  7759f8b   checkpointing work in progress on MLComp dataset integration
       new  ea20d1c   remove labels handling from vectorizer code
       new  cba81c1   more work on document classification dataset loader
       new  a50bc25   smaller default dim: faster to load by default, need experimental setting to find good tradeoff
       new  6f6b2f8   make it easy to find the raw source document
       new  c9f1b61   better parameter ordering
       new  434eed6   example usage of MLComp document classification datasets
       new  a45a01e   use compiled re pattern
       new  5d2a7fd   small fixes
       new  2f17c0a   MISC: Make sure that the tests pass on numpy 1.2
       new  5735f26   cleanup hmm module: made properties compatible with Python 2.5, etc.
       new  b10ad1c   changed hmm.trainer usage: each hmm object must have a _default_trainer property which can be overridden by passing a different trainer into hmm.train()
       new  26ca669   changed "train" -> "fit".  Removed HMMGMM for now.
       new  c3d66b6   removed references to gmm.init() in gmm docstrings
       new  3cb6f7a   fixed random seed in hmm unittests
       new  56dc4d7   removed init() method from hmm classes
       new  55ea905   minor tweaks to make hmm.GaussianHMM look like gmm.GMM
       new  429149e   Merge branches 'master' and 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  70be452   fixed bug in HMM viterbi logprob
       new  834aeba   typos
       new  4cc088d   added MultinomialHMM, unit tests
       new  96d8bee   fixed *HMM.fit() to include *all* parameters by default
       new  7b9ecbf   removed ndim argument from gmm.rvs()
       new  6a981e2   added validate_covars back to gmm.py
       new  b5f0228   added ndim property back to gmm to keep it consistent with HMM
       new  95df2dc   Merge branch 'hmm'
       new  7948ad6   added support for HMMs with GMM emissions
       new  abe8330   Merge remote branch 'upstream/master'
       new  41b0562   add the ability to use stop words for text classification, but does not improve accuracy hence not enabled by default
       new  b77a973   typo in comments
       new  d025a71   faster and better accuracy with hinge loss of doc classif example but not sparse anymore since l2 reg...
       new  fdfe084   ENH: making data contiguous in memory in coordinate descent
       new  622ee91   ENH: adding affinity propagation algorithm moving clustering methods in clustering.py
       new  152487e   removing pl.show()
       new  6e7a326   ENH: adding exception raising
       new  1082fa4   Added description of input parameters in svm.SVC docstring
       new  8cf4148   Add documentation to svm mdoule.
       new  09ccaab   FIX: bug in blas_opt detection.
       new  1d117cc   make the features package a first class citizen
       new  d672980   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  a22cbe2   Link against compiled cblas in case this is not in the system.
       new  e6a14c2   Bug fixing in setup.py
       new  750f44e   Apply changes made by Olivier.
       new  d404072   Merge branch 'master' of github.com:agramfort/scikit-learn
       new  a251463   using staticmethod rather than property
       new  2bb1340   cosmit
       new  e788f11   setting array as fortran in lasso + enet coordinate descent
       new  f497c99   MISC: Comsit + replace some global seeds with RandomSate
       new  0d8f1ca   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  06911df   MISC : renaming affinity propagation example
       new  a3f4f6b   MISC: Rename to let the underscore RULE!
       new  59fa8e9   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  d5505fc   API: Create a base estimator class.
       new  da3a328   ENH: improve base class
       new  df165d3   ENH: Temporarily remove the typing for the base_estimator
       new  8e6a9ca   broke glm to improve model selection
       new  5fb7223   TEST/BUG: Test the BaseEstimator class and fix the repr
       new  4a20e78   Merge branch 'master' of http://github.com/agramfort/scikit-learn
       new  d9f5c76   Cosmit
       new  6bd895a   MISC: Remove the #$! import *
       new  a178da5   ENH: convert all GLM estimators to the BaseEstimator class
       new  232bfd4   BUG: Fix the OLS regression
       new  8ec5898   ongoing work on glm with crossval
       new  8a0e8a2   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  3978883   BUG: Fix constructors with arguments.
       new  ad1fdec   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  1b92a48   BUG: Syntax error
       new  c9f8c0c   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  9480379   BUG: str of linear models now working.
       new  6575711   API: Change the type of params: turn this into a frozenset: unmutable and matches better our use than a dictionary
       new  65294eb   continue improve glm cv
       new  164bf65   Merge branch 'master' of http://github.com/agramfort/scikit-learn
       new  d93be9b   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  f43f933   ENH: Change _params to frozenset
       new  3818399   API: Change argument controling whether intercept should be fitted or not in GLM to 'fit_intercept' in order to disambiguate with the value of the intercept (intercept_)
       new  36c6bf8   fix glm cv
       new  8ca262f   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  e83d972   Cosmit in tests
       new  2eac428   fix frozenset
       new  391bdd5   Merge branch 'master' of http://github.com/agramfort/scikit-learn
       new  be82ca3   API: Change the BaseEstimator and parameter signature logic.
       new  5d7d530   Cosmit
       new  e0872ef   Cosmit
       new  8856323   ENH: Convert LDA and clustering to use the new BaseEstimator
       new  d947fd6   MISC: Change the title of the documentation.
       new  c7793cd   Cosmit
       new  19442b0   ENH: Make the clustering more usable
       new  6905e38   ENH: Add an example of playing with the stock market
       new  611554c   ENH: more efficient stopping criterion coordinate descent GLM and comparison with python-glmnet
       new  5889ff1   BUG : fix in affinity propagation API : affinity propagation now returns the indices of cluster centers
       new  5a5d201   BUG : fix in stock market example
       new  0fd05c8   BUG : fix with blas on mac os x
       new  7a344ac7  ENH : moving bench_glm.py to benchmarks folder
       new  1132797   ENH: Make SVNs fit to the  BaseEstimator API.
       new  241e080   ENH : glm coordinate descent with BLAS
       new  b4976fe   Added deb_cython_0.12.1 to use code cythoned with up-to-date cython 0.12.1
       new  f6e9a2a   Build-Depend on python-numpy providing numpy-ext (Closes: #589590) + policy 3.9.0
       new  fd50b37   make nosetesting not suppressing output from code
       new  ed48e90   adding a workaround to allow BFS (Closes: #588595)
       new  9eff1da   Changelog for all previous commits
       new  91dfe47   actually doing arith operation and printing to avoid failed comparison
       new  2b8ddd8   Merge branch 'debian' into debian-release
       new  fcf7286   adjusting copyright file to actually address 3.9.0 policy + moving packaging under BSD
       new  35f241e   adjusting changelog
       new  959e267   Merge branch 'debian' into debian-release
       new  9c20589   blas-ification of elastic net + ensure that the gap is initialized and evaluated
       new  2ad4f95   BUG : fix blas support in setup.py with coordinate descent
       new  3c92c56   One more tests on LibSVM with precomputed callable kernel.
       new  21b6ed1   ENH: Make SVNs fit to the  BaseEstimator API.
       new  4f31787   Refactoring of LibSVM bindings.
       new  fb91d9a   Test for libsvm margin.
       new  ec5eca3   More bugfixes for blas detection in setup.py
       new  8727393   FIX: numpy 1.4 fixes.
       new  3867063   Mark as known to fail some tests in test_hmm.
       new  92eea9c   Use numpy.testing instead of unittest to skip failing tests.
       new  cb9e15e   ENH : adding stratified cross-validation object
       new  c704e3d   ENH : fix doctests in glm, svm and lda
       new  524b532   Cosmit
       new  d7e6408   MISC: Put the nearest neighbors estimator to the BaseEstimator convention.
       new  52feded   MISC: rename base_estimator.py to base.py
       new  965c2c3   Refine cblas detection on OSX.
       new  31b36e8   FIX: compatibility fixes for py3k.
       new  2fd21d5   Initial support for sparse matrices in SVMs (scikits.learn.sparse.svm)
       new  f27f234   BUG: Make sure that the docs still build with recent versions of numpy and Matplotlib
       new  51543b1   BUG: Make sure the docs still build with recent versions of numpy
       new  835de69   fixed GMM examples
       new  39ac7d6   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  78ea762   BUG: Adapt the sparse SVM to the rename of base_estimator.
       new  675967a   MISC: Remove warning when compiling docs.
       new  450b419   MISC: Adding titles to examples.
       new  bf2f812   DOC: Document best practices/coding guidelines to make it easier for people to contribute.
       new  8631c5e   adding grid search code
       new  a2c98ee   Cosmit
       new  0787e25   MISC: Put the nearest neighbors estimator to the BaseEstimator convention.
       new  c6b9b28   MISC: rename base_estimator.py to base.py
       new  8e44322   Refine cblas detection on OSX.
       new  2490597   FIX: compatibility fixes for py3k.
       new  89cad61   Initial support for sparse matrices in SVMs (scikits.learn.sparse.svm)
       new  c14be56   BUG: Make sure that the docs still build with recent versions of numpy and Matplotlib
       new  b5fd8ae   BUG: Make sure the docs still build with recent versions of numpy
       new  735e11d   fixed GMM examples
       new  8ad8771   BUG: Adapt the sparse SVM to the rename of base_estimator.
       new  619a64d   MISC: Remove warning when compiling docs.
       new  a45354e   MISC: Adding titles to examples.
       new  e8c77c0   DOC: Document best practices/coding guidelines to make it easier for people to contribute.
       new  b5da779   BUG : fix doctests in neighbors
       new  94b67bf   BUG : fix doctest in datasets/base.py
       new  b99c765   add pipeline
       new  8ada0f2   ENH : using digits in grid search example
       new  069a945   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  0562558   Cosmit
       new  73d4685   ENH: Make the grid_search take instances of estimators rather than factories.
       new  b752013   API : renaming GridSearch to GridSearchCV BUG : fix grid search example
       new  af5d0ab   Add a setup.cfg to specify default nosetests behavior.
       new  e5bb430   MISC: 80 character bordel!
       new  b17d5fc   New feature selection
       new  182690e   WIP pipeline
       new  7db08f5   Last version of univariate selection
       new  92b11aa   Merge branch 'master' of git at github.com:vmichel/scikit-learn
       new  c5ef47d   Example of feature selection pipeline
       new  7bd4b98   Merge branch 'master' of github.com:vmichel/scikit-learn
       new  4f0adf7   Add a setup.cfg to specify default nosetests behavior.
       new  b880e2c   MISC: 80 character bordel!
       new  c1f222e   API : cross val generator in now given in fit in grid search object
       new  4b81ba1   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  c75f88e   Cosmit: rename grid to iter_grid
       new  9cbdfb8   Corrections of indentation in univariate_selection
       new  af76054   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  61f7c42   remove old univariate_selection
       new  d1baaf5   ENH : update grid search example
       new  12ccf23   DOC: Beautify example
       new  a2ca1e9   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  2a448f8   Correct nosetets for univariate_selection
       new  dcddaa5   MISC: Beautify examples.
       new  f7950c8   Cosmetic on Pipeline
       new  2646957   Merge branch 'master' of github.com:vmichel/scikit-learn
       new  2453a0a   Add doc to univariate_selection
       new  7a4e4f4   Merge branch 'master' of git at github.com:vmichel/scikit-learn
       new  dd92df6   FIX: bug fixing on sparse.svm.
       new  8f1c3fc   Cosmit
       new  c83447b   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  cd73b6a   Merge branch 'master' of git at github.com:agramfort/scikit-learn
       new  314ba8d   Merge branch 'master' of http://github.com/vmichel/scikit-learn
       new  e8cc2ff   ENH : first draft of RFE
       new  bcc3c85   Merge branch 'master' of git at github.com:agramfort/scikit-learn
       new  aaecda1   Add rfe example
       new  7aafa3d   ENH : fix RFE + example
       new  fb7b47b   update example
       new  e29b629   ENH : improve RFE
       new  4dccda5   Merge branch 'master' of git at github.com:agramfort/scikit-learn
       new  120b1bd   Merge branch 'master' of git at github.com:agramfort/scikit-learn
       new  ab1220c   Corrections in rfe
       new  2d63185   Remove feature selection
       new  71c1b24   Add ranking_
       new  cf2dc37   Add Crossvalidated version of RFE
       new  c346f8a   ENH : adding loss functions in metrics.py
       new  026a766   Merge branch 'master' of github.com:agramfort/scikit-learn
       new  7390148   Add example of RFE CV
       new  dd68397   Merge branch 'master' of github.com:agramfort/scikit-learn
       new  d5f9eb9   ENH : fix RFE and RFECV
       new  811563a   FIX: more bugfixing in sparse.svm.
       new  c3e7b53   Doc updates to the svm module.
       new  0ea8827   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  9dcd719   ENH: optional computing of estimated covariance of LDA classifier.
       new  2e10705   MISC: add an unfinished toy example to compare LDA with a (not yet implemented) QDA.
       new  828c5b5   ENH: rework univariate selection to reach a compromise between ease of use and flexibility.
       new  05ad102   ENH : allow grid search to work with lists of grids
       new  fe9eaf5   ENH : using BaseEstimator with GNB
       new  fcf13ec   cosmit'
       new  e08e595   ENH: First go at a help for cross-validated evaluation of a score.
       new  4d40365   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  c77024d   ENH : adding BaseClassifier and BaseRegressor base classes
       new  495fb3a   ENH : using mixin rather than base class to bring score methods to estimators
       new  b2f3f61   Merge branch 'master' of http://github.com/agramfort/scikit-learn
       new  faeb984   ENH: Add an example showing the dependency of SVC+Anova on the number of features selected.
       new  94ec683   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  2bfe517   BUG: Fixed example after last API changes
       new  01886dc   BUG: add missing call to pylab show function
       new  b87d10a   BUG: Fixed pipeline feature selection example after last API changes
       new  a3bd419   ENH: Add joblib as a bundle dependency.
       new  e157a45   BUG: Fix doctests in pipeline.py
       new  9b220b8   BUG: Fix doctest in GMM
       new  bd6e1e1   ENH: Add script to update joblib dependency
       new  782aa4a   Cosmit: rename MixinClassif to ClassifMixin
       new  a48f922   ENH: Make sure that in cross_val_scores the StatifiedKFold is used only for classifiers.
       new  7d72553   MISC: lda: Y -> y
       new  e6272b0   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  079c7c5   BUG: Fixed example after last API changes
       new  a3d0a29   BUG: add missing call to pylab show function
       new  40cd7a9   BUG: Fixed pipeline feature selection example after last API changes
       new  8c7b1bf   ENH : fix in svc.coef_ + cosmit
       new  d2f8923   cosmit
       new  e1d9796   work in progress on sparse vector extraction for document datasets
       new  c36e0f7   exclude scikits.learn.external package from top level nosetests env
       new  5c2298e   missing pl.show() in rfe examples
       new  c6c91f2   more missing pl.show() in examples
       new  13b1fc7   ENH : fix in svc.coef_ + cosmit
       new  d8815c5   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  05db1fc   MISC: Small change to contribution guidelines, suggested by Mathieu Blondel.
       new  3b9b8b4   using a separate class for the sparse version of the hashing vectorizer
       new  4efdd77   readd the dense version of the vectorizer
       new  d53a668   checkpointing work in progress on the sparse version of the document vectorizer
       new  6433cee   more scalable TF-IDF computation unfortunately using a python for loop
       new  8c61c97   new example to demonstrate sparse TF-IDF + sparse SVM on 20 newsgroups (too slow right now)
       new  b97e07c   Merge branch 'sparse-documents'
       new  c0047b5   Restrict to python >= 2.5 to avoid dealing with --install-layout absent in 2.4
       new  13b319c   avoid useless allocations in dense_to_sparse conversion
       new  23811af   COSMIT: For the sake of underscores
       new  619821a   MISC: comment
       new  31bd96c   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  fdea794   experimenting with character n-grams features (basic morphological analyzer)
       new  ddc2791   ENH: Add parallel computing for cross validation.
       new  0efc0ba   BUG: Get the BaseEstimator to work even if there is not __init__
       new  8bf6a74   ENH: Make the parallel cross validation more efficient.
       new  f7c215e   BUG: Fix cross_val_scores for unsupervised problems.
       new  982c3a0   ENH: Improve cross_val in parallel
       new  79d634e   Merge branch 'cross_val_gael' of git at github.com:GaelVaroquaux/scikit-learn
       new  d5452da   Misc
       new  70cd263   ENH: Improve the repr for the BaseEstimator
       new  c076dd0   Accepting Nico's patch to account for NumPy transition
       new  8d975f2   changelog entry for -3
       new  f13dba1   Merge commit 'debian/0.4-3' into debian-release
       new  a2fefd4   ENH : using np.logspace instead of np.linspace in paths
       new  8e4290f   ENH : using np.logspace instead of np.linspace in paths (after merge)
       new  f36fdb7   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  f5db5f5   fix one simple import in doctest for GMM
       new  c2c76a7   simple Makefile for repetitive dev tasks on POSIX OS
       new  466b2b4   fix broken/unstable sparse SVM tests
       new  e296f11   even more stability fix for sparse SVM
       new  55d49ba   fix broken doctest in HMM
       new  0624f49   disabling broken doctest in Gaussian Mixture Models
       new  181fdc2   fix HMM doctests
       new  ae4e81e   cosmit
       new  0a5ad9e   ignore coverage output folder
       new  aff2b96   trailing spaces
       new  7ffab5c   skip remaining failing tests in HMM test suite
       new  0236687   fix inline comment
       new  299a8b8   API : making Y optional in fit for OneClassSVM ENH : new OneClassSVM example
       new  b314cc3   FIX : removing duplicated example
       new  ff989b8   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  497f1fd   ENH: add QDA classifier, some docs, examples and tests. LDA has been reworked a bit to follow the API of QDA and avoid useless operations.
       new  64d4a19   Merge branch 'master' of http://github.com/GaelVaroquaux/scikit-learn
       new  1a9cbeb   Merge branch 'master' of git://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  49bd6c3   Merge branch 'gael' of http://github.com/agramfort/scikit-learn
       new  8eae7e9   BUG: Fix a bug preventing from LinearModelCV to print.
       new  8c6293a   ENH : new SVR example
       new  8257fb1   ENH : new SVR example
       new  6b218d1   ENH: optional computing of estimated covariance of LDA classifier.
       new  4451bba   MISC: add an unfinished toy example to compare LDA with a (not yet implemented) QDA.
       new  ec65ba2   MISC: lda: Y -> y
       new  511bee7   ENH: add QDA classifier, some docs, examples and tests. LDA has been reworked a bit to follow the API of QDA and avoid useless operations.
       new  9769090   Misc
       new  7281b3c   Merge branch 'temp'
       new  9617c98   ENH : improve QDA (taken from Matt Perrot)
       new  b110ebb   ENH : improve LDA (taken from Matt Perrot)
       new  92b8b3a   ENH : improve LDA QDA example (taken from Matt Perrot)
       new  9ab5bd5   MISC: cosmit in LDA, QDA
       new  ac5184a   ENH : new example for LDA vs QDA
       new  26cdfa4   ENH : removing old  example for LDA vs QDA
       new  c16e867   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  69dd905   BUG: Fix forgotten import in example
       new  894c26e   ENH : attempt to have a default parameter for bandwidth in MeanShift algorithm
       new  976f6e1   Merge branch 'master' of git://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  858da50   cosmit in LDA/QDA
       new  36604ca   MISC: Remove pointless ellipsis directive (doctest)
       new  7b2f998   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  e05dc94   MISC: vectorize priors computation for LDA and QDA
       new  58efcf4   Merge branch 'master' of ssh://revilyo@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  f959fd3   MISC: remove debug
       new  e5f56af   Remove unused imports in qda module.
       new  72682f0   Some doc for the svm module.
       new  41d842f   Add target in to Makefile.
       new  07bc525   Fix names and missing parameters in LinearSVC.
       new  48be43b   Add support for sparse matrices in liblinear bindings.
       new  b4102d5   Add a reference to density estimation in GMM docs.
       new  b8d080e   Use relative imports inside scikits.learn.
       new  89f6a7f   Remove unused imports from hmm module.
       new  da9edc5   Showcase the new LinearSVC wrapper for with sparse liblinear bindings in the 20 newsgroups document classification example
       new  e150a68   tracking changes in master branch
       new  ad360fc   BUG: Fix lda and qda on 64 bits.
       new  acd12ec   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  2fa7cb0   Merge branch 'master' of ssh://gvaroquaux@scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  82ccb9c   TEST: Make sure that doctests for bundled dependendies pass.
       new  0e1b792   Refinement and bugfixing in the liblinear bindings.
       new  263a454   More refactoring and bugfixing with liblinear.
       new  28ca33e   More refactoring in libsvm + liblinear.
       new  b993498   remove unused imports from setup.py
       new  95b90a3   run all tests suite through nose.
       new  db3a4f3   move liblinear into its own folder
       new  947691c   Bug fixing in liblinear bindings.
       new  e1531e1   Added some failing tests.
       new  9835d5e   Bug fixing in liblinear bindings.
       new  dbe6579   XFail tests that fail (or are plainly wrong).
       new  c854e8a   Refactor layout of developer docs.
       new  cd84d91   Revert unwanted changes (aka ooooops!).
       new  17c908d   Added tests to trigger failure on classes using liblinear.
       new  8bf9ae3   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  81c2b0a   BUG: Make sure that joblib does get installed.
       new  a0b508c   ENH: Make sure tests get installed.
       new  e5aa871   ENH: Improve the repr for the BaseEstimator
       new  5779fcb   TEST: Make sure that doctests for bundled dependendies pass.
       new  baaa94f   BUG: Make sure that joblib does get installed.
       new  505738d   Fixed a typo
       new  de451f7   Refinement and bugfixing in the liblinear bindings.
       new  2c57f55   More refactoring and bugfixing with liblinear.
       new  efc32a5   More refactoring in libsvm + liblinear.
       new  285d7f2   remove unused imports from setup.py
       new  2c8188f   run all tests suite through nose.
       new  63ec11e   move liblinear into its own folder
       new  1e1452d   Bug fixing in liblinear bindings.
       new  3b0e439   Added some failing tests.
       new  49ba346   Bug fixing in liblinear bindings.
       new  3d0f46d   XFail tests that fail (or are plainly wrong).
       new  8509088   Refactor layout of developer docs.
       new  29f7c8d   Revert unwanted changes (aka ooooops!).
       new  ace4884   Added tests to trigger failure on classes using liblinear.
       new  3a67251   ENH: Improve the repr for the BaseEstimator
       new  668094e1  TEST: Re-enable external tests.
       new  f2273c5   BUG: Fix doctests to account for change in BaseEstimator repr
       new  e9f6767   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0e944a5   Update the developer docs.
       new  4e79a78   Refactoring & bug solving in liblinear.
       new  037c769   fix broken test for text features extraction
       new  2dec81d   fix broken test for text features extraction
       new  582c812   FIX: fix liblinear predict in the multiclass case.
       new  b3901e1   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  da623db   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  e047f16   Merge branch 'cross_val_gael' of git at github.com:GaelVaroquaux/scikit-learn into cross_val_gael
       new  e9e4a83   Add reference to pybrain from the ann docs.
       new  2221881   Update numpydoc (sphinx extension).
       new  3ae5b86   Update svm rst doc.
       new  5eef725   Add rst doc for logistic (empty for now).
       new  258b351   FIX: fix shape of support vectors in liblinear sparse.
       new  dd37fb3   fixed broken doctests in gmm.py
       new  714f61b   updated gmm.py to comply with scikit-learn API.  fixed pep8, pyflakes errors
       new  3ed43f0   DOC: fixed typos in developer documentation
       new  10b2fef   Merge branch 'master' of ssh://scikit-learn.git.sourceforge.net/gitroot/scikit-learn/scikit-learn
       new  f9247e9   Update git information.
       new  82bb324   ENH : adding doc to clustering module API : adding trailing underscores to estimates in clustering classes
       new  35146dc   ENH: adding test for RFE and reaching 100% coverage
       new  d1acbe1   ENH : adding doc for grid_search module
       new  5dc8d7b   MISC : cosmit nfeatures -> n_features, nsamples -> n_samples, nclasses -> n_classes
       new  0b70487   Do not compare LinearSVC and SVC for exactly equal classification.
       new  b166d11   FIX : adding missing doc file
       new  1760798   Update git information in README.rst.
       new  818ed82   Update sphinxext/docscrape from numpy's trunk.
       new  b6e2e13   Refactoring in the svm module.
       new  85c6ead   Re-enable probability estimates in logisitic regression.
       new  570787d   tracking changes from master and restore broken SparseHashingVectorizer
       new  918992a   add ability to compute token ngrams too
       new  e194de4   BUG: fixed failing test in GMM
       new  d7c61d9   remove GMM.lpdf method
       new  0f3081a   merge
       new  a3961e9   fix broken doctests for SVC / NuSVC
       new  88795a5   Merge branch 'master' into char-ngram-features
       new  96e8ffb   cosmit
       new  753a6f5   update hmm module to comply with scikit-learn API
       new  10308fa   remove hmm.HMM factory to simplify hmm module's interface
       new  c716c75   merge hmm_trainers into hmm module
       new  15ab406   finish merge of hmm_trainers with hmm and remove hmm_trainers
       new  f5a6620   remove extraneous tests from test_hmm.py
       new  9332344   FIX : fix in subplot index in plot_iris.py
       new  a29f9cc   speed up hmm unit tests, add test for GaussianHMM with priors
       new  9fd9179   fix GMMHMM bugs.  speed up tests
       new  86807f3   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  80b28f2   Rename failing example in order to build the doc.
       new  a770fec   FIX: fix generating the examples with some tricky uses of pylab.
       new  3493a40   Update install information.
       new  0582253   Fallback to plain html for image rendering in index.html
       new  f7119ca   Fix bugs in dev docs.
       new  bdb4491   Updates on install doc.
       new  9b9d0aa   Update mailmap file.
       new  609e2e5   Updates on sparse.svm.SVC
       new  8b26c79   MISC: Update joblib to 0.4.4
       new  ae05aa1   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  99c03a0   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  78ec57a   Remove install_requires line.
       new  0f0a66b   Update README. Remove unused dependencies.
       new  b4dc31e   Use str for printing parameters.
       new  bc0cfb0   Update setup.py.
       new  29196ff   Change make setup to run setup.py
       new  b8b033a   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  3367693   Use repr for arrays in representation of classifiers.
       new  97bbe31   ENH : removing unused preprocessing routines
       new  35d2340   FIX : in Makefile that calls now nosetests directly
       new  20ebc14   Use nosetest as testing tool in README.
       new  f14f306   ENH: Make sure that the QDA inherits from the ClassifierMixin
       new  f64e035   Merge branch 'cross_val_gael' of github.com:GaelVaroquaux/scikit-learn into cross_val_gael
       new  5786df2   ENH: Make sure that the QDA inherits from the ClassifierMixin
       new  dcb3692   ENH: Make sure that the logistic regression does inherit from ClassifierMixin.
       new  fad03fb   ENH: Add image to graph feature-extraction helper, and some basic graph manipulation functions.
       new  1a1adee   ENH: Make sure that the logistic regression does inherit from ClassifierMixin.
       new  f953067   FIX: removing useless imports
       new  034f741   ENH: Add some code to compute a graph Laplacien on sparse and non sparse matrices.
       new  f674464   Merge branch 'master' of github.com:scikit-learn/scikit-learn into cross_val_gael
       new  f997664   BUG: Temporary fix for 'array does not own memory' in SVM
       new  005d588   ENH: First implementation of spectral clustering.
       new  0c58de2   BUG: Temporary fix for 'array does not own memory' in SVM
       new  bccdf97   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  8f008f3   ENH: Clean up the image clustering code and add an example on lena.
       new  86ee17f   DOC: Add documentation for spectral clustering.
       new  7467555   ENH: Add an estimator object for the spectral clustering.
       new  daea67c   ENH: Add k-means cluster with clever initialization. API: Break clustering in a package with sub-modules.
       new  0983539   Allow setting variable PYTHON, NOSETESTS in Makefile.
       new  c0a72ab   ENH: K-means algorithm with good initialization, more polished spectral clustering, and a synthetic example.
       new  6557000   MISC: Restore an example that is now working again.
       new  68ca67e   API: Change 'clustering' to 'cluster'
       new  5db18a5   Merge branch 'master' of git at github.com:GaelVaroquaux/scikit-learn
       new  1e8b363   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  c087f64   BUG: Add a forgotten setup.py line
       new  0f24c7a   ENH: Backport a fast graph connect component algorithm from scipy.
       new  b32a968   MISC: Cosmit based on comments from Olivier and Alex
       new  2edd6ed   Doc: correct size of intercept in svm.
       new  96eeee6   Keep shrinking and probability as booleans in SVM.
       new  2bac37e   DOC: Add some notes on complexity of clustering algorithms.
       new  03d8ec4   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  e824d9f   BUG: Adding missing setup.py file.
       new  893ba32   BUG: Remove UTF8 character checked in by mistake.
       new  ba0c787   BUG: Make graph laplacian and spectral clustering work in 64 bits.
       new  4e13e86   Refactoring: put all gmm examples in its own directory.
       new  e5cf2e4   Some love for the rst docs.
       new  26ef94e   Create new class NuSVR.
       new  c62aa88   Some patches for k_means.
       new  1d645ea   ENH: For numpy >= 1.5, use np.linalg.slogdet as a fast_logdet
       new  d1a9d8e   BUG: Fix bug with numpy >= 1.5 introduced by my previous (stupid) commit.
       new  1686dea   Backport changes in sparsetools to compile under python2.7.
       new  df8c593   Add a pure-python version of LARS and refactor structure in glm.
       new  e87eb20   add README for gmm examples.
       new  d240064   Refactoring and doc for svm module.
       new  61905f2   FIX: fixes for the lars lasso code.
       new  a31685c   ENH: Remove 'import *' in glm/__init__
       new  f2f0d88   ENH : more work on LARS (doc + examples)
       new  8b53b92   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  565852c   BUG: Fix tests broken by last commit
       new  3feef6d   ENH: Make spectral clustering tests more robust
       new  667e508   ENH : continue refactoring of GLM module (doc, moving files, config etc.)
       new  121014b   ENH : more refactoring of GLM module
       new  2ad861c   FIX : fixing __init__ files for examples
       new  664ad0a   ENH : cosmit + fix examples for doc generation
       new  7a42bf2   Cosmit (PEP 8)
       new  0b1ad2a   cosmit in examples
       new  df6ccf1   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  581549c   BUG: Fix glm/setup.py so that the glm sub package installs right.
       new  e401e05   TEST: make the test location consistent.
       new  3556457   API: Add cluster as an import of the main __init__
       new  bc5604a   Fix build system.
       new  a9bdcc4   Fixes on the Lasso-LARS algorithm.
       new  6e59a59   Add benchmarks for the LARS algorithm.
       new  3e2e464   Change score function and add docstrings.
       new  c1ebd63   Some work on the rst docs.
       new  9cef659   more doc love.
       new  29ab733   BUG: Fix warnings module not imported in coordinate_descent. Thanks to Pietro Berkes
       new  edc0774   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2d942d1   BUG : fix in Lars at the end of path + more tests (not working yet)
       new  cfc090b   ENH : using explained variance as score for regression problems
       new  b44e1c8   ENH: on the use of explained_variance in mixin regressor class
       new  6e708ab   DOC: more work on svm module.
       new  7d3a01e   Fix in LARS: specify manually number of interations for full path.
       new  780b92f   Remove "debugging" traces...
       new  394fab4   Fix doctests.
       new  a32d10c   DOC: some doc glm module.
       new  1f25d0b   ENH : New version of Bayes Ridge
       new  0d1d2e3   Merge branch 'master' of git at github.com:vmichel/scikit-learn
       new  7993c24   FIX : fix in handling of intercept in glm base and ridge
       new  71fe575   Convert to ndarray in Ridge
       new  34bdfc3   DOC: glm module.
       new  abf805a   FIX: fix intercept in LinearRegression.
       new  fff3465   Doc: Add stub file.
       new  05765bb   TEST : adding test to ridge with no intercept
       new  b2d0a77   Newer (and faster !) version of Bayesian regression.
       new  d81b619   Merge branch 'master' of https://vmichel@github.com/scikit-learn/scikit-learn
       new  cc4e824   ENH : New version of Bayes Ridge
       new  ebf568f   Newer (and faster !) version of Bayesian regression.
       new  7ce1c60   Update tests for bayes
       new  f3d743e   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  be92b0f   ENH : draft of what could be a preprocessing routine (done by hand for now)
       new  b6c7ad4   MISC: Move AUTHORS to AUTHORS.rst so that it displays better on github
       new  9cb4cab   BUG: Make sure computations do not get executed at import time, so that the tests still run.
       new  0794c1a   Cosmit
       new  3a4ae80   BUG: Remove failing doctest.
       new  150e615   MISC: More tests and more docs for preprocessing.
       new  56306dc   FIX : prevent pipeline.score to do a fit which was wrong
       new  01856ce   FIX : it may happen that pipeline.estimator do not implement predict
       new  36a70f7   BUG: Cater for NaNs in SelectPercentile.
       new  ee4b532   Cosmit: 2 lines between function definitions
       new  a19a432   ENH: Make sure that the cross_val_score uses StratifiedKFold for pipelined classifiers.
       new  a9e73ad   ENH: GridSearchCV: add an 'iid=True' and open the option to optimize average score accross fold.
       new  3c996a9   MISC: 3-Fold cross-val by default
       new  99e3fb0   ENH: Make sure that grid_search uses a StratifiedKFold by default on classifiers.
       new  15c2a35   BUG: Fix doctest
       new  4b56aef   DOC: better example for SVM-Anova
       new  68b73d3   moving ridge out of bayes.py
       new  9213838   Benchmarks for some Glm classifiers.
       new  9db0e12   DOC: more on glm module.
       new  41165ab   Fix typo
       new  62ac028   ENH: Make sure docs build on older versions of sphinx
       new  9aa3e84   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  8d88c28   DOC: Prettify
       new  c02afbe   DOC: Make first page more compact.
       new  a21f0e8   DOC: Update the developer guidelines.
       new  12e8e47   Backport total_seconds from python2.7 to use in benchmarks
       new  bb34e5f   Refactoring in glm.benchmarks.
       new  5691026   Rename nSV_ --> n_support_ in svm module.
       new  e545c6a   Some more doc for the glm module.
       new  6782310   Update doc svm.NuSVC
       new  29b1720   MISC: Tweak front page
       new  21d5516   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  dd4872b   ENH : adding PCA filter
       new  4f12dec   ENH : adding computation of percentage of variance explained by each component
       new  059fb22   Make BSD find happy.
       new  f9815ce   Compat: Add function copysign in the case of numpy < 1.4
       new  d04b33d   MISC: Cosmit on PCA tests to get understandable errors from the buildbot.
       new  47a1ca2   MISC: Delayed import of pylab, to work on the buildbot
       new  d59506f   MISC: Relative imports
       new  a1b47d4   BUG: Fix tests to be moroe robust
       new  bc01343   Do not import pylab globally in benchmarks
       new  530bef9   Test for function utils.fixes._copysign.
       new  6ba3b80   DOC: Fix typos in svm module documentation.
       new  941ae5a   FIX: fix previous (stupid) commit.
       new  4b1a1e4   FIX : for doctest in PCA
       new  971b5b6   Remove Y from fit in OneClassSVM.
       new  8d16947   Welcome Virgile Fritsch
       new  d0792b9   cosmit + trailing spaces + improved some comments
       new  31e170f   Add first draft of variational bayes
       new  a54a873   Add variational inference
       new  dbb3aa1   pep8 spacing
       new  cbc5f44   more cosmit
       new  6e5e140   Cosmit
       new  e7ae40f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e9d356d   MISC: Add a warning in the spectral clustering if pymag is not present
       new  c21d0a6   introduced gael's implementation of fast_ica and debugged GS orthogonalization
       new  e259a80   Update docstring for BayesianRidge.
       new  4cd777c   ENH : adding ledoit-wolf for robust covariance estimation
       new  d72c27d   cosmit in fastica, that created a bug -- to be fixed
       new  1444269   updates in fastica and more tests
       new  98ec3b9   DOC: Update doc for bayesian regression + examples
       new  eeb9dbd   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  27477f8   DOC: update docstrings in glm.bayes.
       new  b92ff2f   Revert "MISC: Add a warning in the spectral clustering if pymag is not present"
       new  edf9429   Revert "Revert "MISC: Add a warning in the spectral clustering if pymag is not present""
       new  a04a141   cosmit
       new  d8caf8a   FIX : Remove reference to Variational Bayes
       new  060bea1   Merge branch 'master' of github.com:scikit-learn/scikit-learn into issue-77-sparse-cd
       new  b21fbbb   starting boilerplate for sparse coordinate descent
       new  4315aee   completed and cleaned the tests
       new  f744094   improved the tests
       new  cade67f   Update docstrings in svm and logistic.
       new  0d32d38   ENH : adding FastICA class + example
       new  5a7a47a   Merge branch 'add_ica' of http://github.com/bthirion/scikit-learn into ica
       new  c75191f   Merge branch 'master' of github.com:scikit-learn/scikit-learn into issue-77-sparse-cd
       new  72f7cb9   fix broken test
       new  89f605d   FIX: some fixes for bayesian doc in glm.rst
       new  4bb6209   BUG: Import stats explicitely to work with scipy > 0.7
       new  8e70946   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6490014   Fix typo on David's name
       new  6b14f85   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4d764fd   DOC: some more fixes on BayesianRegression doc.
       new  6a6bbf0   Better error message in fit svm.
       new  4ec011b   Fix failing tests (sparse svm).
       new  d291408   ENH : more on ICA (examples + doc)
       new  4ff781f   Merge branch 'ica'
       new  375be9d   Add a reinitialization function for estimators + write test for GridSearch.
       new  3bbc5fe   More doc in bayes.py, fix bug in high dimension, add score
       new  83df999   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  5bc337e   FIX : fix ica vs pca example
       new  8c9f1c7   Change test name for the _reinit() method.
       new  10a86a1   checkpointing work in progress
       new  2d7aaaf   avoid confusing cython extension names
       new  20bb335   solved conflict in test_fastica
       new  9e7d8e9   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b20abdd   ENH: Add _set_params/_get_params in Pipeline
       new  d74a093   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  300d473   Merge branch 'master' of github.com:GaelVaroquaux/scikit-learn
       new  c4c8ffd   ENH: Implement _reinit on Pipelines.
       new  ea66aab   ENH: Use new pipeline framework in SVN-ANOVA
       new  ba0e2af   added probabilistic PCA and associated tests; works reasonably well
       new  48b0a00   BUG: Import stats explicitely to work with scipy > 0.7
       new  1c59054   ENH: Finish the __repr__ for the pipeline
       new  b2423f9   ENH : adding example + refactor in covariance module
       new  bcdb2a1   splitting ledoit_wolf.py in two files
       new  79f8ddb   ENH: Add error management to the KFolds
       new  120fc3a   oups missing example file
       new  8be939f   FIX: missing covariance.py
       new  ab622c2   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4a67caa   fixed various issues with sparse datatype handling in previous checkpoint
       new  da285ea   ENH : change the convergence trigger
       new  24e9b6a   More coverage for bayes
       new  743159d   Merge http://github.com/GaelVaroquaux/scikit-learn into gael
       new  06cbe38   cleaning the handling of the intercept in GLM linear models
       new  edaa6a7   better note
       new  4bd590d   first stab at the sparse CD
       new  fab8518   leave sparse evaluation of the dual gap for later
       new  b06f78d   forgot files from previous checkin
       new  e4bf8d4   Merge branch 'master' of github.com:scikit-learn/scikit-learn into issue-77-sparse-cd
       new  e7c8ba6   ENH: Port the nipy k-means with some cleanups and enhancements.
       new  a5b75b3   ENH: Change the initialisation heuristic for k-means: in general random is faster.
       new  3421872   check that sparse API for coordinate descent also work with dense list-based input
       new  60d326d   BUG: Adapt spectral to new k_means API
       new  4ae3119   Merge branch 'master' of github.com:scikit-learn/scikit-learn into kmeans
       new  a0a4375   ENH: Add error management to the KFolds
       new  7a78d61   one more test for sparse CD
       new  c7692d8   Be able to do _get_params and _set_params in a recursive way.
       new  7574a48   Cosmit
       new  3a869a2   Cosmit
       new  a5b807e   sparse dual gap too!
       new  aa9297c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into issue-77-sparse-cd
       new  26643c5   DOC : create and start doc for cross-validation. Still lack of examples...
       new  1ee1e29   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  122d338   cosmit
       new  eb18b3d   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bff2c6e   fix broken doctest in cross_val
       new  e65c4dc   DOC : add changes in classes.rst
       new  1437a40   Merge branch 'master' into issue-77-sparse-cd
       new  f565966   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e7f43ba   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b568f5d   Fix imports.
       new  e900a13   more robust tests
       new  5e98aa8   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e85c625   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  1ef468c   Merge branch 'master' into issue-77-sparse-cd
       new  7b5fb57   fix in ppca
       new  e82e9e0   cosmit in pca and test_pca
       new  20fe292   ENH : avoiding computing a pinv at each iteration in BayesianRidge
       new  4adfad6   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  ebc414e   missing import
       new  8020a93   TEST : bayes
       new  b1da4fe   FIX : imports in __init__ of glm
       new  d1c196d   cosmit in ARDRegression
       new  f23267f   Merge branch 'master' of github.com:scikit-learn/scikit-learn into pipeline
       new  d52c5c8   ENH: Change pipelines so that they are simpler and address subobjects using the routing logic of the BaseEstimator
       new  a81ba85   BUG: Fix bug introduced by previous commit
       new  9ed4b89   BUG: Fix doctests.
       new  fc6e9ac   TEST : removing lda.py
       new  e3e458b   COSMIT : PEP 8 in PCA
       new  1c95067   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0c6b21b   DOC: Better docstring.
       new  cb35cc8   merged origin and fixed a conflict
       new  f2d7455   ENH: Make setting nested parameters on Pipeline really work.
       new  dea6e40   Cosmit
       new  30ffff1   TEST: Add a smoke test for cross_val_score
       new  92dddfa   added sparse Lasso utility class
       new  d9d3735   OPTIM: fix a typo and some suboptimal cython constructs in dense coordinate descent
       new  d5962d2   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  1e5074d   DOC: Explain the _set_params method in BaseEstimator class.
       new  b17cf79   FIX : RegressinMix score had flipped y_true and y_pred
       new  6761568   EXAMPLE : adding model selection example with train/test error graphical illustration
       new  13ebec2   EXAMPLE : making only one figure in model selection example
       new  f201bce   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4ae80d4   Temporary fix for Table of Contents not showing.
       new  6d2e473   MISC: Fix spelling.
       new  6dda5f5   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0dcd737   DOC: Add more links to the index
       new  c5ffd7e   DOC: Add forgotten link targets.
       new  9fa6521   DOC: prettify the nearest neighbors docs
       new  0625780   DOC: Prettify the SVM docs.
       new  e7672b9   MISC: Add 'Python' in the examples page
       new  fc7c55b   DOC: fix the PCA iris example.
       new  4bec379   MISC: Remove non-necessary lines from PCA example
       new  e2957a5   DOC: Prettify the clustering documentation.
       new  6b282e0   DOC: Fix the reference classes documentation
       new  fd315bc   DOC: Prettify the GLM docs
       new  4ee8b88   share the same cython impl for both lasso and elastic net CD
       new  a8e72b1   make d_w_max early stopping criterion scale invariant
       new  4e9810a   cosmit: s/nsamples/n_samples/g and s/nfeatures/n_features/g
       new  aa1b766   group stopping criterion related boilerplate in the same place for readability
       new  a4748e1   MISC: Quiet down the tests.
       new  a9f8b48   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2ab74cd   ENH: Add a tester to the scikit.
       new  54e8b77   ENH: Make doctests pass with numpy tester.
       new  a4f50ea   MISC: Make sure tests always run.
       new  8ab4f91   FIX: make CD lasso robust to zero valued columns (useless features)
       new  0de7e5d   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  332f63e   MISC/DOC: fix reference
       new  e3dc8d9   Do not import pyalb in the benchmarks.
       new  f6c6724   add duration to glmnet benchmark output
       new  9176cab   removed useless includes
       new  8a3f06a   make d_w_max threshold independant of the squared norm of y to make it useful in practice
       new  a555f22   Merge branch 'master' into issue-77-sparse-cd
       new  86d1048   port latest bugfix and optims from dense CD to sparse CD
       new  4287862   Rename PyBallTree* --> BallTree* in BallTree.cpp.
       new  e882140   TEST: Fix test_pca sign error.
       new  e535011   Added filters to WordNGramAnalyzer.
       new  34c2758   fix NORMALIZE_WHITESPACE issues in doctests
       new  95546e7   more robust and understable CD elastic net test using explained variance score instead of RMSE
       new  2e1fc63   forgot to setup the good value of rho in last checkin
       new  bd72bdb   Merge branch 'master' into issue-77-sparse-cd
       new  ba2356b   FIX __init__.py of glm.sparse
       new  bb2e066   EXAMPLE: add example of dense vs sparse Lasso on dense and sparse data
       new  7a636fc   TEST: Fix whitespace in doctests.
       new  545f47c   FIX :  example of dense vs sparse Lasso on dense and sparse data
       new  39e4208   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  e9b5733   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  fe93d47   Added a non-hashing dense vectorizer object.
       new  6053027   LARS refactoring speedup Work In Progress!!!
       new  8144df3   more on lars optimization
       new  b591602   more on lars speedup WIP
       new  f5bb8ab   LARS with precomputed kernel working.
       new  c463cea   more work in lars optimization.
       new  a79553e   More on LARS performance: triangular solving and cholesky deletes.
       new  7c24dd7   A more challenging example.
       new  3c884fa   cleanup and fix some tests.
       new  bf789be   ENH: GridsearchCV, Pipelines and cross validation
       new  780c166   More on LARS.
       new  aa5732d   more optimizations
       new  601293f   cleanup
       new  a9b03b9   Added transform() method to Pipeline object.
       new  0f79a86   Updated dense vectorizer to follow transformer API.
       new  d1a9f55   passing Gram in LARS and LassoLARS
       new  ca9c569   ENH : more doc in lars.py, handling of intercept
       new  6f8039c   Merge branch 'fabian/python_lars_fast_2'
       new  9e3b36c   fix doctests in lars.py
       new  805d184   more on lasso benchmark
       new  b5f7356   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  ac02b33   merged origin
       new  452f723   ENH: Make sure that Pipeline and GridSearch objects are indeed recognized as classifiers, if they contain a classifier.
       new  a4ed78e   ENH: Make sure clone works on pipelines
       new  0b8aa6c   Merge branch 'textextract' of git://github.com/mblondel/scikit-learn
       new  3fb62b9   Support fit_transform() in pipeline.
       new  20bd2b9   Support lists for training data in grid_search.
       new  3dbe9f3   Use fit_transform and use iterables for documents.
       new  0675770   Remove uncessary code.
       new  88a52a1   Save memory when the matrix is built.
       new  5c7f3fb   Added fit_transform() to pipeline.
       new  e790866   ENH: Implement a score for the GridSearch.
       new  91f0814   Vectorizer should implement fit_transform.
       new  fd60658   ENH: Make sure that a GridSearchCV has a score
       new  5135fca   Update CBLAS files: add rotg, rot, trsv, remove tpsv.
       new  5c17058   Good bye minilearn.
       new  8be8a8a   Fixes for ref atlas.
       new  9719271   Some fixes for the atlas we ship.
       new  e872975   FIX : preprocessing : scaler should not be allowed with axis=1 (opt removed)
       new  cbdbda6   Add missing cblas_dcopy files.
       new  b44a386   skipping BayesianRidge failing test
       new  7b12364   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2138c51   Merge branch 'cross_val' of github.com:GaelVaroquaux/scikit-learn
       new  4b71675   Merge branch 'master' of http://github.com/fabianp/scikit-learn
       new  e8e2b1a   FIX: failing gmm tests
       new  20937eb   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  192d243   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9590adb   better docstring + cosmit for the RFE module
       new  d5a71d2   pep8
       new  78fbd20   move the RFE module to the feature selection package
       new  5aa6098   merge textextract branch from mblondel
       new  a9f98dc   Fix broken merge by ogrisel :-P
       new  dce7bf7   PEP8 + missing fit methods
       new  9209225   make analyzers inherit from BaseEstimators to get better repr and parameters management
       new  69ca71e   using diabetes in lasso/lars examples
       new  e8cc69e   removing assert for debug
       new  f1530fb   ENH : speeding up the LARS
       new  0e193e8   BUG: bug fix in LARS Lasso mode + speed improvement (we can still do better)
       new  e174b32   adding LARS with Gram to benchmark
       new  87644db   ENH : speed in LARS by forcing X to be fortran ordered + cosmit (unactive -> inactive)
       new  c9aaaaa   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6e982be   pretifying the LAR / LARS examples to match with results on wikipedia page
       new  41d3f72   work in progress: refactoring the document classification dataset API to remove the feature extraction step
       new  82ba7a1   Merge branch 'master' into textextract
       new  231cb36   cosmit in docs of glm module
       new  42fca0f   New theme for the web page.
       new  daf8ed9   Add what's new page and a nicer sidebar for index page.
       new  043cd96   Move glm related benchmarks to a common location.
       new  4c0d270   Performance improvements for LARS precomputed Gram matrix.
       new  df54c39   Remove weight_label keyword from SVR.
       new  a22d25e   Remove Minilearn C sources.
       new  4884978   Add some developer information for the CBLAS we ship.
       new  a5703a1   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7417610   Cosmetic changes to the cblas README.
       new  1080466   More love for the new web page theme.
       new  0f37104   Refactoring in the svm module.
       new  d1d2cef   Also remove Windows python extension (.pyd) in make clean.
       new  ef26ea4   move sparse.svm into the svm module to match glm.sparse.
       new  4089a47   Make a reference page in the docs.
       new  1d22339   Update class reference.
       new  08d1a4d   Cleanup in setup scripts.
       new  017311d   Bugfixing in setup.py
       new  972a272   cosmit.
       new  1be5131   Cosmetic changes to svm tests.
       new  aa2a523   Correct default value of gamma in the svm docstrings.
       new  2461026   Refactoring in sparse SVM and bug solving (default value of gamma).
       new  06c74d2   Refactor svm tests.
       new  35ce1eb   Fix doctest failing by last bug fixing.
       new  558d878   Add target test-doc to Makefile to test the RST docs.
       new  dbbe2ef   Remove obsolete debugging code from grid_search.
       new  41f8c0b   Remove obsolete comments from doc.
       new  5bbd6d8   Cosmetic changes to grid_search_digits.
       new  a07e818   ENH : improvements in bayes
       new  366de15   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  44597a9   fix doc generation on plot_lasso_coordinate_descent_path.py example (pb on my box)
       new  cc1b30c   SparseCountVectorizer, SparseTfidfTransformer and Sparse Vectorizer
       new  edf4fd4   Reduce import time.
       new  21debd0   Polynomial kernel also uses keyword gamma.
       new  be36a15   Fix typo in svm docs.
       new  7fa2603   Fix wrong link in doc.
       new  612e9e4   DOC: update doc about LARS.
       new  2a70aec   Add LARS, LassoLARS to class reference.
       new  7f760b7   Update funding info.
       new  d6f75e5   Update API changes in feature_selection doc.
       new  5f8978d   Remove the ann module.
       new  f72be0d   Remove obsolte css code from the docs.
       new  6695c5b   Update docs on sparse svm.
       new  02a9dfb   Correct spelling errors in svm documentation.
       new  ac62bf0   Fix spelling errors in glm.rst
       new  1459c70   Remove non ASCII characters from the docs (problem in latex output)
       new  5725d3c   Fix non-html (latex) generation of the docs.
       new  2ccfa9f   Fix RST (numpydoc) markup.
       new  68b078d   Update organization in index.rst
       new  f9aa89d   Update doc of neighbors module.
       new  0182bce   Update links in svm doc.
       new  67f3468   Update theme in web page (sidebar color)
       new  4b98c35   Fix malformed RST in BallTree.cpp
       new  8d45f3b   Change doctests that are machine dependent.
       new  5cb8e6c   Update joblib to 0.4.5
       new  65569b7   Comment out fragile joblib tests.
       new  61c81a2   Add test.py script that runs nose.
       new  cce8adc   Remvoe printing statements from tests.
       new  652550a   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  3d29d64   Adopt numpy naming scheme for __version__ attribute.
       new  4b025a0   Compatibility fixes for utils.graph.
       new  4def939   Use by default np.unique.
       new  3baf985   Compatibility fixes for scipy <= 0.7 and numpy <= 1.4
       new  ed1dd5c   Compatibility fixes for old scipy.sparse.
       new  70ac94a   Do not include Makefile in final release.
       new  a64f488   Add missing files to setup.py
       new  cd86703   Add missing images
       new  26bdc72   Rename features --> feature_extraction to match module feature_selection.
       new  e7d4035   Update information on testing.
       new  874e228   FIX: bug in setup.py file from glm/sparse/
       new  615de15   Update web page.
       new  48af367   FIX: fix imports in example for renamed modules.
       new  8df4238   Add function template for doc.
       new  c0812b6   Update web page theme.
       new  739bda7   Update mailmap file.
       new  138709e   Add Feature Selection classes to the reference docs.
       new  83057b7   0.5 changelog (Work in progres).
       new  337c87f   DOC: updating doc on Univariate feature selection
       new  27de186   DOC: better link that literalinclude.
       new  2df892f   Combine user guide into a single file.
       new  04b05b7   Update changelog.
       new  dd34f5d   Add png logo.
       new  5426dd3   Update MANIFES.in file.
       new  a658f83   Update test.py and README.
       new  79793fc   normalize option for TfidfTransformer
       new  d8b1db1   fix garbage
       new  cca1515   Simplify test machinery.
       new  c3b7b35   Use ELLIPSIS for machine-dependent results in joblib.
       new  1bccc55   Comment out machine-dependent tests from joblib.
       new  4f93e16   Update Makefile
       new  3b7d38b   Welcome Mathieu Blondel.
       new  399a371   Fix doctests from the tutorial.
       new  d6b4444   0.5 release candidate.
       new  1ef7288   FIX: some setuptools oddities.
       new  6058971   0.5.rc2 release.
       new  8a900a9   Still fixing distutils oddities ...
       new  eba9984   0.5.rc3 release.
       new  6ea8b88   pruning joblib external
       new  9322dc6   Merge commit '8b26c79e7ba67dd614e84c3333ce845ee2edb049' into dfsg
       new  726aa6d   Web page update.
       new  92258ec   Add sparse ti glm/__init__
       new  7206e80   Fix typo in docstring.
       new  94ba2be   Merge 0.5.rc into DFSG (pruning externals: joblib)
       new  4a9c3b0   Merge 0.5.rc3 into DFSG (pruning externals: joblib)
       new  c2c3e4a   changelog for 0.5~rc3-1
       new  4edd2b5   adjusted gbp.conf since we now build off debian
       new  1b1906a   * Deprecated patches: - up_workaround_numpy_cython_issue589652 - deb_cython_0.12.1
       new  6a15013   force distutils build system + to run tests against "installed" version + run doctests
       new  a31c5b5   boost build-dep on sphinx >= 1.0~ and cython >= 0.12~
       new  9d02e90   doc extensions fix + tentative dfsg rule
       new  c29cd70   Merge branch 'dfsg' into debian
       new  7b78b54   pruning joblib in favor of external package
       new  50902c4   tracking changes from master
       new  f022ba6   adjusting gbp.conf
       new  9353116   added patch deb_use_system_joblib
       new  ba758cb   tentative patch for runtime verbosity control of libsvm
       new  5ae7c5f   removing carried version of libsvm in favor to use systemwide (otherwise svm.h inclusion conflict)
       new  64bfb19   prune copy of libsvm sources from dfsg
       new  14c5dc1   Merge branch 'dfsg' into debian
       new  7d0fc75   finalizing libsvm verbosity control patch
       new  9fc59ef   pruning stale patch deb_cython_0.12.1
       new  22dc5e3   adding in TODO to take care about external liblinear
       new  b64a55d   trying to modularize rules a bit, move of .so away happens only at dh_pysupport time now
       new  da60c0c   splitting svm control into 2 patches (cythoned changes separate)
       new  5c4b52a   renaming libsvm verbosity patch -- will forward upstream
       new  6aeb742    up_system_libsvm_no_svmhdepends -- do not use shipped svm.h
       new  2de4cd8   use installed version (not build/) + python-test% separate rule
       new  bfd7010   still need to install dummy joblib subpackage
       new  5eee643   tune up testing against installed version
       new  8a08afa   Adding author and description for the patches
       new  8e0f543   filled up patches headers according to DEP3
       new  f6eff0f   debian/rules -- tests -- run in the same environment to preserve set PYTHONPATH
       new  bc7cae5   move doc building into installdocs, order rules in the order of their invocation, install changelogs
       new  8852e3a   DOC: removing a stale request for subversion write permissions
       new  415c5e9   Allow to build _libsvm.so against system-wide LIBSVM's svm.h
       new  6aa2e61   API to control LIBSVM verbosity without patching
       new  ce31b5b   recythoning _libsvm.pyx for previous commit
       new  1ec827a   revert change to libsvm -- now verbosity is controlled via API
       new  f4dd0a6   enable more doc testing for test-doc Makefile rule
       new  f1f4bd3   adding acknowledgement to Dr.Haxby for my support ;-)
       new  d2ce657   Fix typo
       new  1553baf   Fix doctests from RST docs.
       new  2bd2a68   Fix links in about page.
       new  66e5b03   Cosmetic changes in install.rst
       new  9b1dabf   You want the truth well here it is.
       new  7e69801   Add a link to the PDF version of the docs.
       new  0f148e0   0.5 final release.
       new  e9d2961   Start of 0.6 development cycle
       new  990bdbf   Add note on executing the test suite.
       new  fecfae5   Update web page.
       new  a3476be   Merge 0.5 into DFSG (pruning externals: joblib)
       new  3df1a2d   Merge branch 'dfsg' into debian
       new  592ed41   removing patches aborbed upstream
       new  76bb672   changelog for 0.5-1
       new  dc72677   Boost policy compliance to 3.9.1 -- no changes
       new  4daa85c   Add a note on complexity for SVMs.
       new  21cd1b8   Fix indentation.
       new  3255d16   Fix cross_val when y is a 2d-array.
       new  06e39c9   Add refit option to GridSearchCV.
       new  2c26ebd   Merge branch 'master' into textextract
       new  1cd2ab3   cosmit
       new  d6dbc80   more precise doc on SVM complexity
       new  bf85063   API changes to precision_recall
       new  b610b3e   Fix consistency problem in the order of arguments for loss functions.
       new  42564aa   Add fbeta_score and f1_score metrics.
       new  601ffe9   Rename roc to roc_curve.
       new  77ef3cd   Merge branch 'master' into textextract
       new  2825308   Add datasets to __init__ file.
       new  ba8e530   Correct typo in docstring.
       new  8bb9849   Allow access to multi-class SVM in liblinear.
       new  1881139   Do not execute test coverage by default.
       new  0a91492   Fix doctest in grid_search.
       new  8e20af9   Merge branch 'master' into textextract
       new  65eeaef   lighten GMM tests.
       new  5c99998   remove n_dim property (use plain field).
       new  bdf4d12   Fix and enable _covar_mstep_full in gmm.py
       new  f60a956   Cosmetic changes.
       new  3cb198c   FIX: explained variance is not symmetric: ground truth comes first
       new  ab87a5e   Add tests for predict_proba in LogisticRegression.
       new  89729bf   FIX: ticket 147 on pb with 2d y in f_regression
       new  4a36865   FIX: ticket 147 on pb with 2d y in f_classif
       new  6d30958   ENH: Small optimization to BaseEstimator
       new  58a432c   initial checkin of sgd package.
       new  b7a1afb   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  65df6c3   adding 'iid' option in cross_val_score
       new  2956f65   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  1146598   Bindings for libsvm-dense
       new  0bd1890   Update svm benchmark with latests libsvm.
       new  3e00d98   Some fixes for libsvm-dense
       new  920b7df   set rho on 1 or 0 if L2 or L1 penalty.
       new  21a99d1   l1 penalty implemented. different loss functions supported.
       new  dfae495   Use filter object.
       new  5f9c342   More accurate info in examples.
       new  8606d6f   Update svm examples affected by latest API changes.
       new  8e6adc9   DOC: Some docstring for libsvm low level API.
       new  354a31f   Add dtype parameter to CountVectorizer and SparseCountVectorizer.
       new  ca5c6ac   Revert "DOC: Some docstring for libsvm low level API."
       new  1a5f412   Revert "Update svm examples affected by latest API changes."
       new  e88a6a6   Revert "More accurate info in examples."
       new  728f3f9   Revert "Some fixes for libsvm-dense"
       new  d1d3a83   Revert "Update svm benchmark with latests libsvm."
       new  fccc2ec   Revert "Bindings for libsvm-dense"
       new  4d2b376   pretifying plot_weighted_classes.py
       new  88ab266   ENH: enhacements in the gmm module.
       new  e8d32c6   Make previous commit work also with old versions of scipy.
       new  1ca639e   No specific need that matrix is upper-triangular in gmm.
       new  d40c1ac   Fix doctests in gmm (skip random ones).
       new  52505e9   Revert "Fix doctests in gmm (skip random ones)."
       new  c7248c8   Revert "No specific need that matrix is upper-triangular in gmm."
       new  3fab72a   Revert "Make previous commit work also with old versions of scipy."
       new  0ac6935   Revert "ENH: enhacements in the gmm module."
       new  629527f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  d6fd98c   A few optimizations.
       new  d5c62cf   Move sparse code to sparse module.
       new  7525e14   Remove Sparse prefix from class names.
       new  a590150   Move preprocessing to its own module.
       new  86fe26c   Add Normalizer, LengthNormalizer and Binarizer.
       new  85a1c4a   Remove normalize option from TfidfTransformer.
       new  626c7e8   Sparse equivalents of Normalizer, LengthNormalizer and Binarizer.
       new  192e755   Bindings for libsvm-dense
       new  58e30ff   Update svm benchmark with latests libsvm.
       new  5dfdd56   Some fixes for libsvm-dense
       new  f4e1f14   More accurate info in examples.
       new  0ddfebe   Update svm examples affected by latest API changes.
       new  565f5eb   DOC: Some docstring for libsvm low level API.
       new  1bd45d8   Compile _libsvm_sparse in the sparse module.
       new  774107a   Add setup.py to svm.sparse
       new  29005e4   Preliminary fix for naming issue in OSX with libsvm.
       new  42c03c0   Add a namespace to svm methods to avoid same name mangling.
       new  5e3fd29   Merge branch 'textextract' of git://github.com/mblondel/scikit-learn into textextract
       new  5a5d9d5   Fix for building libsvm in a portable way.
       new  75ac75e   ENH: s/filter/preprocessor/ + docstring cosmit
       new  1319279   ENH: more docstring love
       new  7a344acf  ENH: implement __repr__ for DefaultPreprocessor so that estimators __repr__ looks prettier
       new  a30698c   cosmit
       new  78e5a01   make the pipeline / grid_search object nicer to introspect in tests
       new  79a860b   FIX: make grid_search output deterministic even in case of tie on the scores
       new  03880d7   Fix hierarchy inconsistency for sparse module.
       new  3ac51e2   FIX: quick fix in predict_proba in LogisticRegression
       new  16437d4   FIX: fix doctest with recent API changes.
       new  c3fe0a4   FIX: fix fragile doctest.
       new  9fa7641   Updated liblinear to latest version 1.7.
       new  4c9170d   Make liblinear quieter.
       new  2a4adcb   Update classes to use new features from liblinear 1.7.
       new  b6f14db   Move logistic into glm and add a sparse version.
       new  1ac91af   Doc: better tests for logistic.
       new  b77189b   Fix imports in example.
       new  97d0af1   removing debug compile flags
       new  2847e3d   added class encoding. added tests added l1 final penalty.
       new  dd82f49   does not belong to the repo.
       new  8ce4c64   Merge branch 'master' of git at github.com:pprett/scikit-learn
       new  0009295   Code review from Alexandre:  - PEP8  - removed cython annotation file from repo
       new  394b22e   Merge branch 'master' of github.com:pprett/scikit-learn
       new  6580023   removed unnecessary print statements. added mlcomp example atheism vs. graphics.
       new  d9ddde0   mereging pprett sgd work while tracking master
       new  b31e742   mark heisentest as skipped: it randomly passes 3 out of 5 times on my box with pyamg installed
       new  414c342   Merge branch 'master' into sgd
       new  68634d2   kill evil tabs
       new  86ef9e0   cosmit
       new  e4ac71e   cosmit on example
       new  4af396d   cosmit: PEP8 + some missing docstrings
       new  d6517b9   more cosmit
       new  94b6d6e   more cosmit
       new  280eaec   sgd module code review
       new  ec168a3   merging with alexandre's fixes
       new  ca77c79   100% code coverage.
       new  0a6f6d1   added doctests to SGD and sgd.LinearModel
       new  a8dbb64   change GMM initialization to use cluster.KMeans
       new  cd1e6c9   change GaussianHMM initialization to use cluster.KMeans
       new  644ef2d   fix broken doctests: they are space sensitive unfortunately
       new  314c687   merge
       new  1d7be88   Merge branch 'master' into textextract
       new  21e5c2b   better way to load folder / files dataset
       new  458527f   porting the sparse document classification example to the new API
       new  a81288d   fix bug in hmm.GaussianHMM mstep update for 'full' covariance
       new  b4ab535   Reapply "ENH: enhacements in the gmm module."
       new  aec6fd3   fix gmm examples
       new  6ae9bcd   initial checkin of sgd package.
       new  0b22231   set rho on 1 or 0 if L2 or L1 penalty.
       new  1e61592   l1 penalty implemented. different loss functions supported.
       new  843fe8b   added class encoding. added tests added l1 final penalty.
       new  02a0711   Code review from Alexandre:  - PEP8  - removed cython annotation file from repo
       new  b99f7f4   cosmit: PEP8 + some missing docstrings
       new  e84b4ea   sgd module code review
       new  29e3afb   100% code coverage.
       new  e0665ed   added doctests to SGD and sgd.LinearModel
       new  6845dad   Fix doctests in sgd module.
       new  a55454b   Welcome Peter.
       new  a95f39e   Avoid iterating over features in gmm.
       new  2ae5c65   merge
       new  03aa2ba   Merge commit 'origin/master'
       new  a4f9428   initial *draft* of the sgd module documentation added.
       new  295cb3c   added Readme so that sphinx stops complaining.
       new  561afe3   additional documentation for sgd (plot of various convex loss functions). TODOs in sgd_fast_sparse.pyx done.
       new  75fa2b1   Move common sparse code to SparseBaseLibLinear.
       new  78febeb   Fix Sparse Logistic Regression.
       new  1de02f8   Import LogisticRegression in sparse/__init__.py.
       new  b0eb68c   Merge branch 'master' into textextract
       new  e4cb713   Activate class_weight option in fit() for liblinear-based classes.
       new  89462bb   Merge branch 'master' into textextract
       new  6634501   adding path example on logistic on IRIS dataset
       new  bd4c595   Merge branch 'master' into textextract
       new  5f40147   math formulation cont'
       new  db9b839   penalty contour plot added.
       new  3fcb7cf   PEP8 + better docstrings
       new  8131033   FIX: add missing README.txt file for the sgd examples
       new  af56466   more SGD documentation added: example, math formulation , implementation details. adopted the seperating HP example from the svm package for sgd.
       new  7685c9e   ENH: more cosmit, docstring, test cleanup for the metrics module
       new  f407247   EfficientBackprop reference added.
       new  c34ca71   cosmit
       new  02b56e7   Documentation for sgd polished. SGD.predict_proba for log loss implemented.
       new  6e02554   Merge branch 'sgd'
       new  694de78   fixed doctests after SGD class index refactoring.
       new  323a5d3   BUG: Make sure that grid_search works with sparse data.
       new  b2842d3   Removed tabs.
       new  9840e55   implemented OVA for multi-class SGD.
       new  49d89c5   implemented OVA for multi-class SGD.
       new  053e88e   register the SGD chapter in the user guide TOC
       new  8698230   PEP 8 in metrics module
       new  6b52139   ignore generated doc elements
       new  297ccc4   some more cosmits / PEP8
       new  decbabe   more PEP8
       new  22a8ecb   helpers to use tags with vim / emacs
       new  9d563a6   Make it possible to pass explicit labels to confusion matrix
       new  766c244   make binary classification recall explicit
       new  9995c11   factorizing code to make it easier to do the multiclass refactoring in one place
       new  c3d7e15   fix bug in GMM._get_covars dimensions
       new  eef4224   make HMM interface consistent with GMM
       new  ddbc043   refactored test_metrics to handle the binary case explictly and make room for the multiclass case
       new  a11b38c   test precision recall for binary classification
       new  002d02a   clean up interfaces in hmm and gmm
       new  dbdb8a6   remove n_symbols argument from MultinomialHMM.__init__
       new  d7b270f   more code factorization: fscore joins the party
       new  9a1cc99   and you thought you could escape the PEP8 screening
       new  9f2057d   missing test for f1 output
       new  748fa6e   cosmit
       new  1689be0   Hello list,
       new  a4c34b8   extract label extraction logics
       new  1a167c1   ENH: make precision, recall and f1_score handle multi-class
       new  e116317   removing test for multi label perf evaluation
       new  48e818e   FIX: area under curve: recall is x and precision is y
       new  9372c35   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  678ae5e   add GMM classification example
       new  08c3a68   Correction of a bug with the management of the dimension of the autocorrelation parameters.
       new  1b98de4   Forgot to retire pdb.
       new  9698787   Merge branch 'master' into issue-155-multiclass-precision-recall
       new  c47cec2   MISC: Cosmit in new GMM classifier example
       new  aee1cae   Merge branch 'master' into issue-155-multiclass-precision-recall
       new  0f2eed4   DOC: Make the plot_ica_vs_pca example richer.
       new  03e75d1   MISC: Some tweeks to the layout so that the docs display better on a narrow window.
       new  18e2a6d   DOC: Fix title level in install
       new  9c08a28   ENH: new utitity in metrics module: the classification report
       new  711fb44   DOC: make the index page content clearer
       new  08b30d3   Fix slicing issue when using sparse matrices.
       new  19f4a8c   Y -> y (capital letter is for 2d-arrays)
       new  2aa6f27   MISC: Explicit acronym
       new  d0828de   showcase the new classification report in the examples
       new  6475d14   MISC: PEP8 in docs
       new  4508428   add detailed performance report to the digits example
       new  08e8c9f   DOC: Change the titles' layout
       new  ba96f17   Merge branch 'master' into issue-155-multiclass-precision-recall
       new  c066b89   increasing precision in plot_logistic_path.py to get nicer path
       new  a52a532   Raise exception when X_train.shape[1] and X_test.shape[1] don't agree.
       new  0ab7588   DOC: Rewamp the tables of contents and corresponding layout
       new  0d90775   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  d4e2303   DOC: Make sure the docstring of pca render well
       new  ac49fff   DOC: Remove empty section
       new  2a76cbe   tracking changes occurring on master
       new  ef09cbe   DOC: Add documentation for ICA/PCA
       new  b73d09a   DOC: Remove useless tables of contents
       new  ca5010d   Add more sanity checks for svm with precomputed kernels.
       new  1b09294   DOC: work on the clustering documentation
       new  64e7324   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b18af3f   DOC: spelling
       new  6631428   ENH : pyflakes on examples to avoid useless imports + addint print __doc__
       new  2d5788f   cosmits
       new  d3bd458   ENH: handle support to do weighted averages of P/R/F scores
       new  c810c95   DOC: Tweak in the clustering docs.
       new  21475c3   scalar scoring functions for P / R / Fbeta
       new  d8d7e5d   s/explained_variance/explained_variance_score
       new  0837e3b   make the distinction between loss and score function more explicit
       new  aeaa85b   DOC: document with more details the GMM module.
       new  c8711cd   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  f5bec85   Merge branch 'master' into textextract
       new  58b7aff   spelling
       new  29bc603   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  576b38b   make the grid search able to use an arbitrary score function
       new  296ab65   clarify GMM classifier labels
       new  def4b87   Merge branch 'textextract' of git://github.com/ogrisel/scikit-learn into textextract
       new  7ad238d   DOC: make the neighbors doc sexier
       new  e003e90   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  cccbc8c   Merge branch 'master' into textextract
       new  1e3df61   removing the hashing vectorizers code that need a full rewrite
       new  ec1514a   ENH : more love in examples (adding print __doc__ + some brief descriptions in headers + fixing Anova SVC Pipeline example)
       new  7a2b803   FIX: fix docstrings in LARS (issue 8 on github)
       new  b233039   Merge branch 'textextract' of git://github.com/ogrisel/scikit-learn into textextract
       new  526c0be   Merge branch 'textextract'
       new  95e6482   Convert sparse matrix to CSR format in grid search.
       new  b18b906   Fix imports.
       new  237370b   Pass kwargs to mlcomp loader.
       new  cf0c08c   Fix SGD-based binary classification example.
       new  c6b564c   Note on fit_transform.
       new  41aef36   Merge branch 'master' into ova
       new  dad8deb   SGD supports multi-class classification using one-vs.-all.
       new  9319bd3   SGD multi-class documentation added.
       new  de4caee   SGD classifier supports multi-class with OVA.
       new  41853ac   documentation for multi-class sgd updated.
       new  012744b   update SGD example to showcase the new OVA implementation
       new  b6cd699   Cosmit: explicit what OVA means as much as possible.
       new  2ceca5c   cosmit in k_means module
       new  1971791   FIX: better k-means tests + fixed broken array init
       new  d22c4f6   FIX: potential division by zero in scaler
       new  1643abe   FIX: fixed more cheesy NaNs than an Indian restaurant in Paris
       new  76f5b54   New example to demonstrate the KMeans API with various init strategies
       new  0e7549a   Test compute Gram matrix with support vectors only.
       new  aead101   cosmit + typos in doc
       new  c84ce87   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  41128af   Activate stop word removal by default.
       new  b862a31   Add vocabulary property.
       new  fada652   Changed docstrings for coef_ and intercept_ in sgd package. Wrap intercept_ in an array in the case of binary classification.
       new  077e7ef   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  3da4b3c   Liblinear docstring modified: deleted irrelevant attributes support_ and changed shape of intercept_ and coef_ accordingly.
       new  13e059f   Added dense implemenation of SGD. Modified plot_separating_hyperplane to use dense instead of sparse version.
       new  8233848   Merge branch 'dense' of github.com:pprett/scikit-learn
       new  9ea143a   Commit broken cython header import. See thread "Sharing Declarations Between Cython Modules" on the scikit-learn mailing list.
       new  cf19455   moved sgd_fast_sparse from sgd/sparse/src to sgd/src. ImportError stil present.
       new  5c49a6a   Moved sgd extension modules from sgd/src to sgd. Updated setup accordingly.
       new  7f153e4   performance improvements in cython files; cython files rebuild. test cases refactored.
       new  ee8d60e   added covertype example for dense sgd.
       new  11b1b79   bugfix in plot_loss_functions (import loss functions). added dense SGD to documentation.
       new  3c82e5e   covertype example now downloads dataset automatically.
       new  9d5e16b   trailing spaces holocaust
       new  e850cf7   let me introduce the culprit of the last checkin
       new  b3c6695   Merge branch 'master' into dense
       new  2720bd8   more trailing spaces cleanup
       new  2ebd632   Updated sgd documentation with multi-class documentation. Added decision surface plot of iris dataset for sgd. changed covertype example.
       new  d2774e6   ignore downloaded data from example
       new  f92bc36   Various improvement in low dim classification example
       new  8a72123   Merge branch 'dense' of git://github.com/pprett/scikit-learn into dense
       new  f92c8da   remaining conflict markers in previous checkin
       new  7c24d50   cosmit in SGD example
       new  64cf4f7   s/libsvm/liblinear/ in classification example
       new  cb46e57   remove the dependency to explicit ABC to keep 2.5 compat + PEP8
       new  2e980bc   make the dense SGD code & docstring more readable
       new  22667ac   more precise docstring in base SGD class
       new  c00b843   forgot to finish a sentence on regularization in a docstring
       new  09fc1e1   more docstring love
       new  4fef7da   cosmit
       new  65361d8   use multi proc in multiclass SGD by default
       new  8a5a50f   cosmits in SGD tests
       new  7b1fc52   more comits in the sgd tests
       new  d6b829f   cleanup
       new  1b74a9c   cosmit
       new  e9ad1ca   better test file name
       new  83bde98   reuse the dense SGD test suite for the sparse variant using test case inheritance
       new  02100fc   cosmit
       new  f35ce7e   add GMM.predict_proba
       new  e3fe8b4   add default initialization of GMM.weights to constructor
       new  eca22e7   cosmits in the SGD pyx files
       new  cd74e32   rename GMM.n_dim to GMM.n_features to be consistent with the rest of the scikit
       new  29356ef   more cosmit in pyx files
       new  1418c7b   add HMM.predict_proba
       new  9583c7a   more cosmit in example
       new  431d8f4   rename HMM.n_dim to HMM.n_features to be consistent with the rest of the scikit
       new  60c2a25   fix pep8, pyflakes errprs
       new  ca7b39a   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9db2bdc   PEP8 in SGD tests + docstring
       new  ba6f0dd   Use n_jobs=1 as default value in SGD module.
       new  1b27811   docstrings: n_jobs defaults to 1.
       new  526a091   better looking docstring for sparse sgd
       new  da54f24   more info on loss and penalty params for sparse SGD
       new  c740c07   Fix small typos.
       new  aa6cffa   cosmit: color of data points matches color of decision regions and OVA hyperplanes.
       new  6446631   ENH: adding partial support for predict_log_proba in Pipeline and log reg
       new  dbbcf92   propagate spelling fixes to the dense SGD docstring
       new  e7c11d9   ducktyping in analyzers
       new  0cbd220   work in progress on vocabulary dimension restriction
       new  5139300   merged the mainr epo
       new  19eb852   small fixes + updated the tests
       new  518f28a   cosmit
       new  05e43c3   add note on fortran contiguous memory optim for the X array
       new  7f42d5d   Merge branch 'master' into textextract2
       new  1822d00   OPTIM: vectorizer with predifined dictionary 5x faster by eliminating scipy.sparse.vstack calls
       new  7ddce98   some optims in the text preprocessors
       new  a7dd03a   OPTIM: sparse vectorizer uses COO a init
       new  1abd070   multi-line print cosmit
       new  426dc63   use a SGD model in the mlcomp demo since it is the fastest for this problem
       new  9498502   cosmit
       new  9ae02dc   make it possible to do fancy indexing on filenames
       new  a21319d   move the mlcomp SGD example as a generic 20 newsgroup classification example with sparse features
       new  0492a5c   Cosmit
       new  7bc6081   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  f3e3fcd   MISC: Recover changes overidden by manual merge.
       new  f302617   Make max_df to 1.0 by default.
       new  7bdfa82   Update matrix type in documentation.
       new  fe058f0   Fix broken test.
       new  7816697   warm start optimization changed from coef_ to init_coef_ and intercept_ to init_intercept_. Modified multi-class SGD: now self is not passed as a parameter to joblib but rahter its attributes. Documentation changed accordingly.
       new  c52da98   Multi-class documentation for module sgd added.
       new  c193506   Unique URL for release-specific doc
       new  893a447   Cleanup in libsvm bindings.
       new  e0ef8e0   Cosmetic changes in gmm.
       new  18534b6   Improve docstrings in metrics.py
       new  428c567   Cosmetic changes
       new  fe0d05e   DOC: Add new installation media and a note for pythonxy users.
       new  dea782c   FIX: prefix with plot examples that produce output image.
       new  ce9f68e   New implementation of LARS algorithm.
       new  1ebb8a2   Add a test for lars_path.
       new  08ad1b3   Fix typo (wantto -> want to)
       new  caf832b   cosmit
       new  60002fd   better pipeline notation in vectorizer + classifier grid search example
       new  60464a5   4 more years!^W^W^W 1 more test for vectorizers with max_features
       new  f061783   Include models with L1 and Elastic-Net penalty. Optional arguments for printing the classification report and confusion matrix.
       new  337f25f   cosmit
       new  fa5cf87   changed init_coef to coef_init (intercept likewise). Changed coef_init to fit parameter.
       new  af313bb   Merge branch 'warmstart'
       new  d36dda3   rewriting f_oneway in the scikit to avoid useless recomputations and 3 duplicates of the data in memory. It should probably be fixed directly in scipy. Adding test to make sure that both functions give the same results + a bit of PEP8
       new  04acf56   Merge branch 'master' into log_proba
       new  44352e8   factorize out shuffling dataset since it might be useful by default
       new  55c2330   new example on how to use pipeline / grid_search for extraction parameters
       new  2f03f9a   adding comment to explain the reimplementation of f_oneway
       new  b8ab93c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2883dc8   Merge branch 'master' into log_proba
       new  5376e39   typo
       new  cd76552   Commit of a 'Gaussian Process for Machine Learning' module in the gpml directory. The module implement a class named GaussianProcessModel. I also add doc, examples and tests (involving a coupling with the cross_val module).
       new  011e0c2   Correction of a bug in test_gpml.py (now runs perfect on my machine!). I just don't know how to involve this test within the whole scikit testing procedure (nosetests). Also add a modification of the TOC in doc.
       new  7dcad4a   Correction of a bug in the basic regression example.
       new  e8619da   Delete the old kriging.py module
       new  f8a8d43   sample run output in the grid_search_text_extraction_parameters example
       new  7e3f1d4   reST formatting of example
       new  6f315a6   cosmit
       new  8277a24   better title for the mlcomp example
       new  0961bcf   better example filename
       new  eeb0168   reference new example in the documentation of the grid_search module
       new  6e474fe   Modification of the score function. The score function now evaluates the deviation between the predicted targets and the true ones. This is for convenience only because it allows then to use the distributing capacity of the cross_val module. The old score function is renamed with the more explicit name: `reduced_likelihood_function` (see eg the DACE documentation).
       new  dd637c2   Modification of the main __init__.py file of the scikits.learn package in order to load the gpml module and tests.
       new  dbfd17b   Renames as suggested by Alexandre. Simplification of the examples. Remove the interactive contour label picking in the probabilistic classification example.
       new  b89c762   Bugged example after modification. Now correct!
       new  cecbb13   cosmit
       new  1167f37   Added new example on modeling the geographic distribution of species. Showcases some advanced plotting tools for python (e.g. basemap). Uses OneClassSVM for density estimation - probabilistic outputs would be nice...
       new  5077393   I Ran the PEP8 and PYFLAKES utils and corrected the gaussian_process module related files.
       new  d81d52d   Can't comply with contradictory PEP8 rules on some specfic code such as: y = function(- x) : it asks for spaces around operator, but spaces after parentheses are not allowed :(...
       new  b597363   ENH: automated class_weight for SVC on imbalanced data
       new  7b6b1f6   class_weight="auto" for liblinear-based and sparse classes.
       new  2aef9d4   Merge branch 'speciesmodeling' of git at github.com:pprett/scikit-learn
       new  cf6fd1e   added species distribution example as plot example. added the two data archives and directories to gitignore (both for root and doc directory).
       new  fa53cd4   if possible, species distribution example now uses basemap by default.
       new  2d2dfb8   deleted old species_distribution_modeling example.
       new  1285539   cosmit: pep8 and author
       new  428d457   removing use_svd option in LDA. Only scipy SVD is supported. + PEP8 in LDA/QDA
       new  d59b35c   Merge branch 'master' into log_proba
       new  7cc2577   ENH: adding predict_log_proba to LDA and QDA + tests to reach 100% coverage
       new  8f4b156   ENH : adding support for predict_log_proba in Naive Bayes
       new  e6e13f1   ENH: adding support for predict_log_proba in SVC and sparse.SVC
       new  58ecc20   ENH : adding predict_log_proba in sparse logistic regression
       new  226a871   remove obsolete bench_lars.py
       new  86eabc1   FIX: replace nsamples --> n_samples in svm docstrings.
       new  68036a7   I removed the time-consuming test and made a regression example from it.
       new  ce38790   FIX: make sure class_weight='auto' do not change the result for balanced problems
       new  845acfa   Merge branch 'master' into log_proba
       new  84b9963   API : implement coef_init as fit parameter in glm.coordinate_descent module. Warm restart is not the default behavior anymore.
       new  8daafda   API: exposing fit_intercept params in LassoCV and ElasticNetCV FIX: making sure that X and y are not centered multiple times in *_path functions
       new  e308765   ENH : adding test in pipeline + increase coverage
       new  13bbcc6   fix doc generation pb introduced by previous commit
       new  971a8d5   Reduced memory consumption in covertype example due to memory leak in np.loadtxt.
       new  68300d8   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7efcff4   Replaced np.matrix(A) * np.matrix(B) by np.dot(A,B), so that the code is a lot clearer to read... Implemented the solve_triangular function accounting for backward compatibility (following Fabian's trick).
       new  dc1c7e7   Removed plotting command from the examples in the GaussianProcess class docstring.
       new  9d35f4c   Remove BaseLib class.
       new  ce2d716   Implement make html-noplot for building the doc.
       new  4e36414   FIX: fix class weight auto
       new  4524896   Added note on the importance of shuffeling. Minor changes in text.
       new  f0de23c   Runtime improvement of species distribution example (fancy indexing).
       new  2f7e6a0   set basemap as default.
       new  ef29d5a   Class weights for SGD similar to svm package. Same heuristic as Liblinear for multi-class (OVA): use only weight for the positive class.
       new  6cb85a1   BUG: Fix metrics to run on 2.5
       new  1622c42   removed parameters `p` and `C` from OneClassSVM (dense and sparse). Completed parameter list of SVC, NuSVR, and OneClassSVM.
       new  238ab74   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  da3f3fa   added tksvm from git://gist.github.com/673953.git.
       new  074be5e   use np.fromstring to load data from large csv text files.
       new  d22c9ec   Fix math rendering in SVM documentation.
       new  264a789   Simplification of input's shape checking using np.atleast_2d()
       new  512e670   Update libsvm docstring with latest API changes.
       new  60f72c9   Rename predict_margin --> decision_function.
       new  600d2a5   Indentation fixes in libsvm bindings.
       new  e42d4ea   changed predict_margin to decision_function
       new  29f3fad   Performance improvements in LARS.
       new  06c2b17   more s/predict_margin/decision_function/ in examples
       new  8993f8a   Changes in format of the fit() input (np.atleast_2d for X, and np.newaxis cat for y).
       new  5233f8e   Force y to np.array before concatenating np.newaxis in fit().
       new  50f9d90   FIX: typo in custom score_func in grid_search
       new  8fbc770   Modifications following Gaël latest remarks.
       new  a95fd66   Added Welch's MLE optimizer in arg_max_reduced_likelihood_function() plus reference in the docstring.
       new  30b8958   Correction of a minor typo error in correlation_models docstring
       new  ba35ae5   initial face regonition example using eigenfaces
       new  f88088b   FIX: better handling of NaNs in precision / recall / f-score metrics
       new  ade2f19   Merge branch 'faces-example'
       new  6b5f857   face recognition example using eigenfaces and SVMs
       new  1b1bee0   more explicit subplot titles
       new  570d3c2   ENH: Cosmetic improvements to the face example
       new  b2b1e9c   Adding random projections SVD to scikits.learn.pca as an option
       new  8cf803d   MISC: Cosmit+Doc in fast truncated PCA
       new  4b310a6   cosmit
       new  fdd5bd3   Better heuristic in LARS.
       new  569aac8   MISC: Remove redundant code and cosmit
       new  85e3370   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e50c37c   ENH: Update embedded joblib to 0.4.6
       new  51438e4   Merge branch 'svmgui'
       new  77605f7   ENH: import symbols on subpackage's __init__
       new  126c038   API: Return self in _set_params
       new  7f4e6d5   added GUI example for SVM. Example now shows support vectors.
       new  b90d36e   added tksvm from git://gist.github.com/673953.git.
       new  fa04de9   added GUI example for SVM. Example now shows support vectors.
       new  c6d7818   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  ac76435   Merge branch 'importanceweighting' of git at github.com:pprett/scikit-learn
       new  ab42993   BUG: svm_gui: C is not defined for OC-SVM
       new  88bad1f   RegressorSGD added. Changed SGD -> ClassifierSGD added regression example.
       new  dede5c4   Merge branch 'master' into importanceweighting
       new  905d0e2   changend "squarederror" to "squaredloss". fixed weighted class example.
       new  61b3753   ENH: Raise error when cloning bug estimators
       new  72a0c56   Add support for np.float32 matrices in lars_path.
       new  16845d9   Add parameter precompute='auto' for *LARS classes.
       new  5571f7e   Some LARS refactoring.
       new  f00bdfa   Rename scikits.learn.gmm to scikits.learn.mixture.
       new  b80b6bb   Improvement of the documentation with a piece of code and reference to the regression auto_example. Add a README.txt file at the root of the examples/gaussian_process directory.
       new  ea9bce7   From: agramfort: don't use capital letters for a vector. Y -> y. Done!
       new  a516587   Forgot to retire pdb... Again!
       new  4bc781b   Forgot one capital Y in the piece of code of the RST docpage.
       new  e771f4e   BUG: Deal with 1D data in preprocessings.
       new  e80be9e   BUG: fix cross_val and GridSearch in unsupervised.
       new  487b25b   automatic refitting on radiobutton change and add example. changed button caption to fit.
       new  8e9b35e   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  d739aa9   Removed trailing spaces in the RST doc page. Removed useless comment blocks in the code.
       new  2877985   BUG: Fix GridSearch in unsupervised
       new  d2ec8a3   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  4d8dff6   metrics.explained_variance was renamed to metrics.explained_variance_score so that I needed to modify this example.
       new  b27d3c5   BUG: Fix the doc-generation of examples
       new  8e9ac4f   Cosmit
       new  f828e65   MISC: Fix example
       new  439f734   DOC: minor changes in gaussian_process docs
       new  22e81ea   BUG: Fix missing gaussian_process subpackage in setup.py
       new  c869ec3   FIX more missing files in setup.py
       new  c37538b   API: Remove long-depreciated function
       new  d6cbe7a   BUG: FIx doctests broken in previous commit
       new  8df209b   Update developers info.
       new  3c264be   Add GridSearch and GridSearchCV to the class reference.
       new  51003b2   Update svm docs (content of dual_coef_).
       new  6826edc   Account for lower=True option in solve_triangular.
       new  c98e6db   Do not import gaussian_process from top level __init__.
       new  5fb02c0   update NuSVC docstring.
       new  eda5098   changed loss function names in SGD (squaredloss -> squared_loss; also for modified_huber). Fixed examples after SGD class renaming. COSMIT in sgd documentation.
       new  9ee59dd   Fix failing doctests in gaussian_process.rst.
       new  ffb77ca   Fix GridSearch does not exist.
       new  2fed2ab   Give credit for web page layout.
       new  68bc58d   glm --> linear_model rename holocaust.
       new  f20d516   Welcome Vincent Dubourg.
       new  26bc62e   Update AUTHORS information.
       new  928306f   DOC: documentation CD Enet fit parameters
       new  ecce971   Initial support for weighted samples in svm module.
       new  f6a22f6   Cosmetic changes to web page layout.
       new  2573ebe   Fix example paths for GMM after renaming.
       new  093ad2a   Update class reference list.
       new  817b17a   Cosmetic changes in documentation.
       new  cb07c3e   Add sgd.* to class reference.
       new  cec8e33   Move benchmarks outside the source tree.
       new  7b2d98e   Fix precompute keyword in LARS.
       new  544258d   Update LARS benchmarks with latest API changes.
       new  482bacc   DOC: Cosmit in docs
       new  a63dfb4   Removal of the submodule relative imports in the toplevel init file. I was able to run all tests without any error nor failure. I was also able to build the html doc (so all examples run fine). @Gael: if you confirm this by pulling my branch, then I'll commit this to the master.
       new  320ee33   DOC: score is reserved to 'better is higher'
       new  33a0f4e   DOC: Better plotting in RFE example
       new  2d0ce09   FIX: actually truncate the SVD to make it faster + add some test
       new  6c1c6c3   forgot the test file in my last checkin...
       new  8fa07c3   drop the warning since useful even if approximate as demoed in the faces example
       new  6a5d4c2   make fast_svd deteriministc by default while allowing to pass rng seeds
       new  1d2a66b   test singular values as well
       new  3976608   Adding the power iteration parameter to fast_svd (to make it better in high-rank very-big very-sparse matrices according to the Martinsson et al survey
       new  29480da   Merging the rng changes
       new  c9abe50   pep8 in plot_weighted_samples.py
       new  01e93e7   ENH : adding kneighbors_graph to build the graph of neighbors as a sparse matrix MISC : pep8 in neighbors module
       new  034a5dd   FIX fragile doctest
       new  fc5da45   new benchmark: comparing SVD implementations
       new  327be9e   remove useless import
       new  c0b4aa4   ENH : adding NeighborsBarycenter for regression pbs using k-Nearest Neighbors API : change api for kneighbors_graph function
       new  531f25f   more documentation on fast SVD + missing reference
       new  6f54797   DOC: adding NeighborsBarycenter to doc
       new  e86ffc1   DOC: better docstring for  barycenter_weights function
       new  9b4281e   DOC: even better docstrings in neighbors
       new  dff3752   MISC: reindenting BallTree C++ code (no tabs + 4 spaces)
       new  eb76161   DOC : more on docstrings in neighbors.py
       new  5a5d2fc   PEP8 + various cosmits in sample generators
       new  078bee9   more tests for the iterated power refinement of the Martinsson randomized SVD
       new  734467a   ENH: make the PCA transformer use the iterated power refinement by default
       new  f068df4   review of gaussian process module
       new  cd8c6b0   one more test for SVD
       new  cb50708   Cosmetic changes in plot_weighted_samples.py
       new  9c4cdc2   Add cross-references between LassoLARS and Lasso.
       new  8ef1fa4   Welcome to Alexandre Passos
       new  46a3108   API renmae k->n_neighbors
       new  dd3ba58   added Oliviers ElasticNet convergence test to SGD.
       new  0dcfcd7   move sgd into linear_model and rename sgd to stochastic_gradient.
       new  fe66255   finalized sgd module renaming. renamed ClassifierSGD to SGDClassifier (same for Regressor) updated setup.py files updated examples updated docs
       new  1d69a72   Barely functional NMF implementation.
       new  6699ab2   Updated the example with doctest tags.
       new  f2dc157   ENH: Small tweak in BaseEstimator repr
       new  e61c7df   OPTIM: do not allocate a (n_samples, n_samples) temporary array with scipy.linalg.qr when (n_samples, k + p)) is all what is needed
       new  cdeee0d   OPTIM: fast_svd now has a auto transpose mode that switch to the fastest impl
       new  1b609c1   cosmit
       new  560c6c7   More rename in the sgd module.
       new  f300323   switching back to scipy.linalg.qr with econ=True to avoid half-installed numpy issues with wrong lapack bindings
       new  e96467c   ENH: prettify web page layout.
       new  fea5d18   Some love for scikits.learn.svm.
       new  0f51919   Cleaned some syntax, implemented more flexibility.
       new  97259f0   FIX web page layout for very long paths.
       new  07fa9c5   Update LARS documentation.
       new  8ace83f   Fix for linear_model.rst
       new  cd0fff3   More love for rst docs.
       new  d102d22   moved sgd examples to examples/linear_model and added sgd prefix. minor fixes in some examples. updated sgd.rst
       new  182a40f   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  2aa32c1   Merge branch 'master' into sgd-rename
       new  e993fe1   ENH: Add control of the dtype in img_to_graph
       new  313cdf2   ENH: dtype is img_to_graph defaults to input dtype
       new  5263595   COSMIT: smaller data points
       new  9df6c7f   updated SGD documentation (referenced in linear_model.rst and classes.rst). COSMIT in stochastic_gradient.py docstring.
       new  fdc0aa0   Merge branch 'log_proba'
       new  cc11039   ENH : improving the speed of ridge with inplace computation + symmetric pos def constraint
       new  52f9e46   Merge branch 'neighbor_barycenter'
       new  46a2d55   fixed imports in non-auto examples. COSMIT import paths in plot examples. Rephrased SGD outline in linear_model.rst
       new  a990780   BUGFIX in sparse.SGDRegressor added more tests for SGDRegression
       new  ac851c7   Merge branch 'master' of git at github.com:scikit-learn/scikit-learn
       new  7fceccb   Merge branch 'sgd-rename'
       new  4384f2d   Like it or not, we depend on setuptools.
       new  085544e   Use original diabetes data as shipped by the R package lars.
       new  ccb58f1   rename lars --> least_angle
       new  2057dd9   Remove duplicates in linear_model/__init__.py
       new  dda59b5   FIX: numerical instability in Rdige regression tests
       new  3a4d8d4   cosmit
       new  dbacdf8   new example: principal eigen / singular vector of the wikipedia graph
       new  913f9de   Better docstrings in the example
       new  ac42e38   simpler SVD benchmark: use the sample_generator utility and fixed effective rank
       new  10bb5bc   scikits.learn.gmm -> scikits.learn.mixture
       new  9847703   moving real word examples to the applications subfolder
       new  c3c0ae1   better gitignore data archives
       new  7b68216   s/_sparsedot/safe_sparse_dot/g
       new  dd71ace   even better .gitignore (teasing...)
       new  5add50c   cosmit on PCA module
       new  fef0018   avoid global variable in test
       new  529c84f   Fix typo.
       new  4cf4025   Use relative imports in datasets.
       new  ee33d7b   FIX: sparse svms do not accept callable kernels.
       new  1d30b96   py3k fixes: callable has been removed.
       new  c7832c5   Py3k compatibility
       new  758e5c5   Remove redundant site.cfg parsing.
       new  f64e31c   Update status of py3k support.
       new  5cee9a3   Cosmetic changes in LARS.
       new  bb9aa25   ENH : coordinate descent speed up when n_samples > n_features in cd_fast.pyx
       new  ea494a0   ENH: make the PCA transformer perform variance scaling by default + update the face recognition accordingly
       new  29d1f2b   ENH : allowing Gram matrix precomputing in Lasso / ElasticNet to speed up fitting when n_samples >> n_features. precompute='auto' is used to do what's best depending on the dimension of the pb like in LassoLARS
       new  c3efcd9   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  ef52e14   FIX: GridSearchCV refit did not propagate the fit params
       new  7e2777d   switch whintening off in PCA by default + ensure unit scale + better docstring
       new  42c246b   use a grid search for the SVM params in the faces example
       new  0992637   updated lasso benchmark to showcase the region where LassoLARS is faster than Lasso CD
       new  86fe600   OPTIM: ensure lasso_path aligns the data only once in if not alread fortran contiguous
       new  494dd3f   ENH : speed improvement in lasso_path with precomputed gram matrix - removing code duplication between lasso_path and enet_path - regenerating cd_fast.c with cython 0.13 (faster)
       new  4f58c2a   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  8139fff   pep8 in coordinate_descent.py
       new  8499379   gaussian_process module changes: - a few cosmits in the examples - implementation of real built-in score function for the GaussianProcess which evaluates a leave-one-out estimate of the determination coefficient (as usually done w/ this kind of model).
       new  5142eb6   pep8
       new  157fc0a   ENH: LassoCV / ElasticNetCV now uses all folds data + example
       new  f88ca8c   ENH: make the LassoLARS and LassoCD path examples easier to compare
       new  81b911c   make MSE plot of LassoCV more readable by scaling the y axis
       new  cfe2ed8   FIX: update broken tests by last checkin
       new  d9c7d80   switch to base 10 for the alpha logs in the Lasso CD path plot
       new  e5ce482   revert the plot style to the LARS paper conventions
       new  8ff35e0   select the best alpha using the mean of the CV MSEs instead of the median
       new  c6822e3   Fixed svd-based initialization, fixed example
       new  1a40057   cosmit: += assignement replaced by plain = in coorinate_decent (more natural, less confusing)
       new  a2639ac   ENH added R^2 coeff which is now used as score function in RegressorMixin
       new  4cffab4   renamed explained_variance_score to r2_score in linear_model
       new  97d3f1c   adding r2_score : fixed typos and doctest
       new  426f3c7   FIX: correctly add depends files to setup.py.
       new  569a39d   Make libsvm recognize labels in increasing order.
       new  f98502c   Correct array size in decision_function docstring
       new  665b2a5   TEST: sanity check on decision_function.
       new  60391e0   Inverse sign in decision_function.
       new  443d91a   No need to sort predict_proba any more.
       new  87ace90   Add a comment on inverting the sign of decision_function.
       new  24d919d   FIX: order of indices of support vectors in multiclass.
       new  0bbf933   Shuffle globally for iris in test_svm.
       new  591bc15   Divide parameter alpha / n_samples for consistency with Lasso.
       new  3a9a72d   Cosmetic changes.
       new  25b7bff   extract the randomized SVD implementation as a toplevel class able to handle sparse data as well
       new  d7609b3   consistently rename n_comp to n_components
       new  b6dfcc0   Merge branch 'master' into sparse-pca
       new  aaf1a06   update doctest to handle the change in regularizer strenght definition in LARS
       new  01be9a5   FIX: typo s/mean/mean_/g in RandomizedPCA
       new  e39f23b   Merge branch 'master' into sparse-pca
       new  604e8b7   sed -i "s/\<n_componentsonents\>/n_components/g"
       new  6074856   SVD benchmark have a consistent filename
       new  90ac3a0   factorized out correlated regression dataset utility function and updated benchmarks
       new  7b94a59   pep8 + N->n_samples and D->n_features
       new  0d65b1f   do not allocate useless memory in make_regression_dataset
       new  17de0a9   launch test on documentation by default when running make
       new  90f962d   cosmit
       new  c5671db   OPTIM: do not precompute r2_score_ in ElasticNet in the fit call
       new  264629b   do not precompute explained_variance_ in linear model: can be too costly: use r2_score when needed instead
       new  9514c28   new benchmark for lasso path implementations
       new  897aca3   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  57da10f   giving more love to benchmarks (pep8, pyflakes, var names, etc...)
       new  d0a1d6c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  45beef4   revert previous commit regarding mpl_toolkits.mplot3d in bench
       new  89377b9   API : maxit replaced by max_iter everywhere API : removing default computation of r2 coef in coordinate_descent (speed up +     can be accessed simply eg. lasso.score(X, y))
       new  0573d9f   merging master
       new  dbd3a32   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  c09ecff   Cosmetic changes in lars.
       new  11de158   Update .mailmap
       new  62e133c   ENH : new scikits.learn.metrics.pairwise module
       new  cb3326c   Merge branch 'gaussian_process_review'
       new  5141451   temporary test fix for refit instability in linear SVC: a bugfix branch will be open to reproduce the issue
       new  c1e00cc   Merge branch 'master' of https://github.com/dubourg/scikit-learn into dubourg-master
       new  2da02ab   pyflakes in plot_gp_diabetes_dataset.py
       new  94f38b9   renaming plot_gp_diabetes_dataset.py as nothing is plotted
       new  2073d9b   Wrote a few test cases.
       new  621f00c   Merge branch 'master' into nmf
       new  3b0a75b   Merged upstream changes
       new  d8620ed   FIX: fix bug in sparse liblinear: bias parameter was not set.
       new  80bfde5   cosmit (reST formatting of the SGD module documentation)
       new  b45d282   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  df0cbb3   more formatting in SGD reST and fixed docstest broken by last checkin :(
       new  40ac544   Added benchmark.
       new  5739282   Fixed problem with big full covariance matrices: sum,log instead of log,prod for loglikelihood computations.
       new  7e51797   FIX : fix extra parenthesis in mixture ...
       new  defdb91   FIX lda, qda: new numpy.bincount requires integer arguments.
       new  100faa3   Started Changelog 0.6.
       new  88d7968   Change link in plot_face_recognition.
       new  8864e0f   Remove example plot_lar.py
       new  f75d231   FIX: do not invert the sign of decision_function in OneClasSVM.
       new  7aa6ad8   Add missing options to OneClassSVM.
       new  32bb221   web page layout fixes.
       new  4e8a6a3   Remove duplicate docs (sphinx generates this for us).
       new  9776c87   Prepare for 0.6 release.
       new  27eda75   Remove generated classes on make clean.
       new  7d16891   Add notes on fluctiations of liblinear.
       new  9b224aa   Add type info to docstrings.
       new  027de8f   FIX: backwards compatibility for scipy <= 0.8
       new  4784e58   Remove Methods from docstring.
       new  a7047ca   liblinear bias/intercept handling
       new  c8f536e   FIX: scipy 0.9 compatibility fixes
       new  dc956b4   FIX: second argument in euclidean_distances.
       new  362de11   Cosmetic changes.
       new  e4df607   Better version detection for scipy
       new  2791ba8   FIX: stupid mistake.
       new  58aa73a   FIX Stupid mistake
       new  148c5ef   More robust utils.fixes.
       new  778fc4b   FIX: docstring.
       new  c66dc10   FIX: np.unique.
       new  ae5561f   FIX: scipy 0.9 compatibility fixes
       new  93a9882   FIX: second argument in euclidean_distances.
       new  372b89f   Cosmetic changes.
       new  8347580   Better version detection for scipy
       new  7a485af   FIX: stupid mistake.
       new  5a9e193   FIX Stupid mistake
       new  3bf5d15   More robust utils.fixes.
       new  8c4ea2a   FIX: docstring.
       new  e9c9cb1   FIX: np.unique.
       new  eb6eb3b   Start 0.7 development cycle.
       new  e0613aa   Change version to 0.6-rc
       new  8018613   Add AUTHORS to web page.
       new  e440030   Note on LinearSVC.
       new  007b013   Web page layout.
       new  22ab9a6   FIX: update to latest API.
       new  55be2aa   Web page update.
       new  c466ef5   Add AUTHORS to web page.
       new  a3af357   Note on LinearSVC.
       new  5612fc0   Web page layout.
       new  6e424a7   FIX: update to latest API.
       new  5402fa3   Web page update.
       new  ed4d04e   FIX tests when run with scikits.learn.test()
       new  2c3d9e2   Update doc.
       new  19bca16   FIX tests when run with scikits.learn.test()
       new  1e5bd09   Update doc.
       new  a4c2d9d   cosmit
       new  1f42388   Do not support precomputed kernels in sparse matrices.
       new  13d288a   Update version information to 0.6.0
       new  ebed6be   Update Mailmap.
       new  b82c30e   Update authors list.
       new  4dd89ec   Update Mailmap.
       new  4b5eb65   Update authors list.
       new  87741a7   Update README.
       new  073da1e   Update README.
       new  b286c10   Add all doc to generated latex.
       new  3c86f5f   Add species distribution modelling to OneClass examples.
       new  6bc3311   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9ab9d49   Add ward algorithm + feature agglomeration
       new  424e1be   BUG: fix GaussianHMM.fit to allow input sequences of different lengths
       new  5f53629   reviewing hierarchical clustering code
       new  195e3f3   adding missing setup.py in cluster
       new  9891914   FIX remove broken test in test_mixture
       new  2aa3835   FIX: compilation issues under NetBSD
       new  124a956   Debug in GaussianProcess.predict for batchwise computation
       new  e7f3d20   Debug GaussianProcess.predict for variance estimation in 'light' storage mode.
       new  c84326c   Merge 0.6.0 into DFSG (reincarnating libsvm's svm.* -- effectively forked)
       new  f7a5f11   Merge branch 'dfsg' into debian
       new  bf83c35   ENH : nicer implementation of StratifiedKFold now usable with regression ENH : better RFECV that now selects the smallest model among models with equal accuracy MISC : PEP8 in cross_val + rfe
       new  61d4b9b   DOC: updating doc for StratifiedKFold + ellipsis in svm support
       new  88433b2   Add documentation on Ward algorithm
       new  34cdc33   Add documentation on hierachical clustering. Update examples + add example (plot_lena_feature_agglomeration) Delete folder feature agglomeration and add it to cluster. Changes in feature_agglomeration and hierarchical. Add mixin structure to kmeans
       new  05f0870   Disabled use of system libsvm - scikits.learn has its own fork
       new  6b4e757   Updated copyright: DEP5 rev153, entries for cblas/ and lib{linear/svm}
       new  489f3a2   Extended TODO (in emacs-org mode): consider system's liblinear (need 1.7 with a fix)
       new  30cae8e   changelog entry for 0.6.dfsg-1
       new  5dc61b3   Added README.source to describe reason(s) for .dfsg
       new  52af9dd   ENH: make it possible to customize the WordNGramAnalyzer token regexp
       new  81e7f16   debian/watch: mangle debian version - remove .dfsg for comparisons
       new  6c41d5b   Fixed upstream release number -- 0.6.0, not 0.6
       new  104eb54   adjusting dfsg rule -- do not prune libsvm
       new  03a85c1   Extending changelog
       new  1cb7799   k_means_ - added optional rng parameter to work routines
       new  5b9605e   Centering data for k-means before fitting
       new  c182b37   k-means - added verbose-level print after initialization
       new  d0084d7   added faster distance-computation algorithm to k-means _e_step
       new  de91e61   Add other ways to contribute to the doc.
       new  6c4ecd8   Little doc improvements to the grid_search.
       new  7016229   DOC: remove duplicate information.
       new  7f509e5   Remove unused imports
       new  5fcbbbf   PCA train() stores eigenvalues associated with components
       new  a5af706   adding James Bergstra as author of k_means_ file
       new  1563c31   k-means adding all_paris_l2_distance_squared function
       new  73c6024   k-means - modified k_init to use pre-computed distances for faster, clearer code
       new  e66a033   k-means - added support for a callable "init" argument instead of copying all the k_init parameters as optional arguments - invite user to use a lambda or something
       new  38bd96e   k-means - fixed misleading typo in error message
       new  9242e46   k-means - added optional parameters "precompute_distances" and "x_squared_norms"
       new  21122d7   k-means - added "verbose" parameter to KMeans class
       new  c2e17ac   k-means - added copy_x parameter to worker routine and BaseEstimator, allowing optional in-place operation
       new  50d58ec   ENH : adding function to test the significance of a cross val score with permutations in supervised problems
       new  75d63f7   DOC: Add scipy in the install dependencies.
       new  d9792b1   ENH : add possibility to pass RandomState MISC : adding reference ENH : add possibility to shuffle within labels ENH : now using np.random.permutation
       new  ffd195c   s/permutation_score/permutation_test_score
       new  be3cf61   fix pb with nose and permutation_test_score function
       new  acb8015   Merge branch 'permutations'
       new  e7929e6   added optional args to euclidean_distances and removed k_means_.all_pairs_l2_distances_squared
       new  eefd4f1   fixed typo in my previous patch to PCA
       new  8783af1   added PCA.inverse_transform and unit test
       new  41e4bc8   added components_coefs_ (eigenvalues) member to RandomizedPCA to match PCA
       new  2447b10   FIX : really accurate pvalue in cross-val permutation test
       new  4e941ae   FIX : even more accurate pvalue in cross-val permutation test
       new  4beecdc   test_pca - modified to use assert_almost_equal
       new  fb7632f   euclidian_distances - repair special case for when X is Y
       new  9eca46c   Merge branch 'master' of git://github.com/jaberg/scikit-learn
       new  2570383   PEP8
       new  416e02e   more PEP8
       new  05c9d4a   more PEP8
       new  78898a8   DOC: Typo in docstring
       new  2ae8517   style conventions for variable names
       new  a04d703   Corrected a few typos in the documentation.
       new  073ef84   DOC: document better similarity matrix of spectral clustering
       new  3461771   s/euclidian_distances/euclidean_distances
       new  2ed3039   Partial Least Square 2 blocks mode A (PLS) implementation
       new  b56dcde   PLS examples
       new  843106b   DOC: typos in docstring
       new  4980b61   Add LabelBinarizer.
       new  2f6bf57   Add sparse Ridge.
       new  776d0b7   Support 2-d Y.
       new  56b60f1   Add RidgeClassifier.
       new  95817f8   Add RidgeClassifier to 20newsgroup classification example.
       new  9ee7f61   FIX: removed obsolete entries and added current ones for top-level __all__ + unittest
       new  d085a9b   typo
       new  2ae49bd   ENH : cross-val generator can now return integer indices
       new  2eba6e7   DOC: better docstring in cross val with indices
       new  bb5c923   DOC: update RST doc for crossval with indices
       new  859fc50   removing print used for debug
       new  5ec3d6f   Add efficient LOO cross-val for Ridge.
       new  4fa80b6   Add sample_weight to fit.
       new  0f32a96   Add reference.
       new  919c789   Add support for custom loss or score function.
       new  c545534   Add label binarizer documentation.
       new  9326b70   BF: k-fold should accept k==n
       new  012d0d4   BF: k-fold should accept k==n
       new  5b2491e   FIX: allow the trivial border case k==n in KFold CV
       new  e7c837f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0b4caa9   Merge branch 'cv_indices' of https://github.com/agramfort/scikit-learn
       new  ca2e210   ENH : speeding up kneighbors_graph function avoiding the use of a LIL matrix
       new  5359bbb   FIX : in hierarchial cluster + Mixin fix + tests + coverage + PEP8
       new  635258b   Test 2-d y case.
       new  854e55c   Support fit_intercept in RidgeLOO.
       new  3f5cf5f   Forgot to use sample_weight...
       new  b6a6117   Default fit_intercept to True.
       new  83fd598   Add sparse RidgeLOO.
       new  f13afc6   Add RidgeClassifierLOO.
       new  e1d81fc   Add class_weight.
       new  40543d2   Add some more documentation.
       new  3520ca2   Add sample_weight.
       new  750e8f4   Add installation instructions for NetBSD.
       new  678458e   Add dense_output option to safe_sparse_dot.
       new  17d3db5   Use safe_sparse_dot.
       new  91fdb16   Fix problem when output is a vector.
       new  5ec33a1   Add safe_asanyarray.
       new  0429494   Handle sparse matrix in LinearModel.
       new  09a5462   Import necessary modules.
       new  5f8caaf   Fix tests for sparse case.
       new  a56ec44   Add RidgeCV.
       new  a230f5b   Merge dense and sparse code.
       new  6827dd2   Rename to RidgeClassifierCV.
       new  8ee253a   Fix 20newsgroup example.
       new  c601344   Make RidgeLOO private.
       new  3eb9a00   FIX : fix pb in affinity propagation when S dtype is not float
       new  1b6f1ee   Fix test.
       new  a91efb2   fix a bug of affinity propagtion, which is caused by incorrect index
       new  c160074   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  09c53ae   PLS mode A : two estimation algo: NIPALS & SVD
       new  812d45c   PLS: WIP
       new  544e531   ENH: KMeans tolerance parameter renamed tol (as in coordinate descent) and made public
       new  06ac5f9   ENH : adding inverse_transform to pipeline + better handling of coef_
       new  69b3b39   ENH : adding coef_ attribute in GridSearchCV
       new  60d51b2   Revert "Partial Least Square 2 blocks mode A (PLS) implementation"
       new  6933d81   Revert "PLS examples"
       new  47bb5c5   Revert "PLS mode A : two estimation algo: NIPALS & SVD"
       new  be20fd2   ENH : adding inverse_transform to univariate selectors + pep8
       new  99c7dbd   removing old svn id tag
       new  1ff5724   ENH : refactoring Ward feature agglomeration to make it work with Pipeline
       new  5edd39f   first attempt to use caching in gridsearch with hierarchical clustering... WIP
       new  823ed5a   PLS : cosmetic changes
       new  e49cd50   ENH: Reorganise the feature agglomeration
       new  cab8be4   ENH: Accept strings as memory
       new  b1714ba   FIX and more tests for PCA and inverse_transform also for RandomizedPCA
       new  bbc80ae   Predict is already implemented in LinearModel.
       new  29581fa   Fix issue in RidgeCV.
       new  a04a202   PLS
       new  5e5a648   PEP8!
       new  a6041c6   PLS cosmetic
       new  3d5dd2e   DOC: Add the logistic regression to linear models doc
       new  7441f0f   DOC: Be explicite about what criteria are used in GridSearchCV
       new  704ed7e   ENH : improving ward for better joblib caching
       new  951b5f1   removing plot_dendogram function
       new  f529548   TEST : fix ward clustering tests
       new  de06c23   in hierarchical : s/adjacency_matrix/connectivity, s/k/n_clusters
       new  f7eacfd   remaining s/k/n_clusters
       new  a81df0a   ENH (ward): return children as numpy array (better for joblib)
       new  5f49bdf   Merge branch 'master' into asaf
       new  6e9e1c7   Some docstrings added to ridge.
       new  50e8a9a   Rename lb -> label_binarizer.
       new  dffe8c4   Add note on multi-class classification.
       new  2e7932c   Add some more doc to LabelBinarizer.
       new  e27a47a   PLS: optimize, compare against R implementation, clrify terms
       new  d6ce0a4   DOC: typos + change name of LDA vs QDA examples
       new  f8554f5   PLS: simplify API + som additionnal test
       new  269ab1a   Some love for lars_path.
       new  99e8706   Add documentation for the RandomizedPCA class
       new  d5a928f   Fix typo.
       new  5e891f4   Add documentation on matrices used for clustering.
       new  be305bf   PLS: add transform function
       new  c338ece   PLS: test_pls fix a bug
       new  1b38e73   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9539f4f   PLS: transform method - In fit method: compute rotations - This greatly simplify the transform function
       new  2a45816   ENH: avoid storing parent and weights in Ward (better joblib)
       new  33a1e3c   Rename _RidgeLOO to _RidgeGCV.
       new  a942f87   Note on efficiency.
       new  166debd   Improve the documentation for LabelBinarizer.
       new  328642a   Add TransformerMixin.
       new  b09ffd7   Use TransformerMixin in LabelBinarizer.
       new  215ed61   Merge branch 'ridge'
       new  04c10f7   PLS : add predict function + some cosmetic
       new  40c62f4   PLS : predict - add some plot - test of "non regression" for the prediction fucntion
       new  e388349   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  842b06f   Fix typos.
       new  6de70d4   PLS : make sure this also works with 1 dimensional response (PLS1)
       new  43df20f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  ba93382   remove quotes "" on columns names
       new  a36fa7b   ENH: Add inverse transform to univariate_selection
       new  bd3a29b   MISC: Make sure that nosetests doesn't try to run the bench
       new  2a2329d   DOC : better docstring in hierarchical clustering
       new  da0807c   adding example to rst doc
       new  1cd924e   better ward rst doc examples
       new  5d6b058   PLS cosmetic: PEP8, etc.
       new  fb9e0bd   PLS, new specific classes: PLSCanonical, PLSRegression, CCA + some cosmetics
       new  f6cfa8c   add new method for fecthing datadir + reorg os related imports
       new  fb66b5a   checkpoint WIP for the LFW dataset loader
       new  ee48733   Fix TransformerMixin.fit_transform.
       new  a105bf5   added arithmetical ordering patch for labels in linear.cpp and test for liblinear predict
       new  d917da3   comment
       new  abf0e67   simplification in liblinear testing
       new  c1371e3   pep8 compliance in liblinear testing code
       new  47e6e90   PLS: computation optimization
       new  4152157   Remove references to y in preprocessing objects.
       new  be52e04   Merge branch 'master' into lfw-dataset
       new  0332881   fix broken dataset description
       new  abd9a19   moving swiss_roll generator in samples generator better 3D axes orientation for viz
       new  74e3e58   removing Return from class docstring
       new  1c88202   s/cord_/coord_
       new  0445a8f   in setup.py s/ward/cluster
       new  1eb3bfc   simplified liblinear prediction function
       new  6970730   two trailing whitespaces removed from an multiline comment :)
       new  8ece111   minor syntacting improvement in liblinear test function ...
       new  2472ff6   one more minor improvement to liblinear test code
       new  074a8e7   I think i've got it this time ...
       new  31dc61e   pep8 compliant at last!
       new  941c18a   Turn off axis in plot_iris.
       new  c91f099   ENH: implement decision_function for libsvm-based classes.
       new  5116c3d   DOC: svm.rst refactoring.
       new  48fa476   FIX: always raise ValueError on deficient input in BaseLibSVM.
       new  36c83e8   FIX: fixes & tests for liblinear decision_function.
       new  54e4381   ENH decision_function liblinear, sparse variant.
       new  5c3a9eb   FIX: fixes for liblinear decision_function.
       new  bfb819e   ENH: Add a benchmark for ward
       new  667f203   PLS API - PLS class becomes private => _PLS - move/duplicate docstrings from PLS to PLSCanonical, PLSRegression and CCA - of numpy.linalg.pinv => scipy.linalg.pinv - PLS_SVD => PLSSVD
       new  c777ee7   Nicer support vectors in example plot_separating_hyperplane.py
       new  164f0f7   PLS: API (2)
       new  8447bcf   PLS : coeficients computation
       new  578154b   PEP8 fixes.
       new  d5b445c   Doctest fixes.
       new  1927389   API: fit params -> class params in GrideSearchCV
       new  ff758e0   Add sample_weight to Ridge.
       new  0c2d242   Improve documentation for Ridge objects.
       new  7d39bdd   Move cv parameter to constructor in RidgeCV.
       new  fedd549   Temporarily disable sample_weight when cv is passed to RidgeCV.
       new  e34e76c   Preserve backward compatibility in GridSearch.
       new  7430d20   Remove obsolete info.
       new  ad4b3b8   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9702349   Fix error in documentation.
       new  094f433   various changes to handle fortran ordering in matrices
       new  f83bf67   some pep8 fixes .... but probably more to come
       new  3c01b54   removed testing thing
       new  0195d95   pep8 stuff as well removed testing stuff
       new  2485225   Squash function in test_svm.py
       new  a5aa0bd   remove unused.
       new  1696b72   Add RandomizedPCA to RST docs.
       new  8c83816   PCA docstrings reestructuring.
       new  04123ef   checkpointing work in progress
       new  a653928   Merge branch 'master' into lfw-dataset
       new  d4f779d   work in progress on LFW: fetching the data
       new  6826401   more work on dataset loader for LFW pairs
       new  eb60275   Remove coef_ and get_support from Pipeline.
       new  3848249   Add SparseTransformerMixin.
       new  1fa7336   Use sparse.base.SparseTransformerMixin.
       new  7bd9bff   Add documentation on model persistence.
       new  0a8457b   Minor fixes in RidgeCV.
       new  16c2d78   Add reference for GCV.
       new  be24f7c   Add Olivier Grisel to metrics.py's credits.
       new  fdc8044   Comment broken test.
       new  f79b27d   Rename SparseTransformerMixin to CoefSelectTransformerMixin.
       new  bb1627b   Do not resize the array on k=1.
       new  f3b6f57   ENH: Neighbors refactoring.
       new  1883b7b   Add parameter eps to NeighborsBarycenter.predict.
       new  40b3799   FIX: fix dimensions in plot_neighbors_regression.
       new  0a9a5fd   Simpler doctest for neighhbors.
       new  f84d5be   FIX: rename adjacency --> connectivity in kneighbors_graph.
       new  56bf582   Change the algorithm used in neighbors.barycenter.
       new  28316b0   small fix in barycenter
       new  0902839   remove unused imports.
       new  94f65cc   Rename barycenter --> barycenter_weights (as it was before).
       new  ecd7190   Neighbors refactoring.
       new  33764ab   FIX: fix collinearity issues in least_angle.py
       new  b6fc307   Can now specify desired percentage of explained variance ratio in PCA.
       new  26a2561   PLS : check for numerical instabilities + force float
       new  5bad62d   get rid of the normalization that should not be part of the load time
       new  bfa5831   Merge branch 'master' into lfw-dataset
       new  6609703   make it possible to load the LFW people dataset using the scikits.learn.datasets infra
       new  3a28fd3   remove stupid color slicing 'feature' and shuffle the examples
       new  186dcbb   pep8
       new  599471e   better default slice values
       new  d7a1ddd   better looking example
       new  28d972e   Regenerate Cython file _liblinear.pyx
       new  fc3118c   MISC: Docstring formating
       new  80051f3   Add a few sanity checks for SVC.
       new  4c430a4   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bcde363   Merge branch 'master' into lfw-dataset
       new  e4227fd   face verification example will be implemented later
       new  8876d34   Add tests for sanity checks in SVC.
       new  9a62759   Merge branch 'master' into lfw-dataset
       new  6c45c01   Remove arbitrary code in tests.
       new  836e182   Simpler check for orthogonality.
       new  40cbdb5   cosmit typo
       new  10116f3   Add pls to __init__
       new  39b3213   DOC: set up barebones documentation for PLS.
       new  e37d528   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0538690   Merge branch 'pls' of https://github.com/fabianp/scikit-learn
       new  c5aa155   Merge branch 'hcluster' into hcluster2 that matches master
       new  c861253   ENH: Tweaks for k_means performance.
       new  30adcac   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  0c51a34   Cosmit
       new  8b66fce   FIX: do not resize array in knn_brute.
       new  34cbb65   Faster Neighbors* in high dimensional spaces.
       new  7d0f0c4   Use squared distances.
       new  50a98f6   FIX: typos and missing info in docstring.
       new  f90d456   metrics.pairwise has right to live.
       new  4a6e16f   Rename inplace --> brute_inplace
       new  76ed436   ENH: better consistency tests for neighbors module.
       new  340761d   FIX: typo.
       new  a5fce11   ENH: n_leaves = n_samples in ward tree
       new  f39f724   MISC: np.zeros -> np.empty
       new  6ff2ee5   FIX: don't import assert_allclose
       new  af124bd   ENH: Avoid big temporaries in hierarchical
       new  9fdbbcb   MISC: Cleanup
       new  167eaac   So this is why people kept posting issues to SF's trac ...
       new  b8e70e7   ENH: Hierarchial: don't compute moments twice
       new  a6e5739   ENH: hierarchical: gain memory with izip
       new  458c2b2   ENH: hierarchical: simpler, faster without connectivity
       new  29efec1   MISC: labels in cluster -> int
       new  a051072   MISC: Fix ward bench vs scipy
       new  46bd9dc   first test for the LFW loader skipped if missing data folder
       new  29c927f   more LFW tests
       new  4be3a60   pep8
       new  0faf690   Deleted code is debugged code.
       new  16dc776   Flip the sign when the user accesses coef_ or intercept_ in the 2-class case.
       new  62bcf5b   Cosmetic changes in decision_function.
       new  80673ed   Rename strategy --> algorithm in Neighbors*.
       new  8c6ece8   fix remaining n_comp
       new  0bf053f   FIX : fixing Lars lasso with early stopping using alph_min + adding test for it
       new  3bf44af   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  a089f38   fix LassoLARS docstring
       new  d7f731c   MISC: Avoid depending on numpy > 1.4
       new  bef60ef   documentation for the LFW dataset loaders
       new  62a5aae   Merge branch 'master' into lfw-dataset
       new  f59fcb5   new criterion for wards clustering
       new  83ea3b3   one single cython module for inertia and ward distance
       new  3498fb5   Merge branch 'hcluster2' of http://github.com/bthirion/scikit-learn into hcluster2
       new  749b13e   adding test scikit vs scipy.
       new  5c9df2c   FIX: ugly bug in connectivity on grids and images
       new  15bee0f   generate fake LFW dataset to fully test the LFW loader even without access to the real data
       new  3594e47   add HTML coverage report
       new  05adf81   more robustness test checks for LFW loader
       new  1bb6e27   BUG: Add missing import
       new  323a423   first stab at factoring the 20 newsgroups dataset loading
       new  7e65409   cosmit
       new  a7a6431   cosmit
       new  dd6103e   fix kw params propagation to load_files
       new  f0a054b   update the grid search example
       new  5c7ea36   remove function autodoc section that breaks sphinx
       new  5b806ea   better name: rename load_files to load_filenames
       new  123c0f8   better name: rename class_names to target_names for consistency
       new  adc9533   merge lfw-dataset to 20newsgroups-dataset
       new  bbb6033   Improve performance of GMM sampling
       new  7c53b1e   Second patch by f0k.
       new  e76eec5   Cosmetic fixes in GMM.
       new  1f9f965   More cosmetic changes in GMM.
       new  5ea578a   Rename ndim --> n_dim
       new  898ab69   Rename nobs --> n_obs
       new  c07f957   Some more docstring fixes for mixture.
       new  b4e11f3   MISC: Less code duplication in lfw
       new  6c4ecf3   cosmit
       new  642dc8f   Merge branch 'lfw-dataset' into 20newsgroups-dataset
       new  1dab341   Merge branch 'lfw-dataset' of https://github.com/GaelVaroquaux/scikit-learn into lfw-dataset
       new  15ea041   cosmit / ordering
       new  827731e   use explicit parameter passing
       new  69c7ed5   merge changes from LFW branch
       new  d9caa90   Merge branch 'lfw-dataset'
       new  0bb99fb   Merge branch 'master' into 20newsgroups-dataset
       new  36eb56d   some more work on the datasets documentation
       new  1f710cf   improvements to the datasets documentation
       new  0c7a5c3   fix: avoid creating a spurious '~' in the current working directory
       new  1721c04   pep8
       new  0d6d377   Examples cleanup: remove pl.close, it is now handled by gen_rst.
       new  aaac3b3   Changelog for 0.7
       new  4b00058   Examples cleanup: remove pl.close, it is now handled by gen_rst.
       new  a084f38   Changelog for 0.7
       new  3e63ed5   More doc on 0.7 release.
       new  a8e76cc   typo
       new  3152483   missing justification for the shuffling of samples
       new  a2a2209   More on changelog.
       new  bf0d2aa   More doc on 0.7 release.
       new  f5abc1d   More on changelog.
       new  a3bc7ea   ENH : factorizing img_to_graph and grid_to_graph there was no bug before but an obscure img_to_graph(mask, mask) that is now replaced by grid_to_graph (makes a connectivity structure for a grid of pixels/voxels)
       new  4037901   ENH : ones on diag in grid_to_graph + fix dtype
       new  06fffc7   cosmit
       new  4463a83   Cosmit
       new  9b3abd2   Minor fixes in changelog.
       new  8d8e020   Add metrics to the doc.
       new  5f86e57   More fixes for the changelog.
       new  317df23   Merge branch 'master' into 20newsgroups-dataset
       new  d086bb2   Some more changelog stuff.
       new  d10f47c   Minor fixes in changelog.
       new  ec37f0a   Add metrics to the doc.
       new  9d07758   More fixes for the changelog.
       new  2533e92   Some more changelog stuff.
       new  c953f1c   restore python 2.5 compat
       new  4820ae7   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  cbb4355   Merge branch 'master' into 20newsgroups-dataset
       new  09940a9   API: SVMs: eps -> tol
       new  ea51378   FIX: make PCA models usable in pipelines
       new  21175e3   MISC: Fix example to adjust to eps -> tol
       new  170af3f   Merge branch 'master' into nmf
       new  826a325   FIX: mxf --> Xinfan Meng
       new  08a914d   Change release name to 0.7.
       new  61eb1ed   FIX: mxf --> Xinfan Meng
       new  0053662   Documentation update.
       new  40f7ba3   ENH: Fixed seed for shuffling in SGD
       new  8a19562   Replace latex with simple syntax in docstrings.
       new  3cfe32c   Documentation update.
       new  76a95a3   Replace latex with simple syntax in docstrings.
       new  b8a6adb   Start of 0.8 development cycle.
       new  7d6b8b1   BUG: fix grid_to_graph
       new  27acf91   Merge branch 'master' into 20newsgroups-dataset
       new  68d9939   add backward compat for old load_files public API
       new  c66b2dc   Building on Windows.
       new  ce89c7c   Build precompiled windows binaries.
       new  1feca95   ENH: make transform() work when no Y is given.
       new  2234464   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  63177aa   boilerplate
       new  c092530   typos
       new  9c3bbb0   Replaced wrong k-means++ implementation with a correct one.
       new  8d89358   Extended docstring, renamed variables from javaStyle to python_style, replaced tab-indents with space-indents, pep8
       new  5bbf853   Use scikits distance functions instead of scipy's. Avoid recomputations of x_squared_norms whereever possible. Completion and unification of docstrings.
       new  b58e9d5   MISC: cosmit + use private prng
       new  39e7a68   MISC: fix typo
       new  ed82d42   more tutorial boilerplate
       new  f1fac7a   missing download script
       new  1dae5af   cosmit
       new  7a9892a   typo
       new  2a3f367   typo
       new  39673b6   some work on the introductionary section
       new  d5c72b3   more work in progress
       new  ebaf6d6   better title structure and missing paragraph on good feature extractors
       new  0430f21   more work on machine learning 101
       new  aa35283   cosmit & typos
       new  b5c9ac7   wording
       new  90aa893   missing fetch script for movie reviews
       new  d859f40   work on exercise 01
       new  5b18c58   better use this a exercise number 2
       new  dee404f   do not forget to introduce linear separability
       new  7703b95   typo
       new  c75b74b   add missing fetch script for the face dataset
       new  71a65de   solution for exercise 04
       new  40beda3   extracted skeleton for exercise 04
       new  8b1b194   more work on the classifier section
       new  a99e49c   starting explaining PCA
       new  d3047a1   useless conf
       new  ed7a25e   more work on PCA
       new  b49570d   better titles
       new  fa2aa8f   style
       new  dc4631c   more work on general concepts (esp. supervised learning)
       new  07f8d62   more syle fixes
       new  232dfd0   more subtitles
       new  d9dbb7f   Remain compatible with numpy 1.2
       new  b8f6e71   Do not import scipy.sparse globally.
       new  7fa40ad   Remain compatible with numpy 1.2
       new  0eeaa3a   Do not import scipy.sparse globally.
       new  8ef9ce8   Implement probability estimates for SVR and OneClass.
       new  5a6d911   Raise NotImplementedError on predict_proba when model do not implement it.
       new  4d31f55   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6cd9679   resolve conflict
       new  787e03d   Merge branch 'pls' of https://github.com/fabianp/scikit-learn
       new  f17ebac   added a converged_ attribute to GMM to indicate whether fit() returned because of convergence or because max_iter was reached.
       new  33ee23f   reset GMM.converged_ when calling fit() again
       new  434370d   Update numpy/scipy requirements.
       new  769155d   resolve conflict
       new  8208b29   Release 0.7.1
       new  d53bb58   samples generators: remove multivariate_normal_from_latent_variables
       new  8972010   Read README.rst for description in PYPI
       new  a8f3977   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b1d326a   split >80 char comment in 2
       new  946daf1   Update numpy/scipy requirements.
       new  df4e5ba   Read README.rst for description in PYPI
       new  910c519   add GMM.converged_ attribute to GMM docstring
       new  73c5a98   ignore workspace
       new  7c847f9   ignore OSX stuff
       new  1df6ef6   updated the link to the official documentation
       new  0f0610a   work on clustering section
       new  a50dc8d   add some subsections separators
       new  31cd361   wording
       new  a0ef2fc   reorganizing notes and adding a scikit-learn oriented complement to the supervised learn flow diagram
       new  caf1876   remove section on density estimation
       new  74e225a   started work on linearly separable data
       new  4651e08   added takeway points section
       new  65d5c0b   cosmit
       new  00bb00d   various improvements
       new  91cfc0e   typo
       new  7ef9e19   various cosmits / wording
       new  123a4aa   typo / wording
       new  8b9f3e3   section on regression models
       new  c28b94d   Merge remote branch 'vincentschut/master'
       new  511ed6c   BUG: Fix bug introduced by PLS
       new  d94119c   DOC: Minor fixes to documentation
       new  21a06cf   section on overfitting and the train / test split
       new  c096796   removing confusing section + slight reorg
       new  ba391d9   wording
       new  ee6f018   wording
       new  28b4d98   work on text feature extraction doc
       new  ae59a7a   one more check in the setup
       new  ebe96db   better intro + more work on text features
       new  cc83dab   more work on text classification example
       new  04d5cd9   fix header levels
       new  14f7590   exercises instructions
       new  499a910   improvements in exercises instructions
       new  19b1256   some practical hints
       new  7101ddd   use the pipeline in exercice 02
       new  b179503   inversed solution and skeleton
       new  1556aaf   sentiment analysis
       new  ae0d700   missing skeleton
       new  fd1cb42   gh pages integration
       new  11edf68   Implement transform in LDA.
       new  d244e49   Add LDA to plot_pca.py and rename to plot_pca_vs_lda.py.
       new  cc921d8   Proper implementation of predict_log_proba in LDA.
       new  2422ed9   DOC: minor spellings and formatting (trailing spaces, consistent spacing etc)
       new  e394ad8   trailing spaces
       new  15916da   pep8
       new  9de3a37   style
       new  60f33dd   add check to the nature of y to have more explicit error messages
       new  9fe5182   explicit ValueError when not enough data for kmeans and some pep8
       new  9d37091   style
       new  e05b795   Added CRO-based initialization, TODO tests, bench
       new  78a6416   make RandomizedPCA work on list data
       new  618a1b5   refactored SGD module (removed code duplication, better variable naming). added interface for sample weight. added seed to interface of sparse SGD classifier and regressor.
       new  dfa75da   Sample weights for SGDClassifier. Additional test cases for class weight and sample weight. Example plot for sample weight.
       new  bc16f01   Additional tests for sample weights. Fixed some typos in doc strings. Changed signature of coef_ for SGDClassifier; now adheres to svm convention that coef_ is shape [1, n_features] in the case of binary classification.
       new  9bf1d7b   Theirs Merge commit '0.7' into releases
       new  e5d9a53   Merge commit '0.7.1' into releases
       new  b4e1432   Merge branch 'releases' into dfsg
       new  622de39   Merge branch 'dfsg' into debian
       new  962427c   changelog for 0.7.1.dfsg-1
       new  f2fcd5c   FIX: the datasets doctest fixture could never skip the tests when required
       new  bf935ec   use WARNING level logs before using network access
       new  430d5dc   make the test display the output on stdout
       new  1f799a4   ENH: add function to clear the data_home cache + tests
       new  6b3e17a   full PEP8 compliance for the scikits.learn.datasets package
       new  637c700   renamed load_* to fetch_* when network connection is potentially involved
       new  82185f1   Merge branch 'lda' of https://github.com/mblondel/scikit-learn into discriminant_analysis
       new  8f3b9ff   add load_lfw_pairs and load_lfw_functions for backward compat and consistency
       new  5b5b709   load_20newsgroups as an alias for fetch_20newsgroups in offline mode
       new  4bab789   re-add self.means_
       new  7170f3d   trailing spaces
       new  908e6a5   break test data symmetry to avoid heisenfailure in RandomizedPCA test
       new  f39c4d6   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  5e06d30   FIX: heisen test failure + some pep8 in test_pca.py
       new  2d2c87f   Fix to sparse SVC with kernel='poly'
       new  bdf3332   oops -- forgot that we are still doomed to experimental due to sphinx
       new  3e9a383   Merge branch 'master' into sgdsampleweight
       new  3db9bbe   pep8 + oliviers remarks
       new  bf4268d   SGD: documentation for sample weights and class weights. Some polishing of the SGD documentation.
       new  833695a   Added Multinomial Naive Bayes classifier
       new  14a66c0   Fix to the documentation of the Multinomial Naive Bayes.
       new  7fe5131   BUG: fix kneighbors method in high dim
       new  dded226   DOC: clearer doc for BallTree.
       new  b7c9201   Add polynomial interpolation example.
       new  ec96240   DOC: improve PLS docs and example
       new  15965e6   better README.rst
       new  3cd8608   use vec.fit_transform instead of vec.transform on the training set
       new  75217f7   Untracked changes
       new  491ba87   Merge branch 'master' into nmf-nnls
       new  0cfdc03   Put CRO inside nmf.py
       new  ceaf255   Pep 8 compliance and cleanup for the multinomial naive bayes
       new  59df234   Merge remote branch 'upstream/master'
       new  ce2e0ef   Sparsity constraints and measures of sparsity
       new  66483cf   Use np.vander.
       new  4c1fb9f   DOC: docstring enhacements for Gaussian Naive Bayes.
       new  4699607   DOC: some documentation for naive_bayes module.
       new  7ae1566   Some more pep8
       new  138e688   Refactoring in svm module.
       new  ade11c3   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  ecb869c   Support multilabel case in LabelBinarizer.
       new  e1ecda6   ENH: better doc and tests for unbalanced svm's
       new  075931d   Python 3 compatibility.
       new  f41112b   Nicer low-level API for libsvm.
       new  476824a   adding boosting and decision trees
       new  5112d7c   adding bagging and gradboost
       new  c6cd221   minor change
       new  1c04b54   Merge branch 'master' into nmf-nnls
       new  30f8d4e   Style fixes all around. Clarified NNDSVD docstring.
       new  d62f1ab   Decreased default NMF tolerance to improve results.
       new  e070407   Corrected sparseness measures in NMF.fit
       new  1d8e3f2   Removed print in CRO.fit; moved utils to top.
       new  eeb9ba4   working on interfacing with Cython
       new  345715e   Possibly fixed errors in doctest (not verified yet)
       new  47663e1   Doctests pass now
       new  c2088bc   Fixed bug in transform (lack of .T), renaming
       new  0e98d9a   Non-negative least squares testing
       new  6168f1b   The derivation of the variational algorithm for the DP mixture of gaussians
       new  f86395a   Renamed tolerance to tol for consistency.
       new  2886f05   minor updates
       new  160c952   MISC: Update joblib
       new  816a257   Ignore OSX .DS_Store files.
       new  95ece32   Revert "Python 3 compatibility."
       new  105e132   FIX: rename eps to tol also in svm.sparse.
       new  2e19d22   Add linear_kernel, polynomial_kernel and rbf_kernel.
       new  80d50d8   Small optimizations for polynomial_kernel and rbf_kernel.
       new  4c54e87   Add KernelCenterer.
       new  cd38b37   added doctests for sparse and dense SVR, NuSVR, NuSVC, and sparse SVC.
       new  0496931   Wrote tests to cover mostly everything
       new  57feef5   ENH: cython bindings for libsvm's cross_validation routine.
       new  ca851cb   Add KernelPCA.
       new  ac8bf20   Add kernel PCA example.
       new  7771006   FIX: make the PIL dependency optional (skip LFW tests if not present) + explicit error message
       new  e8b2b21   lfw: import imread from new location in scipy
       new  3c9113d   always use scikits ward algo when no structure is provided
       new  344c782   Merge branch 'master' into kpca
       new  ff93c02   ENH: Add verbosity to the gird_search
       new  0f103f5   ENH: More parallelism in GridSearchCV
       new  b25978a   ENH: GridSearCV: better verbose
       new  732a64b   TEST: Fix trivail doctest failure
       new  4ecde1e   BUG: iter on complete grid (GridSearchCV)
       new  867fec5   Revert "Python 3 compatibility."
       new  23e118c   FIX: rename eps to tol also in svm.sparse.
       new  7c0319a   added doctests for sparse and dense SVR, NuSVR, NuSVC, and sparse SVC.
       new  0203171   ENH: cython bindings for libsvm's cross_validation routine.
       new  94a189a   FIX: make the PIL dependency optional (skip LFW tests if not present) + explicit error message
       new  54a7309   Beginning the code; so far only doing the E step
       new  a06c42b   Update the installation guide with Ubuntu related info
       new  c47200a   Fix a Broken link in the documentation
       new  afc0715   FIX: workaround broken PIL installs
       new  cceac31   Add KernelPCA documentation.
       new  5fba12e   [base.py] Do not break while trying to pprint not existent attribut
       new  3057073   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  aaeadfb   MISC: html-nodoc default target
       new  f1692db   Merge remote branch 'origin'
       new  e9a3722   Add test for precomputed kernel.
       new  a998e62   NMF example on faces dataset
       new  72d4317   Optim in polynomial_kernel.
       new  321fc8b   Implemented fit_transform
       new  d2fecaf   First draft of the code; untested
       new  af84a63   Efficient fit_transform in PCA.
       new  d7ae8aa   [base.py] revert previous commit, as the error is raised when object does not follow scikit API
       new  345ff41   Merge branch 'master' of https://github.com/yml/scikit-learn
       new  28997f1   Tweaked plot aspect ratio
       new  b9a4154   Refactoring of the covariance estimators modules.
       new  61a39cd   Merge branch 'mblondel-kpca' of https://github.com/vene/scikit-learn into kpca
       new  b58911b   Without pl.show, the figure won't be displayed.
       new  edcfbb5   The dp is already fitting properly
       new  9556b55   Cosmit.
       new  6009beb   Fixed broken tests due to interface change
       new  6940635   Fixing indentation bug
       new  cac1865   Use TransformerMixin in KernelPCA.
       new  62b7144   Tests now behave better
       new  00ba725   TEST: Ellipsis on numericaly instable docs
       new  8b210aa   FIX: cross val return array size.
       new  64961c1   Initial implementation of cross validated SVC
       new  0ec5887   [feature_extraction] Refactor text/* to text.py
       new  d4fd5f7   Python 3 compat, this time with npy_3kcompat.h
       new  5a45d03   Revert "Initial implementation of cross validated SVC"
       new  4c03d21   tiny updates on lda (checks, numerical stability)
       new  ee8b41d   Wrapped BallTree in Cython. Note that this wrapper does not include the BruteForceNeighbors() function from BallTree.h
       new  e0d6767   Merge branch 'cython-balltree-wrapper' of https://github.com/thouis/scikit-learn
       new  92f6d03   Cosmetic changes in base.py
       new  fa3f402   FIX: py3k compat.
       new  5bdc5a4   Renamed for backwards compatibility, fixed C++ Exceptions to propagate to python
       new  46ce94b   removed the unused inerta stuff
       new  cbb0a1b   balltree - be explicit about return types' width
       new  ec8edb6   I won't import scipy.sparse globally.
       new  c34264b   Changing the DP derivation to rst---equations don't work
       new  5f32ef3   Renaming; removed numpy 2-norm
       new  8f3985d   Fixed the math
       new  fd4bbaa   Removing useless whitespace between methods
       new  c6cfce4   Removed useless _fit_transform
       new  acc13ed   CRO inherits from BaseEstimator
       new  00ff9ed   Reorganizing the directory structure
       new  fbe83f8   OAS estimator of covariance + new example.
       new  fcd5f8f   Adding variational inference for a finite gaussian mixture model
       new  3ef468a   Cosmit
       new  bacbe96   Merge branch 'hcluster2'
       new  5cdc7dd   I'm returning precision, not covariance matrices. Make that clear
       new  632364c   cosmits with trailing spaces
       new  d79c683   Editing the documentation
       new  ebf827f   Making it clear that the covariances don't work
       new  1150d28   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6c47a31   BUG: doctest the joblib in externals not global
       new  f5a54e0   Fetching every figure generated by the example scripts.
       new  1ff16f5   Since we changed the name of the figure names, changing the rst files.
       new  54283ad   Applied suggestions; updated bench and example
       new  ebe22cd   Updated doctest
       new  4f9a126   Merge branch 'master' into variational-infinite-gmm
       new  2b5d9ba   pep8 fixes
       new  26d809d   Abbreviation expansion in benchmark
       new  e290c85   Fixed comments in NMF example
       new  0decf02   pep8 on test_nmf
       new  744f25a   NEW: Function to automatically download any mldata dataset given its name
       new  b5d5a89   ERF: load files in "mldata" subdir; some documentation improvement
       new  ec08e4c   Removed comments.
       new  5f171e9   ERF: Error checking in fetch_mldata
       new  73e31ba   Removed CRO for now
       new  4d404de   ERF: fetch_mldata allows to use natural mldata.org names for datasets
       new  c999ea6   FIX: trying to reverse-engineer mldata.org conventions
       new  d7144b8   Added nndsvda and nndsvdar options for NMF.init
       new  189b15e   Merge branch 'nmf-nnls' into nmf-lite
       new  1a95f21   Benchmarks and more pep8
       new  14823e3   Fixed benchmark, removed unused import.
       new  4262312   MISC: cosmetic -- setup.py is now pep8 safe
       new  29c20b0   MISC: cosmetic -- cross_val.py is now pep8 safe
       new  ceadb3e   MISC: cosmetic -- fastica.py is now pep8 safe
       new  d5919d8   Fixed NMF benchmark colors
       new  0e0b917   pretifying nmf plot
       new  71f3f51   MISC: cosmetic -- pca.py is now pep8 safe
       new  2ac779c   MISC: cosmetic -- scikits/learn/setup.py is now pep8 safe
       new  96f9161   MISC: cosmetic -- pls.py is now (almost) pep8 safe
       new  dae7882   Merge branch 'nmf-lite' of git://github.com/agramfort/scikit-learn into nmf-lite
       new  afd1d55   Merged
       new  a758f5e   MISC: cosmetic -- hmm.py is now pep8 safe (getting tiring, next time I'll show up earlier at the sprint ;-)
       new  79a8b97   Fix benchmarks printing of error for alt-nmf
       new  3804b01   MISC: cosmetic -- base.py is now pep8 safe
       new  6842353   Fixing small bug
       new  bd50465   MISC: cosmetic -- grid_search.py is now pep8 safe
       new  2f12774   Documentation. Discussed fixes. Set default to ar.
       new  02e08c3   MISC: cosmetic -- grid_search.py is now pep8 safe
       new  33cd3f1   MISC: cosmetic -- more pep8
       new  677f255   Added KPCA citation.
       new  9013e28   Added NMF to classes.rst
       new  22a251b   Fixed non-ascii characters
       new  0b76350   BUG: restore ellipsis in doctests
       new  416a4a2   Change PCA test to fit just once
       new  2e5dde4   DOC: add the show-source back on html
       new  d0f69c2   Updated documentation with references
       new  dafb32e   Added y=None in fit for pipelining
       new  e3ea050   Fixed relative URI in NMF doc refs
       new  4b27ddd   Clarification of example in NMF doc
       new  4b6195b   Merge branch 'nmf-lite' of https://github.com/vene/scikit-learn into vene-nmf-lite
       new  bae0826   Capitalized Gram, added y=None in fit, pep8 test.
       new  48b0c62   Docstring formatting in test_nmf.py
       new  b61cc5f   Docstrings in nmf.py
       new  1a0c5b6   Merge branch 'nmf-nnls'. Docstring fixes, mainly.
       new  75a5735   Clarified NNDSVD in docstring
       new  9f82387   Documented NNDSVD. Fixed ar perturbation range.
       new  eaced83   Corrected error in docstring re: nndsvdar
       new  d98a159   Added disclaimer in nndsvdar docstring
       new  1b3d505   Merge branch 'nmf-lite' of https://github.com/vene/scikit-learn into vene-nmf-lite
       new  1e5ae89   BUG: fix multiple figure plotting
       new  e242e74   MISC: cosmetic -- setup.py is now pep8 safe
       new  dfad6f8   MISC: cosmetic -- cross_val.py is now pep8 safe
       new  a1f3926   MISC: cosmetic -- fastica.py is now pep8 safe
       new  4e169bf   MISC: cosmetic -- pca.py is now pep8 safe
       new  95b872e   MISC: cosmetic -- scikits/learn/setup.py is now pep8 safe
       new  f4c54dc   MISC: cosmetic -- pls.py is now (almost) pep8 safe
       new  e85fa48   MISC: cosmetic -- hmm.py is now pep8 safe (getting tiring, next time I'll show up earlier at the sprint ;-)
       new  3d0a9e8   MISC: cosmetic -- base.py is now pep8 safe
       new  b572463   MISC: cosmetic -- grid_search.py is now pep8 safe
       new  40f4c2e   MISC: cosmetic -- grid_search.py is now pep8 safe
       new  5a6d410   MISC: cosmetic -- more pep8
       new  b43a30e   BUG: restore ellipsis in doctests
       new  12fa60c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7f4d42b   Clarified invalid sparseness parameter error msg.
       new  56ca515   Clarified init parameter error message.
       new  dcd1975   Transposed shape of components_ attribute
       new  4073e15   ENH: plot eigencefaces in face recognition example
       new  491861a   Renamed NMF to ProjectedGradientNMF
       new  3364960   Updated authors
       new  30f3742   DOC: update rst docs to use multiple figures
       new  79aa6c5   DOC: front page link: Ward lena
       new  078d3c1   DOC: better docs for Ward
       new  e5d8ded   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  2bc0e76   pull from upstream
       new  8868aaa   Merge branch 'master' into nmf-lite
       new  4628b4f   DOC: Added both plots to NMF doc, tweaked plots.
       new  2461a6f   ENH: Avoid gathering old images in docs
       new  0b8aa44   DOC: Made plots look better.
       new  9913182   pep8 in plot_kpca
       new  cf29b7a   ENH: do not download LFW when building the documentation by default
       new  ab506f1   Attributes renamed and documented.
       new  bc4a14d   Merge branch 'master' into vene-nmf-lite
       new  7a80e9f   Merge branch 'nmf-lite' of https://github.com/vene/scikit-learn into vene-nmf-lite
       new  2b47592   Merge branch 'text' of https://github.com/vmichel/scikit-learn into vmichel-text
       new  ee41646   FIX: update the examples to match the new text feature extraction API
       new  f726f6e   FIX: feature_extraction.text is now a module instead of package
       new  61afdc9   MISC: reduce disk consumption when generating docs
       new  98d59ec   COSMIT: make the layout a bit cleaner in NMF docs
       new  5eafac9   BUG: giving up on cleaning build
       new  51fc7ec   DOC: Better center of images
       new  b5e6d4a   FIX: forgot to update the documentation after the feature_extraction.text refactoring
       new  647bce2   FIX: decrease disk usage in LFW data folder
       new  4ecd8e9   ENH: factorize some plot code in face recognition example
       new  2eff92e   Adding example; adding explicit lower bound computation; optionally monitoring convergence; full and tied work, somehow spherical and diag diverge.
       new  44b6890   Began work on decompositions package.
       new  5020fbd   Merge branch 'master' into lda
       new  f577304   Merge branch 'lda' of https://github.com/bthirion/scikit-learn into lda
       new  d47d236   Fix doctest.
       new  37a322c   pep8 love (integrism?).
       new  a20a85e   Add test for invalid kernel.
       new  b3c377c   FIX: very confusing internal naming in NMF
       new  8d4e32a   Rename plot_kpca.py to plot_kernel_pca.py.
       new  5954ffa   Add comment regarding PCA's fit_transform method.
       new  638d1d1   Merge branch 'nmf-fix' into decomposition
       new  ea2aa80   Decomposition module WIP
       new  15d60c7   Merge branch 'master' into nmf-fix
       new  87ebf1e   Merge branch 'master' into decomposition
       new  f88f3e1   Refactoring of the covariance module and examples + add OAS.
       new  1f26c70   Working decomposition package
       new  c6b2284   Merge branch 'covariance' of github.com:VirgileFritsch/scikit-learn into covariance
       new  c91de7b   MISC: pep8ification
       new  2c982c9   Missed one reference
       new  97c4d20   Add note on sign ambiguity in PCA.
       new  8794bb9   Merge branch 'kpca'
       new  13bf29b   FIX: broken link to plot_kernel_pca kernel in the documentation
       new  a5dbe94   Merge branch 'master' into decomposition
       new  aa15792   Using a smaller example to speed things up
       new  0863c5f   Simplifying the code a bit
       new  0c7cd30   Some cleaning in libsvm sparse bindings.
       new  c39a017   FIX: KernelPCA plot in doc
       new  693a4b9   FIX: forgot to track init file in tests
       new  77a7aa2   name consistency in sparse svm
       new  53c92b0   API: components_ shape fixed in PCA classes
       new  8ddde2f   More covariance refactoring: separate MLE computation from object.
       new  1ed5718   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  8747939   First draft of the mini batch KMeans - works, but a lot of cleaning up to do
       new  914d3b1   Refactored: deleted the batch_k_means function, and created an option for the batch_k_means to avoid code duplication - Added some documentation
       new  5a71b60   implemented AdaBoost
       new  239cf8b   refactoring
       new  9a8cb21   minor fix
       new  17a06c7   minor fix
       new  9fa26da   Added test one the batch k_means
       new  6761e37   Improve documentation
       new  42a891e   ENH: More accurate and clean numeric code in PCA
       new  73b1a3f   almost done...
       new  d72d838   it compiles\!
       new  16d3990   now it really compiles
       new  129702f   minor fix
       new  a097bf8   working on segfault
       new  174fc2f   now it works
       new  2657354   trying to fix score bounds
       new  77ba9c6   Cosmit in fast_sgd.pyx
       new  c126a39   ENH: More avoidance of np.dot for diagonal entries
       new  e703800   language detection gives slightly better results without IDF
       new  c429a0b   ENH: low-level API of libsvm.
       new  6166e2e   updates
       new  058f5ae   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  1bb3617   Upload to unstable of -2 + fix of objects.inv and numpy dependency only for -lib
       new  a480d43   sanity check in adaboost
       new  79e7f87   Batch K-Means
       new  86ca325   more sanity checks in adaboost
       new  8067ea1   fairly stable now
       new  311ce7b   Cleanup in libsvm helper.
       new  82bb80e   FIX: important fix for sparse SVC (weights were not initialized correctly).
       new  0c5c87a   Update the docstring to reflect the package name changes.
       new  5062f41   fixed failed doctest. SVR parameter `p` renamed to `epsilon`.
       new  cd01f48   Renamed fastica.py to fastica_.py
       new  c2f1557   Merge branch 'master' into decomposition
       new  39d65f2   Rename BaseCovariance as EmpiricalCovariance + reviews comments.
       new  b56ad22   FIX: Explicit docstring inheritance
       new  fdb417e   [batch k-means] Changed the algorithm to compute the centroids.
       new  79d1830   fixed bug where node cuts were not set but left at 0
       new  98d920c   [batch k-means] Fixed the computation of the batch kmeans centroids
       new  cf10808   working on limiting cases
       new  4d1c12c   updates
       new  39e1d3e   fixing bug in adaboost
       new  05b2395   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  8dc785f   updates
       new  212c37b   resolved init
       new  11b6192   initial import from milk
       new  f05035b   renamed, additional import
       new  e7d4519   started conversion to scikits
       new  a7fb04c   updated information gain and set_entropy functions
       new  aaa5ca4   modified base classes
       new  ee5f93c   updated docstring to reflect use
       new  b6604ef   updated load_iris to return features
       new  c68ad19   enh: updated decision tree classifier and associated example
       new  f64cb7a   updated default impurity measure
       new  5ab8fe6   added new impurity measures
       new  3980abf   updated random forest classifier to operational status
       new  7868f38   updated cython script to calculate gini measure
       new  c9a49a5   minor change
       new  60ec05b   Add kernel PCA and linear PCA equivalence test in its own function.
       new  9f4a53f   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  8a1a996   bagging now implemented
       new  c278f17   removing committee for now
       new  b7267f0   updates
       new  6c74aa1   adding tests
       new  b6c89ee   better demonstration in test module
       new  f641291   minor change
       new  12c577f   bugfix
       new  3beac76   Don't hardcode n_jobs.
       new  fa43d48   Remove useless calls to np.asanyarray and improve computation.
       new  c2f3ac9   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  b8d1be7   removed classifier.py
       new  81b4bce   MISC: Two figures in plot_pca_vs_lda
       new  deed432   minor change
       new  a4a2ee3   typo
       new  f11ee12   pep8
       new  af66795   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn into decisiontree
       new  0b1bdc7   MISC: style fixes in NMF
       new  d6ace52   DOC: move working notes to wiki
       new  add386a   DOC: tweak sidebar
       new  0fa6645   ENH: improved contributors guide
       new  d55aee6   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  c36dd42   ENH: add coverage install command
       new  8a0e6dd   DOC: More sidebar tweaks
       new  5593bc4   DOC: fix link to bug tracker
       new  ed4bdcb   DOC: tweaks to developers notes
       new  b96f027   DOC: move KMeans to top of clustering
       new  dc6f382   DOC: Less warnings during build
       new  1d63e3d   cosmit
       new  23f27d3   DOC: Fix more warnings
       new  3c2e91f   BUG: fix links to examples
       new  f453f51   MISC: cleaner generated code in doc/examples
       new  58299a4   MISC: separate decomposition examples to new dir
       new  c2ba06a   DOC: first stap at the performance chapter (full of TODOs)
       new  e68aca0   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b3d70e5   DOC: missing class reference
       new  d180f8b   DOC: cosmit
       new  0faf093   MISC: another style fix for a private function in nmf
       new  b5143ca   DOC: add sample python profiling session
       new  86f4822   DOC: note for later
       new  d0848df   DOC: add some missing reference in the performance guide
       new  2fe9262   DOC: rmk on Sphinx version
       new  db123c7   ENH: avoid the use of lambdas in NMF to get a more informative profiling output
       new  66b3299   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6619a73   DOC: fix small inaccurracy
       new  f0191da   Minor edits.
       new  8c6d3c9   DOC: add tiny docstrings where missing
       new  09851fc   DOC: Fix warnings
       new  850644f   DOC: module entries in reference documentation
       new  cdfaa98   DOC: fix indentation
       new  b48532c   DOC: link fixes in kernel PCA
       new  cbea7c6   DOC: more warning fixes for the classes reference toc
       new  208915f   FIX: stupid statement in plot_face_recognition
       new  9e60a3f   Added documentation for the Naive Bayes classifiers.
       new  7ad9bcd   DOC: make the face recognition example static (to avoid having to download the dataset to build the doc)
       new  fcca076   Cosmit
       new  731b4c9   Add regularization in the computation of barycenter weights.
       new  f57b26e   [MiniBatchKMeans] Starting refactoring code after the review
       new  5c6a453   Added sparse MNNB and modified the textual examples to benchmark it.
       new  cdd0ef7   Modified the Naive Bayes nose tests to the new location of the module and added sparse test.
       new  a42876f   fix broken doctest
       new  a30197c   [MiniBatchKMeans] Small fixes
       new  5e11d75   Merge branch 'master' into batchKMeans
       new  56fc25d   [MiniBatchKMeans] Small fix in the initialisation for the random initialisation of the centroids
       new  65a455a   [MiniBatchKMeans] Fixed the tests for the new API
       new  8b19136   l1 logreg (liblinear): minimum C calculation
       new  19b8f58   MISC: style fixes in NMF
       new  33d7e74   DOC: move working notes to wiki
       new  7b5e1da   DOC: tweak sidebar
       new  bc5f546   ENH: improved contributors guide
       new  13126c1   ENH: add coverage install command
       new  3d20d25   DOC: More sidebar tweaks
       new  81a24d5   DOC: fix link to bug tracker
       new  9878e58   DOC: tweaks to developers notes
       new  ebd2e4d   DOC: move KMeans to top of clustering
       new  636c68c   DOC: Less warnings during build
       new  1d18223   cosmit
       new  9a0ecea   DOC: Fix more warnings
       new  4a1d5fc   BUG: fix links to examples
       new  3427ff4   MISC: cleaner generated code in doc/examples
       new  95cbfae   MISC: separate decomposition examples to new dir
       new  db9a3c0   DOC: first stap at the performance chapter (full of TODOs)
       new  9470b5e   DOC: missing class reference
       new  45f4c41   DOC: cosmit
       new  365fb94   MISC: another style fix for a private function in nmf
       new  c428a7d   DOC: add sample python profiling session
       new  816d560   DOC: note for later
       new  930d269   DOC: add some missing reference in the performance guide
       new  081035d   DOC: rmk on Sphinx version
       new  4c5d26f   ENH: avoid the use of lambdas in NMF to get a more informative profiling output
       new  144b109   DOC: fix small inaccurracy
       new  a21711c   DOC: add tiny docstrings where missing
       new  b4fbd09   DOC: Fix warnings
       new  10458f6   DOC: module entries in reference documentation
       new  bc929a3   DOC: fix indentation
       new  5d46d78   DOC: link fixes in kernel PCA
       new  be0b052   DOC: more warning fixes for the classes reference toc
       new  1e2c6b7   FIX: stupid statement in plot_face_recognition
       new  030e189   DOC: make the face recognition example static (to avoid having to download the dataset to build the doc)
       new  78733bf   Add regularization in the computation of barycenter weights.
       new  a0085eb   l1 logreg (liblinear): minimum C (sparse version)
       new  91de1c2   review of min_C doc strings
       new  d93c959   DOC: refined the python profiling example
       new  560efef   Fixing last bugs in the bound and updates; improving docs
       new  b2a5669   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into variational-infinite-gmm
       new  0ce6f89   DOC: fix / add more class reference links in perf doc
       new  8d3baaf   wording
       new  0f4aa9e   Fix docstring find&replace issue; restoring VBGMM
       new  788b2c3   Adding reference in the derivation
       new  ae5f820   DOC: started intro YEP
       new  8d7f071   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  b6921f9   pep8 dpgmm.py
       new  d1a16e0   Fixing test failures in mixture
       new  deb9b99   Fixing pyflakes warnings
       new  4052ef0   Adding complexity note to the documentation
       new  052ed25   DOC: use uppercase for project / language names
       new  b07bdee   numpy/scipy idioms as suggested by agramfort
       new  b2b8c2b   pep8 compliance
       new  0c86e6b   min_C: reworked _y calculation
       new  d58cddd   min_C: check for ill-posed problem _y * X == 0
       new  7cc9673   min_C: let's avoid scipy.sparse top level import
       new  0b3a04a   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into l1_logreg_minC
       new  e034f09   min_C: fixes to the doc strings
       new  bda15f5   s / shape = / .reshape() /
       new  c8d5baa   removed float64 and int32 conversion
       new  b21a30b   docstrings updated
       new  99f7701   fix for "removed float64 and int32 conversion"
       new  3897b3e   FIX: last char in char analyzer, max_df behaviour
       new  0de2082   FIX: doctest
       new  4bb175d   Do not build documentation in -a mode, to prevent timeouts on buildd servers
       new  5a1e1f4   Use JOBLIB_MULTIPROCESSING=0 to suppress use of multiprocessing by joblib while running tests to prevent failures in chroots without /dev/shm mounted
       new  7f4a004   Merge branch 'batchKMeans' of https://github.com/NelleV/scikit-learn into NelleV-batchKMeans
       new  3568d20   Handle integer type case for the estimation of covariances.
       new  8fdb47c   Replacing DP by dirichlet process
       new  e99c5a0   Don't use np.linalg
       new  6302e0a   Explaining what is dpgmm
       new  f07d5de   Adding see also sections to the mixture models
       new  9831402   Fix the 'give' in plot-dpmm
       new  f2fc81e   Editing a single example for the GMM and DPGMM explaining the difference
       new  7dcaced   [BatchKMeans] Small fixes following Olivier & Gael's review
       new  6eaecac   Merge remote branch 'scikit/master' into batchKMeans
       new  d2702e1   [MiniBatchKMeans] Removed the unnecessary import in examples/cluster/mini_batch_kmeans.py
       new  c90a79a   ignore 'cython -a' HTML reports
       new  ab11115   Merge branch 'batchKMeans' of https://github.com/NelleV/scikit-learn into NelleV-batchKMeans
       new  4de0dbf   ENH: style, pep8, docstrings comments, variable names
       new  62c5fe6   ENH: more interesting batch size
       new  43e8379   ENH: more fixes for variable names
       new  6470049   ENH: fix example docstring
       new  ee125a6   Remove copy and paste errors from nearest neighbors example
       new  0466882   Making the documentation findable
       new  c81364b   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into l1_logreg_minC
       new  11a8110   got rid of np.where
       new  e31db06   reimplemented l1_min_C as a function
       new  21dceb2   removed old version of min_C
       new  e3040cd   DOC: more work on the performance chapter
       new  62b7faf   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2183c4e   cleanup tests
       new  5332582   some more cleanups
       new  f378c2a   bound on C can be calculated also with one class
       new  a8fe71b   cleaned up tests
       new  76f0441   fixes to docstring (as for Fabian comments)
       new  8b81dc9   l1_min_c import in svm/__init__.py
       new  ad7ae97   libsvm low-level API refactoring.
       new  65b8952   PEP inquisition.
       new  61a2523   [MiniBatchKMeans] Now checks the validity of the data only when initializing the centroids. When the data is empty, return immediately
       new  ec13325   Merge with Olivier's branch
       new  7183c30   [MiniBatchKMeans] Documentation fixes
       new  a6a2097   DOC: Make sure that mpl is not interactive
       new  cbb9f0d   DOC: Tweak DPGMM docs
       new  244e01a   Some fixes for web layout.
       new  724a426   Remove obsolete information.
       new  0a898c3   More low-level refactoring.
       new  8e69f43   Editing the documentation substantially
       new  137f216   Adding doc to VBGMM
       new  8ed0046   Adding usage note to dp-derivation
       new  a1592a5   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  b87d8a5   Merge branch 'master' into alextp-variational-infinite-gmm
       new  414b802   Adding some test coverage. For some odd reason some tests fail on 'make test' but pass on 'nosetests scikits/learn/tests/test_mixture.py'. Any idea why?
       new  726a9aa   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into variational-infinite-gmm
       new  188490b   Fixing the docs
       new  36cf63c   Return first score in case of ties.
       new  69c05ea   rename grid_points_scores_ to grid_scores_ in GridSearchCV
       new  9ebbd25   Some tests for the things I changed in GridSearchCV.
       new  8e16dcd   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  420bf0a   Configure sphinx to be able to load extensions
       new  f876ad8   Fix OptionParser import
       new  0d349f1   Merge branch 'master' of https://github.com/ametaireau/scikit-learn-tutorial into ametaireau-master
       new  5e5fab5   Changing the image url in the doc
       new  c5bcc67   Even seeding the RNG in setup_func doesn't make the tests consistent
       new  778df05   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  42b40c6   There was a bug in the setup, now things are working deterministically
       new  c8c3e64   Deleting stray print statement
       new  58b78c9   Adding an rng parameter to the GMM classes
       new  a78bad6   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  36c7fa2   Fix little typos in the general_concepts document.
       new  4f84d43   merge with upstream
       new  01d35b6   Fixing the imports
       new  9295526   Inlining the helper norms
       new  17f8ab1   Beginning to vectorize the code
       new  8972598   more vectorizing
       new  153ed8d   Finish removing quadratic dependence on n_states; update docs
       new  2c8093d   Adding norm to scikits.learn.base, using that
       new  e83bd83   Putting norm in utils
       new  27206e0   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into l1_logreg_minC
       new  5c9bc07   Vectorizing parts of the VBGMM, which I had skipped due to it being a lot less useful than DPGMM
       new  2828888   Use np.cov instead of empirical_covariance in covariance module.
       new  7a93edf   Incorporating some caching and vectorizing to improve performance as per line profiles
       new  4a9040b   Fixing typo bug
       new  0b96751   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  710cfc9   Caching another computation
       new  783e347   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  b23a205   Small typo bug in _bound_z
       new  85b6204   a no-op that fixes tests
       new  897af6d   Change monitor to verbose, better output
       new  b862779   Merge branch 'variational-infinite-gmm' of https://github.com/alextp/scikit-learn into alextp-variational-infinite-gmm
       new  796e2c8   ENH: more informative test error message
       new  1fd29ab   typo
       new  a547a6a   Fixing typo-bug in the full covar update. There are still a couple of nondeterministic bugs to be taken care of
       new  e64326b   Making test_sample stop failing for no reason
       new  c00ed7b   Removing the square from norm() and creating helper sqnorm() in dpgmm
       new  5d9a225   Prevent setting the covariance parameters
       new  f43eda3   Caching the computation of the constant part on _bound_pxgivenz
       new  a8fcffb   Caching part of the bound for diag that was missing
       new  659a0a6   COSMIT: lognormalize->log_normalize
       new  75f7d31   COSMIT: avoid 'import as'
       new  4f67002   COSMITs
       new  2397ab9   COSMIT: avoid one-liner
       new  4a0e4ec   COSMIT: Local imports last
       new  2da48b4   Revert "COSMIT: avoid one-liner"
       new  af697ce   MISC: move mixture's test to sub directory
       new  affa6a9   COSMIT
       new  a484baa   COSMIT
       new  2d3569a   COSMIT
       new  d6337ff   ENH: Vectorize bound computing
       new  3c92923   ENH: Speed up _bound_z in DPGMM
       new  14bbe7e   moving some parameters from fit to __init__.
       new  1ee7f99   COSMIT: DPGMM: Move bound computing to functions
       new  39ad9b6   ENH: Speed improvements in DPGMM
       new  f66733b   DOC: Improve the GMM vs DPGMM example
       new  289a9e9   BUG: Fix bug introduced by moving test_mixture
       new  f2a3432   DOC: Fix layout
       new  a9f55f9   Merge branch 'variational-infinite-gmm' of https://github.com/GaelVaroquaux/scikit-learn into gael-variational
       new  703599f   DOC: Tweaks in mixture
       new  66ce055   BUG: fix testing (heisen) bugs in hmm
       new  9cce6f3   Fixing the names in the hmm test
       new  b255741   Merging gael's branch
       new  026cd4f   TEST: Heisen-bug fixing
       new  7b9311e   Merge branch 'variational-infinite-gmm' of https://github.com/GaelVaroquaux/scikit-learn into variational-infinite-gmm
       new  65c2077   Renaming bound_pxgivenz
       new  84d4e49   Renaming covar to prec
       new  8c04c47   Finishing the renamings
       new  70acebe   Adding a squiggly curve example for the mixture models
       new  028d121   Improving the coverage of dpgmm
       new  09ab842   Testing lognormalize
       new  6d21252   Splitting test_mixture
       new  525dc1f   ENH: spectral clustering doc and style improvements (pep8, docstrings, references, variable names)
       new  92fe82e   cosmit
       new  274dcc2   cosmit
       new  5709c56   ENH: style / pep8 / docstring fixes in s/l/utils/fixes.py
       new  7bae140   ENH: new make_rng utility function to help make PRNG seeding explicit
       new  f6089f1   TEST: forgot to checkin the unittest for the make_rng function
       new  cd5acaf   ENH: add test for picklability of the spectral clustering model
       new  a426dbf   Merge branch 'master' into l1_logreg_minC
       new  c6189e8   DOC: added reference to l1_min_c
       new  e44915f   the l1 logreg example now works with l1_min_c
       new  ed33845   coverage 100% + pep8 small fix
       new  e538b0d   Merged pull request #135 from paolo-losi/l1_logreg_minC.
       new  bf61233   DOC: fix links to l1_min_c
       new  3600482   Reintroduce empirical_covariance function + docstrings + cosmit.
       new  e346bc3   FIX: reference to l1_min_c
       new  762d270   Merge branch 'covariance' of git://github.com/VirgileFritsch/scikit-learn
       new  f3f7b47   Cosmetic changes in covariance.
       new  a048079   DOC: add low-level methods from libsvm.
       new  7b70505   FIX: fix rename of grid_scores_
       new  44f3d70   Do not open file write file until download is complete.
       new  c674b6c   FIX: make normalizer use the real l1 norm on each row (without assuming positive values)
       new  39149ea   Add tests for libsvm.cross_validation.
       new  a7cc486   There is no scikits.learn.feature_extraction.text.sparse
       new  a34ba1a   Fix minor typo
       new  fa9ccf3   Spelling error
       new  f7afbcd   Merged pull request #4 from larsmans/master.
       new  4e103c8   Add a small part about cross-validation + copyedit
       new  e59e8b9   Copyedit on the "working with text" chapter
       new  7826fff   Merged pull request #5 from larsmans/master.
       new  c88b5bc   ENH: Update joblib
       new  adcbdfd   Merged pull request #140 from fabianp/lfw.
       new  6c3402c   Copied the Sparse PCA file from the gist
       new  425436f   updates
       new  f890a41   [MiniBatchKMeans] Added a benchmark
       new  266a6e4   DOC: typo in line-prof package name
       new  c5d20ba   FIX: broken import in bench_plot_nmf
       new  b43b502   DOC: fix doctests to make them work with numpy 1.5 and olderw
       new  803dc33   Make ball tree code safer and 64-bit clean
       new  1409e01   Cleanup lib{linear,svm} C helper routines
       new  af6ab92   Add optional parameter n_class to load_digits.
       new  ef42c40   merged master
       new  c86a507   MISC: Prettify GMM example
       new  6deb88e   COSMIT: Pep 8 and remove useless imports
       new  2f40141   Merge pull request #144 from larsmans/balltree-cleanup.
       new  5df0a96   DOC: trim_doctests_flags = True for sphinx
       new  44915af   Spellcheck and formatting in developers' docs
       new  4496ed8   Merge pull request #147 from larsmans/master.
       new  8e7ac15   Merge (theirs) branch '0.8.X' into releases
       new  658c6bb   Merge branch 'releases' (from 0.8.X) into dfsg
       new  69e764f   Merge branch 'dfsg' into debian
       new  cd05e74   just a placeholder changelog for upcoming release
       new  cc500d5   ENH: avoid useless computation and warnings
       new  ec41def   FIX: missing import in plot_covariance_estimation.py
       new  80b2b86   Py3K: use explicit floor division
       new  97bc011   Boost policy compliance to 3.9.2 -- no changes
       new  e404c35   RF: use joblib.logger submodule itself while accessing its function in grid_search
       new  5990619   RF: use joblib.logger submodule itself while accessing its function in grid_search
       new  f624f9e   Merge pull request #148 from yarikoptic/master.
       new  6ee9d68   rename rng to random_state
       new  baa5e8f   cosmit
       new  ed40b2d   Preventing underflow in wishart_logz
       new  8cdd02c   Fixing 0* problem in z log z
       new  455057b   delayed check_random_state in k means and spectral clustering
       new  c2cb46c   Fixing another underflow bug in digamma. Now the bound for spherical covariance never diverges as a cluster gets empty
       new  904b4c2   Also, no warnings when running these tests
       new  f1c6265   Return also t from swiss_roll generator (needed to plot colors)
       new  5b19ddb   CSS style tweaks.
       new  9aa0e2a   more CSS tweaks.
       new  733cbac   Some more CSS tweaks
       new  642ddc1   Initial implementation of Locally Linear Embedding.
       new  2717b5b   check input arguments to BallTree, and be more careful in dealloc'ing
       new  2bab0b4   Re-generate .cpp from ball_tree.pyx
       new  677d88b   pep8 clean.
       new  3fd831b   typo
       new  27dfe95   Merge pull request #154 from larsmans/master.
       new  2df3f2a   SGD module supports two additional learning rates: constant and inverse scaling.
       new  326a81a   Added SGD regression benchmark
       new  db8fb49   Fixed doc-tests and added doc strings for SGD learning rates.
       new  66ef1bb   added notes on learning rate schedules to sgd.rst.
       new  89b396f   added learning rate arguments to docstrings.
       new  10febc4   Run cython on sgd_fast and sgd_fast_sparse.
       new  1f0424b   pep8 compliance
       new  422d162   cosmit: removed trailing whitespace
       new  8b5a027   kill trailing spaces
       new  455ad64   ROC fixes for trivial classifiers (always predict one class) and input checks (raise ValueError in case of multi-class).
       new  632b97b   [MiniBatchKMeans] Added chart showing the speed and the inertia / total number of points depending on the chunk size and number of iteration
       new  8ee00c6   added doctests for roc and refined documentation.
       new  52c4cef   merge with master
       new  f1928f7   [MiniBatchKMeans] PEP8 Compliance
       new  d85aba6   cosmit: pep8
       new  569e804   cosmit: beautified plotting.
       new  623a4d7   [MiniBatchKMeans] Fixed typo in attribute: cluster_centers_
       new  fd8353c   [MiniBatchKMeans] Added some documentation and example
       new  b6ab08a   [MiniBatchKMeans] PEP8 compliance
       new  4f0158b   merge master
       new  20f1495   FIX: python2.5 SyntaxError
       new  ca558df   FIX: tuples have no .index in python2.5
       new  ae05286   FIX: more python2.5 SyntaxError
       new  f38fd2c   merge from master, update random_state API + pep8
       new  d8bc0ae   Fixing test failures resulting from the merge
       new  a90819e   FIX: explicit linking against std++ breaks under mingw32.
       new  72a7b09   FIX: fix import paths in doctests.
       new  94e2883   Merge pull request #150 from pprett/learningrate
       new  2572441   docstrings: added note to roc_curve, changed y_scores to y_score.
       new  3148626   docstrings: changed signature of metrics.auc from fpr and tpr to x and y.
       new  dfefb01   track changes from master
       new  7613070   Merge branch 'rocfix'
       new  4237752   Fixing some under and overflows; this doesn't fix all test errors yet
       new  555ef24   Merge pull request #157 from fabianp/joblib_fix
       new  21206f4   Removing some more underflows, still not all
       new  0ea7d89   FIX: compatibility python2.5
       new  b4a6b57   DOC: add docstrings to BallTree.
       new  856e9a1   Update neighbors with latest changes to BallTree.
       new  cda0f87   Updated installation instructions
       new  5b26456   Merge pull request #160 from larsmans/master
       new  cbab7cc   Compressed README.rst to make it an executive summary
       new  55e3867   started work on homogeneity, completeness and V-measure as clustering metrics
       new  a371fd9   working implementation of V-measure, still needs doc and updated clustering examples
       new  5bb104f   [MiniBatchKMeans] Added a fit method to the MiniBatchKMeans
       new  92257fd   Merge branch 'master' into batchKMeans
       new  fd2fdc2   use V-measure metrics in K-means example
       new  fda0db3   add missing return info in swiss roll docstring
       new  9a5f5c4   Merge branch 'master' into batchKMeans
       new  1d66d0b   illustrate clustering metrics on affinity propagation example
       new  f4f09c8   100% test coverage for the new clustering metrics
       new  71ad011   more tests
       new  5b8aeb2   add more documentation for the new metrics
       new  4d0c331   typo
       new  cef6ccb   typo
       new  a56ca0b   [MiniBatchKMeans] PEP8 compliance and small fixed
       new  8f76efd   split some tests to make them more atomic
       new  a36f2fb   Merge branch 'master' into clustering-metrics
       new  c34cf5d   Trailing white space
       new  4ba9b8a   pep8
       new  f6b6e4c   DOC: Fix layout
       new  c243d21   Revert "Remove references to y in preprocessing objects."
       new  69f629e   Merge remote branch 'upstream/master' into revert_preprocessing
       new  7bd5b38   DOC: Fix layout
       new  096b966   DOC: Fix layout
       new  d1962c8   DOC: Fix layout
       new  79dce03   TEST: test for scaler in Pipeline
       new  689f268   Merge pull request #163 from paolo-losi/revert_preprocessing
       new  6a65388   Be more explicit about coverage testing
       new  348d9aa   DOC: improve datasets information
       new  1c7c0114  MISC: links to upstream datasets
       new  244abfb   [MiniBatchKMeans] Small fixes
       new  a3521a5   [MiniBatchKMeans] Added an example
       new  ceeaf73   Fixed Lasso call, all is still not right
       new  faae928   DOC: Documentation about covariance estimation.
       new  a9644cf   LARS _update_V fixed by Gael
       new  39c12d7   PEP-8
       new  7242ec2   Initial factoring into SparsePCA class
       new  6463492   Update .mailmap
       new  1f2a807   Layout fixes.
       new  1c5d8aa   Add analytics code to web page, SF discontinued web page stats.
       new  af5b05e   Implemented transform, fixed confusion
       new  d671948   Changelog
       new  2773f9a   Some doctest fixes.
       new  5900508   More docstring fixes.
       new  f21efa9   FIX: change doctest to avoid results with NaN
       new  4285ab7   I have no idea why, but this fixes the broken doctest.
       new  2ac5af8   Release 0.8
       new  da4f5de   Merge branch '0.8.X' of github.com:scikit-learn/scikit-learn into 0.8.X
       new  9ea3e10   Version number.
       new  5dee67b   Start of 0.9 development cycle
       new  41c406f   typos
       new  d90ba25   Merge branch 'master' into clustering-metrics
       new  7b661c4   typo
       new  198be6c   Welcome Lars & Edouard.
       new  fcc6988   FIX: pls docstring.
       new  b3c9687   FIX: pls docstring.
       new  f7c9f24   Remove failing docstring for release ...
       new  fa6fca8   FIX: for SGD log loss
       new  05afc5d   FIX: partial revert of the SGD log loss fix
       new  6bc68a2   DOC: added section on complexity for LLE.
       new  81f2fbe   Rename embdding_vectors_ --> embedding_
       new  39e1816   Make the author file more consistent.
       new  f6b4a84   Merge (theirs) commit '0.8' into releases
       new  d12807f   Merge branch 'releases' (0.8) into dfsg
       new  d67787a   Merge branch 'dfsg' (0.8.dfsg) into debian
       new  d321362   changelog for 0.8.0.dfsg-1
       new  2596c4d   changed semantics of LossFunction.dloss.
       new  595bf83   Merge remote branch 'upstream/master'
       new  f7d5317   FIX: reflect SVC API change (eps -> tol) in doc/tutorial.rst
       new  524a009   cosmit: pep8 + doc
       new  699fab0   cosmit: changed docstring of svm_gui.py.
       new  9de7b66   Merge branch '0.8.X' into debian
       new  965214e   cosmit: removed requirements in svm_gui doc.
       new  8b2aaf0   disable failing hmm test + changelog entry
       new  f03fee2   Merge remote branch 'upstream/master'
       new  3dd1d00   naive bayes name change MNNB->MultinomialNB
       new  3f4c3c6   FIX: reflect SVC API change (eps -> tol) in doc/tutorial.rst
       new  28c30c3   cosmetic change to ball tree C++ code
       new  47cbfbf   FIX: strings are not necessarily singletones + catch mistakes earlier
       new  bbbe225   DOC: minor spellings fixes in pls.py
       new  67ff4ef   Merge pull request #164 from yarikoptic/0.8.X
       new  4b8947e   Add submodule for manifold.
       new  e70a3dd   Cosmetic changes.
       new  10c6e14   Merge branch 'batchKMeans' of https://github.com/NelleV/scikit-learn into NelleV-batchKMeans
       new  1375582   Test KernelPCA support for n_components.
       new  cade238   Add support for n_components in KernelPCA.
       new  49ccebf   PEP8 fix.
       new  2db54dc   Merge pull request #167 from bsilverthorn/fix-kernelpca-ncomponents
       new  5ea0643   BUG: Make SVMs work on non contiguous arrays
       new  82aa58e   COSMIT: Better fix for continuity in SVMs.
       new  8eaf698   FIX: lars_path -- assure that at least some features get added if necessary
       new  9967d57   test case for previous commit
       new  8ddf771   minor -- pass verbose into LARS in the test case
       new  19f0295   fixed bug in BallTree cython wrapper
       new  d33ee3e   fixed small bug in cython wrapper for BallTree
       new  7eae6e0   Merge pull request #168 from yarikoptic/0.8.X
       new  6d2b4ce   FIX: lars_path -- assure that at least some features get added if necessary
       new  55b4584   test case for previous commit
       new  1ed238d   minor -- pass verbose into LARS in the test case
       new  ad2ee4d   MISC: Recythonize the ball_tree
       new  049a979   pep8
       new  c876e83   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  91751a6   more pep8
       new  8849716   better docstring for the LabelBinarizer in the multilabel case
       new  7af1c77   DOC: Fix links in covariance
       new  b4c2bdf   started work on normalizer API simplification
       new  1d0b8df   [MiniBatchKMeans] Updated the example to compare BatchKMeans and MiniBatchKMeans - added the copy_x option to the BatchKMeans
       new  29578c7   work in progress on package structure
       new  d89d3bc   FIX: rounding issues on python 2.6 in clustering metrics doctests
       new  95c3b4b   ENH: add a note on the symmetry of the metrics
       new  7cac98e   ENH: simpler import statement in example
       new  d3e5262   ENH: simpler import statement in example + explicit square
       new  6161e9a   ENH: add links to the reference guide
       new  323fb80   MISC: Remove unused imports
       new  da14f8e   ENH: Make LLE work with older PyAMG
       new  71819eb   MISC: Prettify the swiss roll example
       new  90b32d4   COSMIT: Remove unused import
       new  c3f33c3   ENH: better docstrings for symmetric considerations
       new  08524b0   cosmit
       new  5b593ca   ENH: better organization of metrics references
       new  8de5f59   ENH: reorganization of the document to be operational quicker
       new  119fded   fix broken test introduced in last checkin
       new  e38ee68   Add sparse.LogisticRegression to class reference.
       new  1dc6c49   new utility function to generate blobby datasets
       new  7f1b1b5   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7dfe065   FIX: indexing bug when labels are not consecutive
       new  34f060a   Merge pull request #3 from GaelVaroquaux/manifold
       new  24ca47e   Compatibility Ubuntu 11.04 (with matplotlib 0.99.3)
       new  ef7a349   More on practical tips.
       new  df15f40   Typo
       new  85fc047   FIX: bad import
       new  3c3db66   Move cache_size out of model parameters.
       new  c61f016   Cosmetic changes in the docs.
       new  839ff53   Docstring for test.
       new  5d4a30f   Test for non-contiguous input for svms
       new  673ca56   Implement predict_proba for sparse svms.
       new  6b074d5   cosmetic doc changes
       new  f10929e   pep8
       new  a79579e   Merge branch 'master' into clustering-metrics
       new  85edc23   FIX: broken doctests
       new  b9b58c0   FIX: doctests in svm doc
       new  93f9e96   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e3b3f28   ENH: new utility function to shuffle data in a consistent way
       new  3e94c85   Merge branch 'batchKMeans' of https://github.com/NelleV/scikit-learn into NelleV-batchKMeans
       new  a43363d   cosmetic: pep8 in utils/ + rewrote factorial (2x as fast)
       new  8618b88   DOC: clarified the default for NMF initialization
       new  e11c4ba   factorial should not use O(n) memory
       new  fa9f1a3   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  42a58c8   Python 3-safe attempted import of factorial and combinations
       new  d54d6ca   Merge pull request #161 from ogrisel/clustering-metrics
       new  3b46e42   Merge branch 'master' into sparsepca
       new  10051e1   Updated transform function, began tests
       new  76d93c6   ENH: small fixes in scikits.learn.utils.shuffle
       new  582fa82   [MiniBatchKMeans] Minor modifications on the examples
       new  cebecf8   [MiniBatchKMeans] Added labels and scaled the axis properly on the benchmark plot
       new  08340af   DOC: Prettify MiniBatchKmean example
       new  7a9eb28   BUG: Fix pyflakes warning in k_means
       new  cd99c13   BUG: params not applied in MiniBatchKMeans
       new  0cb216b   merge with master
       new  7e13274   ENH: MiniBatchKMeans: avoid useless computation
       new  f8525b8   BUG: MiniBatchKMeans: Error in stopping criteria
       new  96d47af   MISC: Cosmit in k_means_
       new  5b39710   Merge remote branch 'gael/batchKMeans' into batchKMeans
       new  cd9e727   DOC: cosmit in MiniBatchKMeans docs
       new  4169e28   TEST: Control seed in fastica tests
       new  7638a39   Merge branch 'batchKMeans' of https://github.com/NelleV/scikit-learn into NelleV-batchKMeans
       new  4b5fc0c   Welcome to Nelle\!
       new  f351280   pep8
       new  dbac5c1   ENH: syntactic sugar for the shuffle utility
       new  fd9919d   ENH: better / simpler handling of shuffling in MiniBatchKMeans
       new  2391d94   updated ball_tree documentation
       new  d22e905   Merge commit 'upstream/master'
       new  9bf47c2   ENH: Capture different data length in grid_search
       new  84961ce   added MLLE, made some small fixes to manifold module
       new  5570a48   wrapped brute force neighbor search
       new  b721758   added cython wrapper to BallTree.query_ball
       new  2ab954c   ENH : using make_blobs in plot_affinity_propagation
       new  3e0aeac   ENH : using make_blobs in plot_mean_shift
       new  b281899   ENH : using make_blobs in plot_mini_batch_kmeans
       new  6ec33ff   FIX : removing useless seed fix in plot_mean_shift
       new  54361ad   Text chapter: load_files renamed load_filenames
       new  7f6eb6e   Merge branch 'master' of github.com:scikit-learn/scikit-learn-tutorial
       new  7f42630   typos in README
       new  82d2560   query_ball -> query_radius, removed knn_brute
       new  2c938f4   DOC: minor typos in covariance doc.
       new  f64427b   Merge pull request #178 from kwgoodman/master
       new  087d6f2   BUG: Avoid NaNs in lars_path
       new  7bfc2a1   typos in covariance docs
       new  a038854   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  d0ae9f8   ENH: support instance of BallTree as input to kneighbors_graph.
       new  e219a7c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into manifold
       new  5631ff9   Implement transform method in LLE.
       new  4412c08   FIX: fix test.
       new  f97cefd   more fixes.
       new  bf23ccd   Merge branch 'master' of https://github.com/amitibo/scikit-learn into amitibo-naive-bayes
       new  e83a8cc   naive bayes: copyedit + rename alpha_i to alpha
       new  0b86ee2   ENH: optional and user-settable priors in multinom naive bayes
       new  2cc1837   naive bayes: minor fixes
       new  2a302fd   Merge sparse and vanilla naive Bayes
       new  80da99a   docs + cosmit in naive_bayes
       new  d967530   speed up BallTree.h
       new  843045d   slight speedups to BallTree.h and ball_tree.pyx
       new  4d70847   Fix spelling in dosctring.
       new  55d2772   Merge pull request #181 from lucaswiman/master
       new  43bcf61   ENH: refactored shuffle to address the resampling with replacement case + more tests
       new  7cbc276   FIX: n_samples bug in shuffle, 100% coverage in utils, missing reference doc entries
       new  1e418a9   first shot at a boostrapping cross validator
       new  afd8546   typos
       new  5592fcb   more typos
       new  03cb369   added unit test for BallTree.query_radius
       new  ed40151   fixed reference-passing bug in BallTree.h
       new  308eeb1   ENH: ensure that training and test split do not share any sample
       new  948c4ef   ENH: better input validation + more representative doctest
       new  280f2af   ooops
       new  54bb8a8   cosmit
       new  baf555b   DOC: cleanup in cross validation doc
       new  a395657   Merge branch 'master' into bootstrap
       new  549d18d   add bootstrap to reference doc
       new  47f4a12   DOC: new section for the Bootstrap cross-validation
       new  8173fed   cosmit
       new  d04dac4   cosmit
       new  0d59180   add see also in resample docstring
       new  4c34265   FIX: make cross_validation_score work with sparse inputs
       new  1d304ed   merge master
       new  1d8bcae   cleanup leftover
       new  41e1e92   ENH: add test for the permutation_test_score with sparse data
       new  54067fe   Merge branch 'master' into bootstrap
       new  e5b94d5   more tests
       new  a61093e   Merge branch 'balltree-wrapper' of https://github.com/jakevdp/scikit-learn into jakevdp-balltree-wrapper
       new  040938e   Merge branch 'bootstrap'
       new  bbbb022   naive bayes: handle 1-d input
       new  eee5910   ball tree cleanup & 64-bit safety
       new  48efee2   Modify the method computing errors on covariances (<cov_object>.error)
       new  e89a024   naive bayes: fix predict_proba bug and change priors behavior
       new  3e4a85c   Fixed issue 82: bug in init of Kmeans. I feel it is a little hacky  since I copy init at the very begining, not in "_init_centroids". But in this function, X_mean would not be available. And I like this solution better than adding another argument to _init_centroids.
       new  6a8dd42   Minor documentation: how passing a callable for init works.
       new  0530495   Changed default initialization method to "k-means++" for consistency with k_means function. Don't know if this is the desirable default.
       new  6f4021c   fix naive bayes docs and example + credit mblondel + vanity
       new  6722d79   typo: interation/iteration + re-Cythonize cd_fast.pyx
       new  cbbb0fc   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7d83a82   vastly improved MLLE speed
       new  5c42bde   added HLLE code
       new  0c55ac8   naive bayes: test pickling
       new  4e928ae   naive bayes: safe_sparse_dot, doc and docstring updates
       new  943439d   rename MultinomialNB params, rename GNB GaussianNB
       new  06cca82   Bug fix: turn <covariance_object>.mse into <covariance_object>.error
       new  151ebbc   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4c54424   FIX: make r2_score and explained_variance_score never return NaNs
       new  4163283   k-means clustering test: changed data points to be far away from zero. Now fails on initialization bug.
       new  84525f6   FIX: fetch_mldata fixed to support non-standard data sets in mldata.org
       new  946ff51   NEW: mldata tests
       new  e4dee4f   ERF: Simplify conversion of mldata.org data set name to filename
       new  bd8c8fe   reformulate MultinomialNB as linear classifier
       new  9ec0824   ENH: add pre_dispatch to GridSearchCV
       new  384fdb0   ENH: Catter for lists in grid_search
       new  97f86b5   sped up HLLE code
       new  a98db44   Merge pull request #185 from amueller/master
       new  9018064   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  5724245   pep8
       new  f4f55ac   add a comment explaining the + 10
       new  57cf25f   NB: add class_log_prior_ and feature_log_prob_ back as properties
       new  a72e89e   Merged Gael's gist newest update
       new  c0c1ee3   Merge branch 'master' into sparsepca
       new  996eff0   A couple of passing tests
       new  b94c8fc   NB cosmit: *feature* independence
       new  6731b64   cosmit: expand MultinomialNB docstring
       new  6dd017e   Safer importing in grid_search module
       new  a8bf9e5   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  a7c3e46   Merge branch 'mldata' of https://github.com/pberkes/scikit-learn into pberkes-mldata
       new  81bcf5a   pep8 / style
       new  4ec4a02   Merge pull request #1 from ogrisel/pberkes-mldata
       new  4854280   FIX: fix segfault in cases of infeasible nu (NuSVM)
       new  1d1b4dc   FIX: Remove column name when renaming in fetch_mldata
       new  c986097   ERF: Improved coverage of mldata, taking into account network availability
       new  b4fab73   DOC: documentation for fetch_mldata
       new  e04d4f8   Covariance errors computation API changes.
       new  33c4d15   ERF: Test mldata download using mock urllib2
       new  d7e64be   Merge pull request #184 from larsmans/amitibo-naive-bayes
       new  fa8ae78   rm references to naive_bayes.sparse in docs
       new  629f763   factored out the example code
       new  2dd421d   NB: rename use_prior to fit_prior
       new  27ee5c4   DOC: a little commenting
       new  b3804e9   BUG: Minor bugs in cross_val
       new  9703c23   FIX: fix pep8 and pyflakes issues
       new  be2d75a   fix broken test in MultinomialNB
       new  e0a1b36   ERF: refactor object mocking urllib2 for general use (to be used in doctests)
       new  55bcfed   ERF: Refactor utility function to test that list of names are (not) in an object
       new  1c7fd2f   ERF: Move testing utilities to make them accessible from doctests
       new  7602e4f   FIX: Doctests use mock mldata.org and do not download
       new  aecc0b4   slightly improved logging in a few easy cases
       new  b094000   FIX: transform method.
       new  6e46e2f   DOC: small fix in datasets.rst docs
       new  ee14654   bugfix: bad Scaler example.
       new  947b1b6   ENH: more readable datasets definitions
       new  325bcfd   ENH: avoid double HDD copy of mocked datasets + style
       new  dfd3a83   rm self.sparse attr in MultinomialNB; not needed outside of fit
       new  f1320bd   fix priors bug in MultinomialNB
       new  96b79a1   2010 is so last year
       new  a40922a   Merge pull request #153 from fabianp/manifold
       new  a97e299   FIX: use NeighborsClassifier in test.
       new  aa84b2d   Merge branch 'mldata' of https://github.com/pberkes/scikit-learn into pberkes-mldata
       new  d3bfaa5   Improved error handling + reduce memory use
       new  dd64030   Simplify intercept fitting in MultinomialNB
       new  df1fda0   Merge pull request #2 from larsmans/pberkes-mldata
       new  85fa92a   Error in MultinomialNB docs
       new  ab29811   merge
       new  3dcfe3c   merge master
       new  1861307   add random projection and PCA to digits manifold example
       new  8d37dcf   use scikit-learn QR compat alias
       new  a0d6fbb   cosmit
       new  29a0913   Merge pull request #3 from ogrisel/pberkes-mldata
       new  87ecbcc   FIX: some bugs in locally_linear_embedding.
       new  ca3e787   ENH: split figures for better reusability and readability
       new  144bde6   Merge branch 'extended-digits-manifold-example'
       new  b89174f   ENH: make the LLE random seeding controllable and deterministic by default
       new  b4c39cf   DOC: remove obsolete information in neighbors.rst
       new  5939ca5   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  a68070a   Add max_iter to LARS.
       new  e91a53e   docstring style
       new  01d10bf   FIX: broken doctests and missing max_iter attribute in LassoLARS
       new  2194925   fixed LARS doctest.
       new  9926845   FIX: broken doctest in the documentation caused by the last fix
       new  2b2f52c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  5f87e43   Added naive Bayes classifier for multivariate Bernoulli models
       new  3522d79   some documentation for BernoulliNB
       new  71cffc4   Merge branch 'master' into preprocessing-simplification
       new  ae418f3   work in progress on SampleNormalizer unification
       new  c151af5   enable test for the sparse variant
       new  f9156e4   getting rid of the remaining stuff in the preprocessing.sparse package
       new  858c43f   more explicit / descriptive low level cython function names
       new  3315aea   cosmits / pyflakes / pep8
       new  0c2ca24   ENH: improve docstring with missing parameters and motivations
       new  05a52b7   factorize a normalize utility function
       new  617d373   added ability to return distances and specify multiple search radii for BallTree.query_radius()
       new  f192cfe   fixed r shape bug
       new  d718744   Better doc for the dataset loaders.
       new  dc6b31c   Make kernels consistent with SVM and add sigmoid kernel.
       new  264b1f1   Fix LDA transform.
       new  d2a1339   Add LDA to the handwritten digit 2d-projection example.
       new  20e6298   Add TransformerMixin to LDA and RandomizedPCA.
       new  6444a43   Cosmetics.
       new  76b8a9e   Merge branch 'manifold' of git://github.com/fabianp/scikit-learn into manifold
       new  4504663   s/SampleNormalizer/Normalizer/g
       new  980fbf7   Merge branch 'master' into preprocessing-simplification
       new  b3e642e   moar tests
       new  b5e8b19   more tests for preprocessing (scaling)
       new  6086c9b   more tests for preprocessing: coverage is now 100%
       new  8020de6   make centering optional in Scaler / scale + fix broken test
       new  7d83db9   one more test
       new  3884786   one more test for preprocessing (no mean centering)
       new  d2d4336   fail early
       new  b3fea3b   pep8
       new  a1fda46   ENH: docstrings for Scaler / scale
       new  b619e5f   pep8 changes
       new  b8c1759   bugfix: sparse_format can be omitted
       new  e418d3c   typo
       new  dd6a4ff   better docstring for Scaler
       new  bc1e469   register the preprocessing utilities to the reference documentation
       new  36bb65b   fixes in See also sections
       new  7fd4084   ENH: give motivations for standardazation in the Scaler docstring
       new  603ecdf   ENH: style fixes and better use of the scikit-learn API in ROC example
       new  58966b7   Merge branch 'master' into preprocessing-simplification
       new  1299ffb   started work on the narrative documentation for the preprocessing package
       new  4279e16   typo
       new  04c9dfa   reorg TODO and notes
       new  4391c99   DOC: section on normalization
       new  77817a8   Do binarizing in BernoulliNB
       new  c26f65a   DOC: section on feature binarization
       new  5492e7c   transpose data on input and sources on output. This makes the interface compatible with the other decompositions
       new  42e6ae5   Adjusted examples to new ICA interface
       new  daf6453   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  ef38ea7   factorize the binarize function + write documentation
       new  91de2ad   format
       new  2d988c0   typo
       new  386e420   I don't really understand this, but it makes the error go away.
       new  51feb21   Added warning to fastica
       new  657b41d   Simplify binarizing in BernoulliNB
       new  e98bac8   pep8
       new  4762cfc   fixed bug
       new  6f85683   pep8
       new  b26393f   typo
       new  c5bb239   mention LDA in docstring
       new  74b762e   print docstring in examples
       new  c3b98c1   typo
       new  3f0f970   pep8 and starting with X in right shape
       new  705b35b   Merge pull request #200 from amueller/minor_docs
       new  c57b123   cosmetic changes
       new  b33ead7   Twenty newsgroups will not create folder if the folder doesn't exist and the files won't be downloaded anyway This happens if the folder doesn't exist and download_if_missing=False
       new  362bf3f   Merge pull request #194 from jakevdp/balltree-queryrad
       new  d7499b8   fix message in document classification example
       new  1ce86f3   Merge branch 'master' into bernoulli-naive-bayes
       new  7880af7   Optimize BernoulliNB + improve docstring + add to doc-class example
       new  dea2a3b   rst fix
       new  08f749e   DOC: fix errors in manifold doc + style tweaks.
       new  ea59ba4   Explicit cmap in swissroll example.
       new  559496b   Merge pull request #203 from amueller/docs_fix_again
       new  575c0fe   Merge pull request #193 from ogrisel/preprocessing-simplification
       new  cf8bf52   Notes on Fortran-ordering in fastica
       new  c9e9523   Better PCA docstrings.
       new  f6db31d   Fix LDA.transform's docstring.
       new  dd00031   Copyedit preprocessing docs
       new  d0750d0   Merge pull request #198 from amueller/fastICA_transposed
       new  fb5d723   Typo.
       new  d53e3f9   Add hinge_loss to metrics.
       new  50e9f18   Initial checkin of sparse.MiniBatchKmeans clustering + document clustering example on 20 news.
       new  1bf591c   enh: compute cache only on samples in current batch. now significantly faster.
       new  cdcc8ea   Refactor MultinomialNB: separate prior estimation and feature counting
       new  765ce7d   Added different compute_cache functions: dot and euclidean distance.
       new  0f7e074   Use unique from utils.fixes in naive_bayes
       new  6bb2b37   added SpectralClustering to document_clustering example. rm compute_cache from cython codebase.
       new  00d7de6   fix: random_state was set to None.
       new  82c4cc3   use provided x_squared_norms instead of recompute (if none euclidean_distances will recompute).
       new  c334708   reuse squared sample norms if possible (_calculate_labels_inertia).
       new  c9a7345   Python2.5 compatibility
       new  1af80ea   Merge pull request #207 from pprett/mbkm-fix
       new  f228fdb   Fix bug in MultinomialNB: output transposed
       new  11b0160   Replace loop in MultinomialNB._count with dot product + pep8
       new  f120d2e   BUG: explicit imports in doctests
       new  02e673b   BUG: binary classification failed in MultinomialNB, +regression test
       new  69f78db   Fix 404 from broken URL in release log
       new  1e2d35c   ENH: fit_transform on TfidfTransformer
       new  857930f   add C parameter to LinearSVC docstring
       new  46c3bc2   Revised text classification chapter
       new  dca2c71   Copyright and project name in HTML footer
       new  5e57728   Merge pull request #6 from larsmans/master
       new  7e71ad9   Fix pprett's website URL (<> caused it to be a relative URL)
       new  4a00ed5   add the test command
       new  02b697a   Use euclidean distance.
       new  2598d39   Merge remote branch 'upstream/master' into sparse-mbkm
       new  2393d67   dense and sparse seed differences: change order of shuffling X and init centroids to ensure consistant results.
       new  2f5248b   changed mini batch representation in dense MiniBatchKMeans - saves mem but increases runtime.
       new  f035f4e   Merge branch 'master' into bernoulli-naive-bayes
       new  ed1bfa4   deleted sparse clustering package. Moved _fast_kmeans.pyx to parent package and renamed to _k_means.pyx.
       new  04c121d   Merged dense and sparse MiniBatchKMeans implementations. MiniBatchKMeans now supports both dense and sparse inputs X.
       new  129fe59   Document clustering example updated.
       new  43c76c3   cosmit: pep8
       new  db3d273   fast function to compute l2 norm of rows in CSR matrix.
       new  8bb2fc2   set max_terms to 10k. Added spectral clustering.
       new  794605a   Refactor MultinomialNB and BernoulliNB: introduce BaseDiscreteNB
       new  f5c0cb5   Fast and memory-efficient loader for the svmlight format.
       new  9de3afe   Allow to user to fix n_features.
       new  333d5ea   Docstring.
       new  8e7369c   vectorize loop in BernoulliNB for 100x speedup in sparse case
       new  1b47641   Important note.
       new  059a129   Propagate errors up to the Python level.
       new  f98bb35   svmlight reader: don't use leading _ in identifiers
       new  cf623ab   FIX: update mldata tests to match recent updates; mock_urllib2 now accepts ordering parameter
       new  2b379fe   Narrative documentation.
       new  a770a59   Merge branch 'svmlight_format' of git://github.com/mblondel/scikit-learn into mblondel-svmlight
       new  aa3e4f9   SVMlight reader: minor fixes
       new  9b3f0f4   SVMlight reader: ensure C calling conventions + docstring
       new  3dc5d24   Update credits.
       new  6582a72   Return false when couldn't read the file.
       new  7eab00f   Fix comment.
       new  8526ab6   Plumb memory leak in SVMlight reader
       new  a1156da   SVMlight reader: one more clear() instead of delete
       new  d7debeb   SVMlight reader: cosmetic
       new  20c15ee   SVMlight reader: skip one level of indirection
       new  24619a6   Simplify and document SVMlight/libSVM data reader
       new  6566057   Use C++ exception handling in SVMlight reader.
       new  8d5e00d   Merge pull request #6 from larsmans/mblondel-svmlight
       new  fb3ecab   finish exception handling in SVMlight reader
       new  ea11518   Extend MultinomialNB tests to BernoulliNB
       new  5ab1fee   Update BernoulliNB docs
       new  e5f4b44   Updated README
       new  c08faad   BUG: broken doctest in BernoulliNB
       new  025ce37   renaming, included tests
       new  dee7edd   Merge branch 'master' of github.com:scikit-learn/scikit-learn into sparsepca
       new  8459e30   Updated init.py
       new  bd6ac1a   one more test and a quick example
       new  1a1510a   Merge branch 'mblondel-svmlight' of git://github.com/larsmans/scikit-learn into svmlight_format
       new  650c464   Fix compile issues on Mac OS X.
       new  446687a   Fix ref counting bug.
       new  d59a06b   More comments.
       new  c92d52c   Glitches in BernoulliNB and DiscreteNB (mostly docs)
       new  841a295   Merge pull request #210 from larsmans/bernoulli-naive-bayes
       new  2cede17   SVMlight reader: memory leak, type test
       new  772b24a   (Hopefully) full exception safety in SVMlight reader
       new  4f059a5   Merge pull request #7 from larsmans/mblondel-svmlight
       new  f70c442   load_svmlight_format -> load_svmlight_file.
       new  c7735e3   pep8
       new  5fd648c   DOC: foundations, prettified example
       new  236243c   added arpack support in scipy_future; wrapped MLLE and HLLE into locally_linear function
       new  26cf261   Merge branch 'master' into pberkes-mldata
       new  535e910   DOC: reorg of dataset page to make it more consistent
       new  a7c75a2   Merge branch 'master' into svmlight_format
       new  483b3ac   datasets/mldata.py is not a script, chmod 644
       new  20d7191   Merge pull request #209 from mblondel/svmlight_format
       new  08b7f87   Python 2.5 and SciPy 0.7 (tentative) compat in mldata
       new  b730d3a   Fix broken doctest in mldata
       new  7ec2f91   removed old files; moved example to examples directory
       new  adff6b1   Doc enhancement, added alpha in transform
       new  491129f   pep8 changes
       new  8bdf678   Merge branch 'master' into sparsepca
       new  159bcb1   FIX: make the dataset doctest fixture modular
       new  dbaea6f   typo
       new  1f99835   Added ridge in transform (factored here for now)
       new  c0187b9   Removed print statement from test. Whoopsie!
       new  ed1215a   prettify plot_sparse_pca.py
       new  d124bde   adding authors in sparse pca
       new  6b0b36d   Merge pull request #2 from agramfort/sparsepca
       new  668094e7  document placement new in SVMlight reader
       new  a455d4d   Documentation fixes.
       new  3642ab1   Initial integration of Orthogonal MP
       new  3e72041   added LTSA method
       new  0cb2e3c   Renaming, some transposing
       new  b215812   pep8
       new  12fcc75   Tests and the refactoring they induce
       new  7b4cdb0   PEP8
       new  09e17cf   Added signal recovery test
       new  69ccac9   rigurous pep8
       new  b3e6f81   Added the example
       new  0264056   Cosmetized the example
       new  e5b5c35   s/nonzero/non-zero
       new  25418ff   Remove ugly BLANKLINE stuff from text chapter
       new  f5c5e9a   Use grid_scores_ instead of _get_params + small fixes in text chapter
       new  4643f25   fixed bug in modified LLE: now works for higher dimensions
       new  a7df3c8   added method argument to digits example
       new  c6f2366   track changes from master
       new  e61a052   fit_transform does NOT return self + other docfixes
       new  d4982ef   ENH: Lasso and LassoCV: fit params -> class params
       new  f2f0cae   DOC: Tweak the cross-val lasso path example
       new  6f8bd76   ENH: LARS: fit_params -> class params
       new  83cf11c   COSMIT: Minor refactor in lars_path
       new  60b2597   ENH: Add a ShuffleSplit cross-validation iterator
       new  c312618   fix a typo in neighbors docs
       new  4d3926b   fix restructured text problems in the developers doc
       new  cb43d65   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  e5d19f8   BUG: Fix bug introduced in 83cf11c
       new  45f38b2   Merge pull request #216 from ametaireau/master
       new  cd694e7   Merge pull request #1 from ogrisel/jakevdp-manifold
       new  2c1c88b   BUG: change alpha scaling in LassoLARS
       new  e65e93e   Add test and cleanup for 2c1c88
       new  873256b   Added Olivier's patch extractor with enhancements
       new  6ee443a   cleanup
       new  06067f6   Tests for various cases
       new  fcb81ab   BUG: LassoLARS: X: modified during the normalization
       new  f1df50f   PEP8, renaming, removed image size from params
       new  4b172d1   Add note to base fit_transform doc.
       new  5bdc3ab   Parallel vectorizing is slower than serial
       new  e304597   Raise error if file doesn't exist.
       new  79ba998   Fix parsing issues.
       new  dafec63   BUG: LassoLARS didn't renormalize the coefs
       new  45f99c6   Test: test for unnormalized predictors.
       new  9bfd806   Add failing test.
       new  cfc0dee   DOC: add reference to FastICA from the ICA docs.
       new  5342cc9   Rewrote SVMlight parse_line with C++ iostreams
       new  32f8c47   More tests for the svmlight reader.
       new  064a800   SVMlight reader: some extra tests + cleanup
       new  5a905e5   DOC: add fit_intercept to LinearSVC docstring.
       new  03b5e9c   Merged Gael's latest update to sparse_pca.py
       new  9c4c63f   Merge branch 'sparsepca' of github.com:vene/scikit-learn into sparsepca
       new  898777a   Merge branch 'sparsepca' into sc
       new  1ff5944   Adapt kNN classifier to sparse input
       new  53cc842   Use new utils.atleast2d_or_csr in naive_bayes as well
       new  38235cf   FIX: update_V without warm restart
       new  1a745d3   FIX: weird branching accident
       new  1de74af   Merge branch 'sparsepca' into sc
       new  e1ef751   Revert "FIX: update_V without warm restart"
       new  a557817   Revert "FIX: update_V without warm restart"
       new  0e83b3e   Revert "Revert "FIX: update_V without warm restart""
       new  2cef229   Merge branch 'sparsepca' into sc
       new  00c976e   FIX: make the dataset doctest fixture modular
       new  2c425fa   typo
       new  7ee358f   document placement new in SVMlight reader
       new  7158ee3   Documentation fixes.
       new  23704df   Initial integration of Orthogonal MP
       new  170cc2f   Renaming, some transposing
       new  aa3bd39   Tests and the refactoring they induce
       new  487a8e1   PEP8
       new  6c74b15   Added signal recovery test
       new  9f5f0c3   rigurous pep8
       new  26fad8d   Added the example
       new  37dc0c5   Cosmetized the example
       new  b8ed9cd   Added Olivier's patch extractor with enhancements
       new  afd860a   cleanup
       new  dde98c0   Tests for various cases
       new  057b7f3   PEP8, renaming, removed image size from params
       new  77b6612   FIX: weird branching accident
       new  c02d81e   Revert "FIX: update_V without warm restart"
       new  7aa4926   Revert "Revert "FIX: update_V without warm restart""
       new  cfd5b34   Merge branch 'sc' of github.com:vene/scikit-learn into sc
       new  1598c40   FIX: update_V without warm restart
       new  34c6585   Added dictionary learning example
       new  43acd7f   ENH : prettify dict learn example on image patches
       new  ad5b4ca   pep8
       new  2a256ff   Merge pull request #3 from agramfort/sc
       new  0fe40e4   Document sparsity in k-NN
       new  019ac66   Correctly document sparse input possibilities in naive_bayes
       new  d8ede1e   Merge branch 'master' into sparse-knn
       new  849ff59   renaming for consistency, tests for PatchExtractor
       new  4e90475   Add sparse k-NN test, fix a bug
       new  6a44386   Extend sparse k-NN test to try pairs of sparse matrix types
       new  a3cdda7   Fix bug in sparse k-NN and add disabled (!) test for sparse regression
       new  bd26fd8   Better document scipy.sparse support in neighbors module
       new  7af69d7   Initial shape of dictionary learning object
       new  bf613b0   Added DictionaryLearning to __init__.py
       new  8550f10   FIX: silly bugs so that the example runs
       new  620b907   ENH: Tweaked the example a bit
       new  0c51fba   PEP8
       new  82a2d3a   Prevent some copying in neighbors + docstring for euclidean_distances
       new  64b04d1   ENH: Update joblib to 0.5.2
       new  5cc0473   Copied the Sparse PCA file from the gist
       new  233f050   Fixed Lasso call, all is still not right
       new  68d4baf   LARS _update_V fixed by Gael
       new  0882e1d   PEP-8
       new  eb769f1   Initial factoring into SparsePCA class
       new  054b2ee   Implemented transform, fixed confusion
       new  88aa176   Updated transform function, began tests
       new  7809996   Merged Gael's gist newest update
       new  ba03b11   A couple of passing tests
       new  0d1a6fd   factored out the example code
       new  ab44438   DOC: a little commenting
       new  36aa5b8   renaming, included tests
       new  d9ca705   Updated init.py
       new  246f95e   one more test and a quick example
       new  90f359b   pep8
       new  ecacbea   DOC: foundations, prettified example
       new  f3de9b3   Doc enhancement, added alpha in transform
       new  f9ed2e3   Added ridge in transform (factored here for now)
       new  213ac0c   Removed print statement from test. Whoopsie!
       new  72f0b7d   prettify plot_sparse_pca.py
       new  3ddf206   adding authors in sparse pca
       new  e437e98   s/nonzero/non-zero
       new  9d5ce8b   Merged Gael's latest update to sparse_pca.py
       new  e19a440   FIX: update_V without warm restart
       new  e9b8347   FIX: weird branching accident
       new  f5d757c   Revert "FIX: update_V without warm restart"
       new  5250036   Revert "Revert "FIX: update_V without warm restart""
       new  7b1e063   ENH: Update joblib
       new  cad248c   MISC: minor cleanups
       new  981478e   more tests for (mini-batch) k-means (99% coverage).
       new  bf41c13   Merge branch 'master' into sparse-mbkm
       new  2fd33fd   FIX : using product form utils.fixes for python 2.5
       new  badddc8   pep8
       new  a9ea545   MISC : fix docstring, cosmit in image.py
       new  98f592d   FIX; missing import in dict_learning.py (OMP in transform in not tested
       new  ead245d   Refactoring in ridge.py
       new  7538049   Rename of cg -> dense_cg and 'default'-> 'dense_cholesky'.
       new  5698a4c   Some docstring updates.
       new  4309c35   minor changes
       new  31a2a11   NeighborsClassifier: changed window_size to leaf_size & updated documentation as discussed in Issue #195
       new  1cd62ba   Use 10 neighbors in k-NN document classification
       new  c9d7a3f   neighbors: check string equality with ==, not is
       new  c7365e2   Copyedit SparsePCA docs
       new  24c3a68   Merge pull request #5 from agramfort/sc
       new  1a51660   Merge branch 'sparse_pca' of git://github.com/GaelVaroquaux/scikit-learn into sparsepca
       new  7f6b1ff   Finished merging Gael's pull request
       new  10cb314   Merge branch 'master' into sparsepca
       new  9847371   Merge branch 'master' into sc
       new  1b76e96   Merge branch 'sparsepca' into sc
       new  c01ea0e   Merge branch 'sc' of git://github.com/larsmans/scikit-learn into sc
       new  e0a9008   Copyedit SparsePCA docs
       new  b45a5ff   Renaming, part one
       new  fd98b57   Merge pull request #219 from larsmans/sparse-knn
       new  302d33f   Some doc copyediting
       new  7e83a53   Renaming, part two
       new  f68b50a   Renamed online dict_learning appropriately
       new  61ee071   Merge branch 'sparsepca' into sc
       new  b579582   Renaming part three
       new  d014aad   Fixed dico learning example
       new  993fbe7   Used @fabianp's ridge refactoring
       new  83771ab   Exposed ridge_regression in linear_model init.py
       new  4804f5e   Merge branch 'master' into sparsepca
       new  d11d4e2   Updated ridge import
       new  b99397a   Merge branch 'sparsepca' into sc
       new  df89fab   FIX: checks in orthogonal_mp
       new  5c3bafd   Cleanup orthogonal_mp docstrings
       new  bcde8c0   Change normalization behavior in TfidfTransformer
       new  c2a2fc2   Docfixes in feature_extraction.text
       new  4783ebe   fixed doc formatting
       new  c3591b3   merged with sparse classifier commit
       new  c131696   merged changes in master
       new  37b8109   pep8
       new  6af8179   merge with previous commits
       new  f75f47d   H_tol/M_tol -> hessian_tol/modified_tol
       new  1b427d9   Move scipy_future into utils.arpack
       new  d922db3   Add Jake to the mainfold credits.
       new  4ebd826   Remove bogus sparse vectorizing tests
       new  49a034d   docfixes in feature_extraction.text
       new  484d9bd   OMP docs, a little broken for now
       new  d673f82   document classification example doesn't demo only linear classifiers anymore
       new  c1e234e   DOC: omp documentation improved
       new  4391ac8   make parse_file in SVMlight reader static
       new  a48b00e   DOC: omp documentation fixes
       new  516812e   Merge pull request #222 from jakevdp/balltree-doc
       new  8fb39dd   DOC: dict_learning docs
       new  786ce12   dictionary learning tests
       new  598aee0   Fixed overcomplete case and updated dl example
       new  7b7f32f   fixed overcomplete case
       new  b5b833e   Explicit cmap for plot_compare_methods.
       new  468e4c2   Cosmetic cleanup.
       new  87364ee   Fix broken doctest in NeighborsRegressor
       new  efa5e37   PEP8
       new  5cdbb8f   FIX: strings are not necessarily singletones + catch mistakes earlier
       new  ff4eff0   DOC: minor spellings fixes in pls.py
       new  06429c5   FIX: bad logic in Pipeline.
       new  e3d00d2   Revert "FIX: bad logic in Pipeline."
       new  2f741ac   online dictionary learning object
       new  847d002   BUG: Make SVMs work on non contiguous arrays
       new  28dcc2c   COSMIT: Better fix for continuity in SVMs.
       new  e17ae1f   Test for non-contiguous input for svms
       new  f96263a   FIX: fix segfault in cases of infeasible nu (NuSVM)
       new  cf5c72e   0.8.1 release.
       new  48a3cd1   Better performance of Ax=b solver when b is 2d and A is sparse, and add new tol parameter.
       new  7a6b239   Fix doctest.
       new  ff671f5   factored base dico object
       new  038cd46   pep8
       new  e3ef711   Merge branch 'sparsepca' into sc
       new  16ac0ec   pep8
       new  db08066   more transform methods, split_sign
       new  7826f94   OMP dictionary must have normalized columns.
       new  853e469   Merge commit '0.8.1' (theirs) into releases
       new  0268272   Merge branch 'releases' into dfsg
       new  2e630ce   Merge branch 'dfsg' into debian
       new  3b48abd   changelog for 0.8.1-1
       new  2749234   ENH : new radius_neighbors_graph to build graph of nearest neighbor from radius
       new  16991d5   Merge branch 'master' into sparsepca
       new  8a553f5   Merge branch 'master' into sc
       new  dba0802   Search tfidf__norm space in text class. grid search example
       new  09d3335   Merge pull request #228 from larsmans/tfidf
       new  514c4f3   Use four categories instead of all in doc. class. example
       new  30683e6   ENH: Add small info on diabetes
       new  1872fb5   changed batch representation from indices to slices.
       new  d08d58b   DOC: improved dict learning docs
       new  f65dec3   Tweaked the dico example
       new  8a524b5   Refactoring in libsvm bindings.
       new  12d5159   FIX: fix bug in LLE with dense solver
       new  e8ed5e7   Update ARPACK from scipy.
       new  24ab71e   Backward compatibility fixes for testing LLE.
       new  d3b3a94   FIX: arpack doctest
       new  7809e44   comment LLE arpack test
       new  75773ff   exposed dict learning online in init
       new  b06f121   working on partial fit
       new  3f4d0cc   Optimize CountVectorizer.fit_transform (+ minor refactoring)
       new  cdf7df9   pep8 feature_extraction.text + rm content word "computer" from stop list
       new  37a6b97   Protect against MemoryError in libsvm.fit
       new  3096b68   DOC: Expand and copyedit naive Bayes docs
       new  d6a5d7d   Recythonize libsvm.pyx with Cython 0.14
       new  9704604   Refactor/simplify CountVectorizer
       new  bad0431   denoising example
       new  f243393   Refactor feature_extraction.text (again) to use Counter
       new  5ae3c91   Annotate the example
       new  7b42c64   partial fit iteration tracking, test still fails
       new  3bdc425   FIX: typo, s/treshold/threshold
       new  7875faa   FIX: doctest Ridge.
       new  baf2977   Replace mixture.logsum with numpy.logaddexp
       new  bbce0fd   FIX: add newline after autosummary:: sphinx directive.
       new  a8c3e61   Layout & consistency fixes linear models documentation.
       new  37d4939   cosmetic linear_model.rst
       new  fcb45b0   FIX doc linear_model.rst
       new  f7b05a9   test for vectorizer_inverse_transform
       new  0694387   Layout tweaks.
       new  32619c9   on demand inverse vocabulary
       new  02c78d5   Implement fit_transform for Vectorizer as well and document it
       new  59253f5   Default argument safety + cosmit in feature_extraction.text
       new  f15de34   DOC: adding radius_neighbors_graph to doc
       new  59e2955   pep8
       new  d1262f8   typo
       new  ce71714   DOC fixes in datasets
       new  958b7df   Reverse coef_ in Ridge.
       new  05f54c2   DOC: new example for Ridge + more rst docs
       new  7d1200d   Changed the default return type of ward_tree from bool to int
       new  83a266c   adding a comment on the test for grid_to_graph
       new  3cad001   Merge branch 'sparsepca' into mblondel-fix_ridge
       new  7a87962   simplify sparse pca
       new  1fd8277   Tweak denoise example spacing
       new  ef41e46   pep8 examples
       new  b584bab   pep8
       new  fcae8f0   Merge pull request #235 from mblondel/fix_ridge
       new  6fa3a6b   Merge branch 'master' into sparsepca
       new  aad908d   Merge branch 'mblondel-fix_ridge' into sparsepca
       new  c154810   Merge branch 'sparsepca' into sc
       new  791d4b5   Merge pull request #234 from larsmans/inverse-vectorizer
       new  5bcbb3c   FIX hmm.py to succeed tests; stopgap, put old logsum.py in that module
       new  e5ca264   FIX and ENH feature_extraction.text.CountVectorizer
       new  9bc1a90   random state control, comment fixes
       new  80b9854   Merge branch 'sparsepca' into sc
       new  8065cf7   random state control
       new  34b3276   clarify lasso method param
       new  e95d364   Merge branch 'sparsepca' into sc
       new  78cf3bf   clarify lasso method param in sc too
       new  4f05f93   default arg safety + docfixes
       new  2982590   Improve Logistic Regression sparsity example.
       new  5fe3ea9   Better test and remove old garbage.
       new  61cb912   s/seed/random_state in patch extractor
       new  cb9dbd0   DOC: fixed patch extraction comments
       new  767de18   ENH: PatchExtractor transform
       new  346330b   d:/progs/Git/s/seed/random_state in dico learning example
       new  a67b287   d:/progs/Git/s/seed/random_state in denoising example
       new  9992f70   Started one-hot transformer
       new  259bd0e   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  7fab6da   FIX broken test for CountVectorizer
       new  94eb7c0   pep8 and using np.bool instead of bool
       new  7d2ab8f   Revert "Started one-hot transformer"
       new  6c01e89   Merge pull request #236 from JeanKossaifi/sparse_matrix_type
       new  c52440a   Don't use np.atleast_2d when interfacing with native code.
       new  9685b97   Allow CountVectorizer to be fitted twice.
       new  415be5d   Remove unnecessary submethod.
       new  145fb81   2011!
       new  b730d3d   FIX: s/V_views/code_views and pickling
       new  7ec38ba   Merge branch 'sparsepca' into sc
       new  09f5b28   DBSCAN clustering algorithm. A density based cluster analysis algorithm that looks for core points in dense neighbourhoods. Comes with a test function. References an example that is not yet created (I'm working on that now). The structure of the DBSCAN code and the test are from the AffinityPropogation algorithm, and takes a Similarity matrix as input the same way (different other parameters though!).
       new  26a441f   DBSCAN density based clustering algorithm (Ester et al. 1996) Comes with test function
       new  273c214   Some documentation for hmm module, and a warning.
       new  38e1071   Merge pull request #230 from agramfort/radius_neighbors_graph
       new  26795bd   Example file based on Affinity Propogation example. Plots core points larger than noise and non-core points.
       new  a11eae6   Fixed noted issues with previous version
       new  a3fde56   params in DBSCAN.fit description
       new  dfaa407   DOC grid_search + pep8
       new  19ccdb3   DOC: more sparse pca narrative documentation
       new  0a5dfb1   FIX: gram when method=cd
       new  470c94b   DBSCAN now takes either a similarity matrix, OR a feature matrix. If a feature matrix, distance is calculated using scipy.spatial.distance.pdist Test added to test this feature.
       new  e433952   ENH: make rng of the LLE tests controllable to hunt down potential NaNs
       new  8838d03   Merge branch 'master' into sparsepca
       new  fe3c8d9   removed fit_transform overload
       new  99a64df   Merge branch 'sparsepca' into sc
       new  f38c5ab   Refactor naive_bayes and don't treat BernoulliNB as linear model
       new  bd72672   ENH show top 10 terms per category in document classifier example
       new  6a5bede   squared loss -> squared hinge loss.
       new  09bb0ce   COSMIT: Simplify the lena ward example
       new  7957572   Cosmit
       new  0219e66   Rephrase motivation for Sparse PCA
       new  a9bd3eb   Misc rephrasings of sparse PCA docs.
       new  a0aa0b2   Remove 'structured sparsity not implemented' comment
       new  2ff6954   Prefix explanation of sparse PCA formulation with 'Note that'
       new  68634c2   atoms -> components for clarity
       new  2ec3f20   Trailing whitespace fix.
       new  d46b820   Rewording in docstring
       new  4d874b9   gradient descent -> coordinate descent in docstring
       new  5f4d6b7   'Returns' section of the _update_code docstring
       new  b521b21   Wrap np.seterr reset in a try..finally block
       new  3edfda7   FIX the IRC chan used is scikit-learn, and not learn
       new  ee56dcb   Merge pull request #247 from NelleV/FIX_doc
       new  cf86e28   some optimizations for GaussianProcess
       new  7a62ae2   pep8 improvements
       new  9e2792b   remove unnecessary parens
       new  1486d24   Cosmit
       new  eb92e52   ENH: l1_distance: gaussian_process -> metrics
       new  4ea4754   DOCFIX typos in svm module
       new  e9e4b09   Doc: fix minor error in docstring
       new  dba41e4   DOC: sparse_pca: put maths at the end
       new  0219296   DOC on sparse_pca
       new  69a5453   DOC: Add l1_distances to classes.rst
       new  aad6f3e   cosmetic changes to DBSCAN
       new  99a0e9b   vectorize loop in DBSCAN with np.where
       new  5ef258b   Cosmit DBSCAN test
       new  760c154   DOCFIX DBSCAN: we use arrays, not matrices
       new  bccb5bf   DOC: consistent punctuation, minor enh
       new  92e5966   DOC: missed a couple of dots
       new  689b339   ENH: verbose and title in sparse pca example
       new  bec7d0a   ImporError -> ImportError
       new  2885a39   TEST: faster tests, and more coverage
       new  ee429af   Merge pull request #248 from dwf/misc_fixes
       new  4867237   BUG: Fix gmm bug + test failures
       new  5a3fd7d   Streamline imports in lfw.py: don't try anything with PIL
       new  c51dfa1   DOC: fixed typo in sparse pca narratives
       new  3baf1f7   Merge branch 'dwf_sparse_pca' of git://github.com/GaelVaroquaux/scikit-learn into dwf_sparse_pca
       new  295fc55   TEST: fake parallelism
       new  da56e4e   FIX/ENH: numerical stability in GMM
       new  9300689   TEST: fake only on win32
       new  5a0c5c4   TEST: no meddling with joblib outside of win32
       new  0d9a3fc   Merge pull request #1 from larsmans/dbscan
       new  2ac8df5   Merge branch 'master' into sparsepca
       new  ee040ce   Lower tolerance in sparse pca example
       new  6947828   DOC: sparse pca transform rephrasing
       new  c2aac46   DOC: more sparse pca transform rephrasing
       new  f3f24cd   Restore conditional PIL import in datasets.lfw
       new  1fbe5d2   Comsit: PEP8
       new  d439d8d   Cosmit: remove useless comments
       new  30f2fab   One big decomposition example
       new  2fcbe5b   DOC: consistent coding method in docstrings
       new  1239216   Merge pull request #7 from GaelVaroquaux/dwf_sparse_pca
       new  1b850e1   TEST: more coverage
       new  3aae4a7   MISC: Restructure compound decomposition example
       new  d115860   TEST: SparsePCA: testing fit_transform useless
       new  c7f4b36   FIX: sparse pca ignored initialization
       new  7242b65   Merge pull request #8 from GaelVaroquaux/dwf_sparse_pca
       new  75b9895   Merge branch 'sparsepca' of github.com:vene/scikit-learn into sparsepca
       new  39cdda6   FIX: typo in example s/cluter/cluster
       new  485ba6a   pep8
       new  e1452d1   pep8 in example
       new  43ab08a   FIX: messed up images in narrative doc
       new  8f2dfea   FIX: example image order is consistent (for now)
       new  fefce5e   TEST: testing HMM more robust
       new  1e56a3c   ENH: predictable ordering in example, included kmeans
       new  a5fd9a4   Merge pull request #212 from vene/sparsepca
       new  4ca0cce   Merge remote-tracking branch 'master/master' into sparse-mbkm
       new  ab3d07f   ENH: mixture: better numerical stability
       new  14842fc   moved _gen_even_slices to utils/init add test codes for checking callable init
       new  40f9170   pep8
       new  b87d586   FIX : fix failing test in comparison between lassoCD and lars
       new  f690851   pyflakes warnings
       new  15ad168   pep8
       new  f497509   Revert "pyflakes warnings"
       new  61fef5d   Covariance with residual at the end for path is zero.
       new  ccacfb3   Cosmit: Fix (some) pyflakes warnings
       new  39ef1e6   Added loader code for (Roweis) Olivetti faces dataset.
       new  c9b5711   Added imports to __init__.py for Olivetti faces
       new  f4b74fe   Documentation for the Olivetti Faces dataset.
       new  d8da0ca   Remove 'load_' alias for 'fetch_'
       new  e0f4acf   Use prints for now instead of logging at Gael's request
       new  d3909e3   Add a shuffle keyword, default False
       new  a1b49ec   ENH: olivetti faces: control RNG in shuffle
       new  ab41d59   DOC: Add a descr to olivetti_faces
       new  854159a   DOC: fix some formating issues
       new  b7e01da   DOC: Fix layout
       new  85d0a81   DOC: fix layout
       new  21f3659   FIX: LARS doctest in linear_model.rst
       new  d6121a8   DOC: fix layout
       new  edce447   DOC: copyedit docstrings in pls.py + (almost) pep8-clean
       new  32738af   pep8 and docfixes in various modules
       new  a99a58b   kernel pca gets its own module
       new  c72b4f3   pep8 and docfixes for LLE
       new  fc5c6b9   suppress division by zero warnings from precision, recall, f1
       new  462e2ce   simplify np.seterr handling in sparse_pca
       new  19825df   label_num is now only calculated once. This corrects a previous patch, which I incorrected half finished a refactoring, breaking the code badly :(
       new  fe62153   added label propagation class
       new  404bc3b   Update rsync command
       new  804a280   Use isinstance instead of the ancient (Py2.1) types module in fastica
       new  babe4a5   Merge branch 'variational-infinite-gmm' of https://github.com/GaelVaroquaux/scikit-learn
       new  ec13d95   Replace logsum by np.logaddexpr in hmm, tweaked some tests.
       new  c54743f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  eec20ae   FIX: add tolerance for lack of numerical precision
       new  f9cc40a   Merge remote-tracking branch 'lemin/sparse-mbkm'
       new  8fdd827   remove leading _ in _gen_even_slices and duplicate implementation in sparse_pca
       new  3262925   FIX handles NaNs in LogisticRegression, and many more classes
       new  ec56ca1   Merge pull request #255 from vene/kernel-pca
       new  3e15bc6   remove assert_warns from test case (not supported by build bots numpy version).
       new  08712d1   assert_all_finite: pre-check if we're dealing with floats.
       new  083d58b   DOC: Missing dot in Pipeline class description
       new  69178ab   Merge pull request #260 from glouppe/master
       new  664f710   Rework assert_all_finite and related functions in utils
       new  58d5b7f   callable now actually allowed in fastica
       new  03f0f37   disallow sparse input in dense liblinear
       new  8398cdb   Enforce axis=1 in Normalizer.transform + doc fixes
       new  1355100   Merge pull request #218 from fabianp/fix_lars
       new  9eddd66   Merge pull request #261 from glouppe/master
       new  b37929f   Add forgotten 'install' for mixture.
       new  1ba34d5   BUG: Fix clone for ndarrays and sparse matrices
       new  aab8745   BUG: Fix clone for nadrrays and sparse matrices
       new  f43279b   Removing unused code
       new  7374407   Merge pull request #259 from larsmans/input-validation
       new  c69a904   switch map and sum commands to numpy
       new  1739738   dbscan_.py file reinstated after accidental deletion
       new  c77bf7b   Rename n_states --> n_components in mixture & hmm + cosmetic changes.
       new  2bd81d1   remove verbose output from GMMHMM test
       new  a7fc975   Merge branch 'master' into sc
       new  1bd217f   DOC: adding note on glmnet parameter correspondance in ElasticNet
       new  6103a5f   ENH: Avoid np.logaddexp.reduce
       new  49bd56b   Function to calculate similarity matrix given either a feature matrix or a similarity matrix
       new  42c1f65   Merge branch 'dbscan' of https://github.com/robertlayton/scikit-learn into dbscan
       new  6d32024   Handle metric="precomputed" in dbscan.
       new  5d0ffae   Fixed documentation, and the input matrix is now consistently called 'X'.
       new  645bfa0   NOW X is used consistently everywhere
       new  6f3a9ae   pep8'd and pyflakes'd
       new  d3502b0   ENH : adding LASSO model selection example based on BIC and AIC
       new  6cd32e0   BUG: s/empty/zeros in plot_lasso_bic_aic.py
       new  6a1e2b3   pep8
       new  9c16517   FIX : _to_graph failed if mask's data was not of type bool
       new  6980380   Test to check that the grid_to_graph function works with every type of mask (and not only bool)
       new  d867fd4   DOC: Fixed issue #110 DOC: Added default values FIX: Removed nu in SVR and epsilon in NuSVR
       new  c7f61bd   cosmit: docstring of MiniBatchKKeans
       new  3ffe86e   remove n_init arg from MiniBatchKMeans signature
       new  8010f4d   fix: doctest formatting.
       new  9855595   fix: remove n_init from mbkm tests.
       new  6429f03   DOC: more precisions in univariate_selection
       new  3fff3c1   Merge pull request #265 from JeanKossaifi/master
       new  c128ca5   Merge pull request #266 from glouppe/master
       new  52b6ec6   BUG: fix dotests
       new  c2d8a71   labels_ doc updated
       new  3041f29   Chmod 644 feature_extraction/image.py: not a script
       new  c87d03a   Added a paragraph in the documentation. o
       new  0b8e809   FIX more useful diagnostics in mlcomp_sparse_document_classification.py
       new  37cf782   DOC: Missing import in doctests
       new  713a004   Merge pull request #272 from glouppe/master
       new  ae43df2   Add χ² feature selection
       new  486044b   Demo chi2 feature selection on document classification
       new  cee1e1b   document Chi2 feature selection
       new  51625b9   ENH test and fix chi2 feature selection
       new  130b84d   Rename f_chi2 to chi2
       new  eebe575   DOC: stress that only chi2 works with sparse
       new  9fe5e60   avoid mutable default arguments
       new  92ed8f4   K-means with transform method. Branch updated to include recent sparse k-means implementations (old pull request was basd on an old branch)
       new  8460e0d   pep8 fix for k_means_.py
       new  5bc5c35   Fixed documentation in example
       new  c56e777   Examples for dbscan in documentation
       new  1fae2ea   fixed broken doctest in HMM
       new  6dfd938   DOC: fixed SparsePCA docstring issue
       new  02375b3   COSMIT: remove unused import
       new  eebc2ec   DOC: Improve the Bayesian regression docs
       new  98ede0d   Typo
       new  96bec4a   Sorry, other typo
       new  47f59c3   s/euclidian/euclidean/g
       new  28fe278   More mutable default args
       new  f801626   fixing up tests, adding "unlabeled_identifier"
       new  b95d719   Much better example with pyplot, thanks to suggestions by GaelVaroquaux. This example is based heavily on the nearest neighbour example.
       new  b7c8845   vq now the default in KMeans.transform
       new  a4b0c29   n_samples used instead of n_points in transform()
       new  2e6206d   ENH decorator to mark functions and classes as deprecated
       new  5ab0f1d   New-style deprecation of datasets.load_files
       new  29e828d   Initial commit
       new  addaea0   fixed bug in calculating tau
       new  a2316af   Brought in OMP from the larger branch
       new  f263155   added functions to classes.rst
       new  00af49f   Remove useless prints in example
       new  621aefc   fix: call set_params in mbkm fit.
       new  9352898   Merge remote branch 'upstream/master'
       new  cd198c1   American spelling
       new  d1c264c   Example now much more likely to return 3 clusters.
       new  8d0e957   added cythonized Floyd-Warshall algorithm
       new  5a12763   API : renaming LARS to Lars
       new  7fdb81c   COSMIT : used implicit continuation inside parenthesis instead of backslashes in BayesianRidge.
       new  3000c7c   Typo : fix the 0.5 coefficient
       new  1e1b7b0   Merge pull request #279 from JeanKossaifi/master
       new  6134371   MISC: s/larslasso_results/lars_lasso_results
       new  0038a8e   pep8
       new  f3c2c76   deprecated decorator won't work on __init__; skip it
       new  243d209   Merge branch 'master' into rename_lars
       new  b0eeaa2   basic features of multiclass labeling up
       new  8ac81f2   speed tweaks in Floyd-Warshall, and renamed graph_search->shortest_path
       new  7df22e0   make deprecated work on classes
       new  966fe5a   Merge branch 'master' of github.com:scikit-learn/scikit-learn into rename_lars
       new  9bd078c   typo
       new  329b54f   Merge branch 'master' of github.com:scikit-learn/scikit-learn into rename_lars
       new  64c3458   Merge branch 'master' into omp
       new  b616ae1   consistency with lasso: s/n_atoms/n_features
       new  26ec2b2   ENH: adding LARS and LassoLARS deprecated classes
       new  2a07ecf   speedup in Floyd-Warshall: unsigned ints to prevent negativity checks
       new  a2cf951   DOC: some fixes
       new  c04e6e0   failing test for expected behaviour
       new  0ca25dc   FIX: LARS and LassoLARS did not accept n_features
       new  bf0ff88   PEP8
       new  d4c3b04   FIX: doctests
       new  2df0620   Merge pull request #278 from agramfort/rename_lars
       new  56239ba   Added Dijkstra's algorithm with Fibonacci Heaps for significant speed gains in path searches
       new  068e415   BUG: `copy=None` in `Scaler.transform` instead of `copy=False`
       new  ef24235   calculate_similarity changed to calculate_distance, moved to metrics.pairwise.py
       new  e79dd9b   Import of calculate_distance in metrics.__init__.py. dbscan.py updated to use this reference.
       new  54cc23e   Merge branch 'master' of https://github.com/robertlayton/scikit-learn
       new  ace0c8c   Merge branch 'master' into lars_n_features
       new  1805d4f   Merge pull request #281 from glouppe/master
       new  9247cb5   bug fix: free allocated memory
       new  a258fb9   changed shortest_path() to accept a sparse distance matrix for more flexibility
       new  8b4fe76   cleanups & pep8
       new  d646ff2   add tests, doc update
       new  9a4a684   fixing the way labeling works
       new  9f93e7a   FIX: support numpy < 1.3
       new  ad01072   Merge pull request #280 from vene/lars_n_features
       new  d87ae1a   FIX: broken doctest in Lars
       new  fa7b914   Remove max_features keyword from lars_path.
       new  822959f   cleared n_features naming confusion
       new  47e8016   s/n_nonzero_features/n_nonzero_coefs
       new  50b0d82   Factored out sparse samples generator
       new  4ce86d6   pep8
       new  4c5622b   OrthogonalMatchingPursuit estimator
       new  fe481a5   pep8
       new  89a81ff   combined manifold examples
       new  436f0d3   checking in minor changes
       new  b53ba4b   manifold doc update
       new  5e852ea   Revert "combined manifold examples"
       new  f52c9e6   fixed bug in shortest path; consolodated isomap examples
       new  c695cc7   ex. change
       new  8716ae3   Merge branch 'manifold-test' into manifold-doc
       new  8980ee3   Use default value for n_nonzero_coefs
       new  f32ca3b   Merge branch 'master' into omp
       new  28a52f0   cosmit in example
       new  33c2693   unified notation
       new  801f1ad   made code consistent with docstring
       new  8d165fe   FIX: coordinate descent stopping rule
       new  32d78bd   cleaned up tests, added count_nonzero to fixes
       new  7435b1d   Added OMP bench
       new  24348af   Refactored decision trees and forests to support CART algorithm.
       new  3780503   Refactored decision trees and forests to support CART algorithm.
       new  15a8987   Tests updated to work with the new distance based method.
       new  4309819   Test using a callable function as the metric
       new  480a380   Multiple small changes
       new  639d5aa   pep8'd
       new  f7ee93f   kmeans example renamed
       new  6887195   Remove hardcoded n_jobs from examples.
       new  a550d6c   Digits example has plot.
       new  2243f92   Added documentation
       new  6a8ce6f   make number of classes explicit
       new  1893641   Merge remote-tracking branch 'sabba/master'
       new  3de1c0a   resolved conflicts
       new  2d28553   Merge remote-tracking branch 'noel/decisiontree' into treemerge
       new  254a172   cleaned up and documented Fibonacci code
       new  f009988   better cholesky management
       new  025e944   pep8
       new  91f6faf   arrayfuncs solve_triangular and EPIC creeping bugfix
       new  729b173   fixed check for None
       new  525ed6a   set random seed to hide odd random test failures
       new  f1139ac   fix more None checks
       new  c1bacae   more clarity
       new  bd1068f   Merge branch 'origin/master' into dbscan
       new  295aa02   Small changes, mostly to wording
       new  d0580fb   Reference to calculate_distances fixed
       new  0ee02b1   Returned line I removed for some reason
       new  43ec924   Deleted line I returned that I really didn't delete.
       new  2fc4c96   Revert "Remove hardcoded n_jobs from examples."
       new  ec7cba4   Don't use n_jobs=-1 in the examples.
       new  51fd7ce   Added early stopping as in reference implementation
       new  9a25d24   ENH: Add a subset="all" to 20news
       new  491c7bb   API: load_20newsgroups is depreciated
       new  01b2a3f   Cosmit
       new  5d1b1e9   API+ENH: load data by default in mlcomp and 20news
       new  e589617   ENH: compression in 20newsgroup caching
       new  39cd55e   DOC: leftover false info in docstrings
       new  796dcbc   DOC: load_filenames -> load_files
       new  8c7cc15   added documentation, reworking tests
       new  0b8b6e0   K-means documentation updated to include information based on this PR Voronoi diagram correlation and Llyod's algorithm referenced. Use of transform method also mentioned. Docs built and checked, and the new test works.
       new  a913997   Extra example removed
       new  44aa564   n_nonzero_coefs defaults to 10% if eps not passed
       new  f7ab53d   began rewriting the tests
       new  bc20c29   transposed generator, updated tests
       new  cdf827d   Modified learn.cluster.mean_shift_.py so that the mean_shift function uses a KDTree to efficiently calculate distance within some threshold. KDTree implementation is in C and is from scipy.spatial.  Tested only using the example located in examples/cluster/plot_mean_shift.py
       new  ed08592   fixed stupid mistake causing the sample generator to be inconsistent
       new  1436e1e   added tests; cleanup; pep8
       new  b5836a4   Added another variant of mean shift clustering (in scikits/learn/cluster/mean_shift_.py that seeds using a binning technique on a grid.
       new  5124486   added missing _set_params call in LineadModelCV
       new  38583ed   Added visualisation and corrected bugs in CART algo
       new  9af7010   unified and simplified path params creation in LinearModelCV
       new  1e75e9a   Merge branch 'enh/ensemble' of https://github.com/satra/scikit-learn
       new  977baf2   Merge https://github.com/scikit-learn/scikit-learn
       new  0f7ebe2   improved nosetests doctest time
       new  c70eb9c   fixed parameters passing of LinearModelCV.fit, with test
       new  567a539   thread safe tests for coordinate descent
       new  f5b1ef1   PEP8
       new  56b1df8   warn when omp stops early
       new  0fa2978   no need for min, it would break on the previous line
       new  47814ff   change matrix order, gram looks ok now
       new  96a5f7b   Modified learn.cluster.mean_shift_.py in the following ways: Replaced old seeding strategy with bucket strategy which should be scalable. Modified nearest neighbor lookup to make it more scalable by adding a maximum number of neighbors -- in most cases this will not make a difference in the results --- the impact of this change is tunable with the max_per_kernel parameter.  It is now possible to force all points to belong to a cluster (default) or only those points t [...]
       new  071d14b   fixing up tests
       new  0c6fba3   DOC: Link the Olivetti docs in the main docs
       new  b60f889   pyflakes/pep8 on coordinate descent
       new  28f4bd4   Merge pull request #289 from sabba/master
       new  7c43e7d   use np.asfortranarray
       new  5949e52   Merged decisiontree and tree_model into tree, random_forests to ensemble
       new  872cf16   tests robust to warnings
       new  11836e7   do not overwrite global warn filters in test
       new  5613a4c   DOC: more explicit docs on alpha/rho in elasticnet
       new  98cbd4c   20% speed improvement by moving _find_best_split to cython
       new  52df7db   Modified learn/cluster/mean_shift_.py in the following ways: Added more efficent and proper removal of duplicate clusters. Took seed detection out of mean_shift function and put it in its own function.  Default bucket size for seed detection is now the bandwidth.
       new  583dfcc   ENH optimize euclidean_distances for memory use
       new  260c438   use np.argmax instead of x.argmax()
       new  74792c8   document and test sparse matrix support in euclidean_distances
       new  110ae3c   while 1 instead of while True
       new  6af2d2a   removed occurances of tree_model
       new  fedc8b9   ENH optimize idf computation in TfidfTransformer using np.binsort
       new  7e2b155   use nrm2 from BLAS
       new  c431799   remove unused imports
       new  d548181   first stab at implementation via KernelPCA
       new  532d760   add arpack support to KernelPCA
       new  498ae50   small efficiency boost to KernelCenterer
       new  d37e1ce   ENH: cv objects created by a helper function
       new  956640c   COSMIT: fix doc indentation to PEP8
       new  286dd14   BUG+COSMIT: rewamp the lasso path examples
       new  f8331ca   Made following changes to cluster.mean_shift_.py: Added documentation for new functions. Made following changes to cluster.__init__.py: this module now imports the get_bucket_seeds function from mean_shift_.py
       new  b0bd263   ENH: more rng instance instead of singleton in tests
       new  8a51be6   FIX: potential division by zero when normalizing non-pruned CSR matrices
       new  7c7b0e8   ENH and DOC TfidfTransformer
       new  acfcd31   It's official: omp is faster than lars (w/o Gram)
       new  06cba3e   FIX add idf smoothing to Vectorizer as well, defaulting to True
       new  1798eea   Small fixes as per ogrisel's comments. Most notable is that DBSCAN() now has components_ attribute, which stores a copy of the actual core points (not just the indices)
       new  7217d30   Merge remote-tracking branch 'remotes/origin/master' into kmeans_transform2
       new  e4ccc3d   Small changes based on mblondel's comments. Nothing overly noticable
       new  3670ea3   Replace points with samples everywhere
       new  ce27885   Use euclidean_distances in kmeans.
       new  7921049   random_state used instead of giving index_order as argument
       new  4f76bb8   Description for components_ attribute. Renamed core_samples_ attribute to core_samples_indices_ to remove confusion
       new  032f76d   Complete rewriting of samples_generator.py
       new  fceb4d4   Split the transform method into a predict and a transform.
       new  11fd404   Cosmit: use dense_output=True.
       new  e23f8d4   Sparse matrix support in kernels.
       new  945efb2   API changes, part I
       new  0398e0a   API changes, part II: Return of the Estimator
       new  1d09255   Fixes for broken tests due to the API changes in samples_generator.py (1)
       new  c5618c4   FIX: precompute_gram=auto
       new  9b162a6   DOC: docstrings fixes
       new  6a95959   Merge remote-tracking branch 'upstream/master' into samples_generator
       new  8226bbd   PCA: fix issue #258.
       new  b711769   pep8
       new  e2844ce   don't use gram in example, useless slowdown
       new  c018c5f   FIX: benchmark was broken
       new  7bcb232   PCA: better doc string for 0 < n_components < 1 case.
       new  7f2d216   DOC: docstrings
       new  fb9d5da   np.random -> RandomState
       new  f43d93b   K_pred_cols -> K_pred_cols_
       new  706d70c   Convert to F-order as soon as possible
       new  ff5ebe8   F-order asap, don't assume any overwriting
       new  4063c97   that was unneeded
       new  32598a6   ENH: Add a LassoCV using LARS
       new  d13a025   COSMIT: Nobody expects the PEP8 inquisition
       new  c08b9ca   API: add import paths for LarsCV and LassoLarsCV
       new  a3c4b8b   MISC: Follow changes to alpha scaling
       new  1eb98c1   ENH: Add normalization of X to LarsCV
       new  0be715c   BUG: Propagate fix 086b58f5 to LassoLarsCV
       new  e098a60   DOC: LARS docstring
       new  3b3c543   BUG: Avoid div by 0 in lars_path_residues
       new  871491d   ENH: Expose eps in LARS
       new  df31ae0   DOC: Tweak the bayesian ridge docs
       new  8edb574   DOC+TEST: LarsCV
       new  8584f98   TEST: Improve test coverage of LarsCV
       new  6304b43   DOC: document eps in least_angle better
       new  6c6e2f9   pep8
       new  d914cac   MISC: LarsCV: preallocate mse_path
       new  124c1cc   ENH: use _check_cv in LassoLarsCV
       new  66736a8   DOC; fix documentation
       new  c8b83e7   MISC: mse_path in LassoLarsCV is now the mean
       new  6260613   DOC: add example comparing LassoCV and LassoLarsCV
       new  a155134   DOC: typos
       new  d4eadd7   clearer benchmark
       new  c496a6e   API: _check_cv -> check_cv
       new  5bca101   ENH : prettify OMP/LARS benchmark
       new  8761428   Merge remote branch 'jakevdp/kernelpca-arpack'
       new  3f8e098   TEST: Robustify LLE tests
       new  379d9da   PEP8 in LLE tests + better assertion failure messages
       new  77001ee   display the eigen solver name in case of LLE reconstruction test failure
       new  38a7183   Partial support for sparse matrices in kernel PCA.
       new  e315d73   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9c5a1bf   Remove unnecessary import.
       new  67e15b8   Merge pull request #11 from agramfort/omp
       new  1a9c78c   Merge remote-tracking branch 'upstream/master' into samples_generator
       new  fd44110   DOC: referenced OrthogonalMatchingPursuit in doc
       new  f514208   Merge branch 'omp' of github.com:vene/scikit-learn into omp
       new  53a9134   Fixes for broken tests due to the API changes in samples_generator.py (2)
       new  3fd8292   Fixes for broken benchmarks due to API changes in samples_generator.py
       new  c595707   Fixes for broken examples due to changes in samples_generator.py
       new  2cd0ccf   updated samples generator according to @glouppe's refactoring
       new  d8abe5a   `seed` renamed to `random_state` and default value set to None.
       new  50e448d   Merge branch 'master' into manifold-isomap
       new  b03c9f7   manifold/shortest_path -> utils/graph_shortest_path
       new  f0ff849   Implement Isomap + transform in terms of KernelPCA
       new  b1f1dcd   added a lot more to label propagation, explained algorithms and differences between the two models
       new  dc7bac9   Merge https://github.com/scikit-learn/scikit-learn
       new  c856519   Added the Boston House Prices dataset
       new  c968558   Fixed imports and run unit test
       new  067fccb   More specific exception in GaussianProcess + regression test
       new  ce167bc   typo s/dictionnary/dictionary
       new  4b88fb9   PEP8
       new  66b95d9   Merge pull request #6 from vene/boston
       new  27b3c32   add description to isomap transform
       new  1f8e603   Corrected import of the data: all 506 columns are now usable
       new  8085437   merge
       new  8b4c9bb   scikits.learn.cluster.mean_shift_.py modified in the following way: improved documentation
       new  24b35fd   Changed plot_mean_shift.py example to use larger data set to show how bandwidth estimation dominates the runtime.
       new  ea5a422   more documentation
       new  e664434   added beginning of examples
       new  5f75ef7   Changed scikits.learn.cluster.mean_shift.py: Updated reference for mean_shift algorithm
       new  1cb2156   Changed scikits.learn.cluster.mean_shift.py: Added Conrad Lee as author.
       new  a1d8259   Merge branch 'master' into omp
       new  c0acd5b   Merge branch 'boston' of https://github.com/bdholt1/scikit-learn
       new  993f52e   Changed scikits.learn.cluster.mean_shift: modified so that complies with pep8.
       new  8310b08   Changed scikits.learn.cluster.__init__.py and examples/cluster/plot_mean_shift.py: modified so that complies with pep8.
       new  d47493b   Changed scikits.learn.cluster.mean_shift_.py: Now uses BallTree because of built in query_radius function, allowing us to get rid of the get_points_within_range function. Changed MeanShift to not use bucket seeding by default.
       new  303a6d4   Hard coded bandwidth to 1.30 because otherwise its calculation is too slow.
       new  62495d5   Changed scikits.learn.cluster.mean_shift_.py: now uses blas nrm2 to compute norm.
       new  f51b652   Merge branch 'dbscan' of git://github.com/robertlayton/scikit-learn into dbscan
       new  d244c483  calculate_distances -> pairwise_distances + goodies.
       new  02f12b2   Improve DBSCAN doc.
       new  7ccf836   Fix DBSCAN example.
       new  ba62ae7   Remove automatically generated auto examples.
       new  f025353   Modified file scikits.learn.cluster.mean_shift_ Replaced a list comprehension and a for loop with numpy operations to improve efficiency.
       new  dfe7791   Modified file scikits.learn.cluster.mean_shift_: removed print lines used for debugging, made code compliant with pep8
       new  4b433a9   Modified file scikits.examples.plot_mean_shift.py: updated reference.
       new  fcf29c2   Test pickability in DBSCAN.
       new  45ff4b1   Test precomputed similarity in pairwise_distances.
       new  17ea86e   Mean shift: now uses norm function from utils.extmath
       new  61e6076   Merge pull request #277 from vene/omp
       new  be6c9cf   FIX: broken samples generator test
       new  88d2463   Added references to functions in the `datasets` module.
       new  36c1857   Merge remote-tracking branch 'upstream/master' into samples_generator
       new  fdfa4c2   Fixed a broken test.
       new  dfc57c2   (Micro)optimization in DBSCAN
       new  da684a2   fix DBSCAN bug (oops)
       new  68088f7   Added tests for the samples generator module.
       new  ae9bc2a   Added references to samples_generator.make_* functions in the documentation.
       new  8321991   ENH: make the file loader keep the filenames information
       new  b48fd86   Merge branch 'samples_generator' of git://github.com/glouppe/scikit-learn into samples_generator
       new  9636722   update doctest in tutorial to match current API
       new  51aa20c   Doc for sample generator cosmits.
       new  9f45c23   ENH: speed up estimate_bandwidth with BallTree + use make_blobs in test_mean_shift.py
       new  dafcb87   ENH : using make_blobs in cluster examples
       new  92cc1f0   pep8
       new  4ef30ea   FIX: cruel bug in OMP, no more unneeded warnings now.
       new  49e6f5e   added "structure" example
       new  bcf48f9   tweaked structure plot
       new  35b04a0   added Isomap.reconstruction_error()
       new  4705624   store BallTree in Isomap for faster transform()
       new  d814dbf   finalized SVM comparison example
       new  840ef6e   all tests pass
       new  d2cbaaa   Merge branch 'master' into sc
       new  0936c55   Updated documentation for boston house prices dataset
       new  755fac5   Added Olivier's patch extractor with enhancements
       new  96dbc10   Tests for various cases
       new  0c19c07   FIX: removed the required parameter K
       new  a428713   PEP8, renaming, removed image size from params
       new  ff88bb6   FIX : using product form utils.fixes for python 2.5
       new  329d1a0   pep8
       new  2fd16ad   MISC : fix docstring, cosmit in image.py
       new  7e5f107   s/seed/random_state in patch extractor
       new  b197771   DOC: fixed patch extraction comments
       new  a1fa061   ENH: PatchExtractor transform
       new  77c72a1   extra blank line
       new  e133798   cosmit on docstring first line
       new  93d1bbd   FIX: broken Gram handling in OMP estimator + minor style improvements
       new  485db45   Merge pull request #295 from bdholt1/boston
       new  9f63c7b   pep8 in test file
       new  f7e0508   image.py authors
       new  eae8df6   speed up tests
       new  b7d63eb   improved warning for invalid max_patches
       new  eda7293   New file: Feature extraction documentation
       new  7e9e023   FIX: dataset description
       new  7658bcf   fix conflicts with master
       new  cde824a   Merge branch 'manifold-isomap' into manifold-doc
       new  6c9f93b   Further optimisation of _find_best_split
       new  3be25ae   Further optimisation of _find_best_split
       new  f8b8aad   cosmit: docstring + raise ValueError if kmeans input is sparse.
       new  2b0e9c9   added document clustering example to KMeans clustering section.
       new  115983d   Refactoring: speedup of decision tree code
       new  b1abfdc   removed some stuff from documentation
       new  2c63ba0   updated pydoc to make behaviour clearer
       new  8c8f3e9   Added feature extraction as a chapter
       new  b8154ee   fix copy paste error in docstring
       new  a9eef34   passed PEP8, using already implemented kernel functions
       new  9abb9fa   update manifold documentation
       new  d52985e   making everything more numpy compatible
       new  f0cdac0   Merge branch 'master' into jakevdp-manifold-isomap
       new  919cc38   FIX: broken dataset generator import + minor styling issues
       new  1412fe5   fix comment
       new  bb5fe7e   Added normalize parameter to LinearModel
       new  4baeb83   Small improvements in the documentation of the toy datasets.
       new  aeee2bb   DOC: improved docstrings
       new  3b90f83   dictionnary -> dictionary
       new  abc35e2   Merge remote-tracking branch 'upstream/master'
       new  bc91b77   Further performance improvements.  Now approx 30 - 50% faster than MILK.
       new  539ba23   Merge pull request #303 from glouppe/master
       new  a6ee968   Added parameter normalize to LinearRegression
       new  c1222ac   Updated documentation, fixed bug in the process
       new  1e8e061   FIX: avoid the dependency on pylab in the doctests
       new  feb7786   DOC: clarified docstrings even more
       new  da19205   Merge branch 'boston'
       new  4c43f8e   LassoLARS now uses the normalize parameter
       new  ebe226d   Updated benchmarking for trees
       new  a525cc5   PEP8
       new  416a675   DOC: added documentation for graphviz method
       new  361cfef   FIX: corrected computation error and typed incoming arrays
       new  52a0158   Completed the integration of the parameter normalize
       new  f1a7454   Implementation of the parameter normalize in bayes.py
       new  1548568   fix the random state of the files dataset loader for reproducible results
       new  a3f7cc4   added parameter normalize to coordinate_descent
       new  a23365c   added parameter normalize to ridge.py
       new  61bdb4b   DOC : fix doc building
       new  d8822e5   Upgrade the setup documentation to match the current master
       new  d071096   updated README.md
       new  60a4bf3   fixed left-over of the previous version
       new  4121347   FIX: corrected graphviz visualisation.
       new  c5ce75a   Merge remote-tracking branch 'vene/patch-extraction' into vene-patch-extraction
       new  f506184   fix broken doctests
       new  4935588   Merge commit 'upstream/master' into manifold-doc
       new  d057963   more details for poor windowsians
       new  5d1429e   Merge branch 'master' into sc
       new  c60d88f   Improvements of the RFE module.
       new  5852e72   Merge remote-tracking branch 'upstream/master'
       new  4e1406a   Accidentally removed a line in a test
       new  0170946   Documentation + PEP8
       new  b4d6af0   More robust test on `step`.
       new  b281aa2   Fixed a syntax error
       new  d0ee04c   Small code simplification.
       new  d4b9698   upgrade to simpler new load_file API
       new  ad81cb8   s/class_names/target_names/g
       new  f6f7300   upgraded exercise 2 to the new API
       new  0331e98   updated faces recognition example to the latest API
       new  dc27354   upgrade inline doctests to new load_files API
       new  d2006f4   reorder exercises
       new  aa77f82   graph construction and example more numpy-like
       new  e53ecc7   fixed other diagonal matrix construction
       new  877068c   rename misnamed "plot" example
       new  12fc1dc   example conforms to pep8
       new  6d4231c   other example conforms to pep8
       new  ac54d65   made test conform to pep8
       new  88dee75   predict() method now numpy friendly (100% numpy friendly now)
       new  fef95fc   more numpy integration
       new  b3c8edb   batch k-means: calculate labels and intertia in chunks to prevent memory errors
       new  295dc3c   Added parameter normalize to omp.py Some changes in Linear_Model
       new  21ab4cb   BUG: Fix a bug introduced in rebasing
       new  971d89c   new-style deprecation of load_20newsgroups
       new  126e09f   pep8 in doc
       new  cf4659d   Merge pull request #305 from vincentschut/mini-batch-kmeans-batch-labeling
       new  0b4272f   fix: if n_samples < chunksize n_batches was 0 and no iterations are performed.
       new  4f93af4   cosmit: rm debug output
       new  6c479c0   BUG: normalize before center in lars_path_residue
       new  06013d5   add smoke test for MiniBatchKMeans
       new  e274935   Added parameter normalize Added parameter overwrite_X
       new  9270eae   Merge remote branch 'upstream/master'
       new  899dfcd   ENH : new LassoLarsIC estimator
       new  f4ddd47   MISC : adding GaelVaroquaux to the authors of least_angle.py
       new  e4f4b90   ENH: addressing @ogrisel's comments on PR 298
       new  f599fab   ENH + DOC: addressing @GaelVaroquaux's comments
       new  dd9bd4c   Fixed some errors (mainly docstrings)
       new  1f2ad60   Merge remote branch 'upstream/master' into normalize_data
       new  3a56be4   changes to manifold doc
       new  22da701   DOC: cosmetic changes to lars-bic doc and examples
       new  5fb850b   DOC: make lasso docs easier to read
       new  43e3229   DOC: clarify doc on BIC/AIC
       new  eb6b74b   Merge remote-tracking branch 'upstream/master'
       new  33190dc   Improved test coverage of rfe.py to 100%
       new  3493ac5   removed everything except the plain and simple decision tree to make reviewing easier
       new  2f9f318   Fixes of minor bugs + improved test coverage (now 100%)
       new  e0852d1   DOC: Updated the documentation to reflect decision trees.
       new  d7f48eb   Corrected newlines, and ensured only tree related changes are in this set
       new  64f785f   Addressed Gael's comments.
       new  4a76f3b   Addresses Gael's comments. (2)
       new  182c365   Addresses Gael's comments. (3)
       new  91226a9   Typo.
       new  f49eec5   FIX: replaced ad-hoc RNG with suggested scikits.learn implementation.  Tidied up dependent examples.
       new  8ec8583   rename coding_method, transform_method to fit/transform_algorithm
       new  b8c330d   fix broken test
       new  31bf0b3   Merge branch 'kmeans_transform2' of https://github.com/robertlayton/scikit-learn into kmeans_transform2
       new  39552b4   Add tests and fix bug.
       new  20d7c3e   Improved test coverage of samples_generator and feature_extraction modules.
       new  345eeaa   Fixed a small introduced due to a previous commit.
       new  d7d36b3   Kmeans transform and predict doc improvements.
       new  5b6b343   Merge with master
       new  0690a0a   Merge https://github.com/scikit-learn/scikit-learn into enh/tree
       new  791b1b2   Merge pull request #296 from bdholt1/fix/feature_extraction
       new  9a88893   Merge remote-tracking branch 'upstream/master' into test-coverage
       new  882fa66   removed function kernel, switched to string for picklability
       new  803e8db   fixed a bug in the circle example
       new  8353bc1   moved label propagation examples to lower subfolder
       new  97b537c   more numpy friendliness
       new  207d25a   speed improvements on LLE variants for high dimensional data
       new  ba5bf00   manifold example updates
       new  3d0da94   typo in HLLE
       new  d7ed1cb   examples: make out_dim explicit
       new  af9eef3   COSMIT: remove unused import
       new  24fde7b   BUG: make lobpcg work with non-sparse matrices
       new  6b98994   COSMIT: tweak plot_compare_methods example layout
       new  d4d6dfa   remove lobpcg from LocallyLinearEmbedding
       new  1d5e5d5   COSMIT: print time in plot_lle_digits example
       new  23163f8   MISC: fix image in manifold doc
       new  44cb7e6   changed digits to faces decomposition example
       new  f37252d   added dict_learning_online function
       new  12386a3   MiniBatchSparsePCA is born
       new  d782152   Removed dict_init in MiniBatchSparsePCA, docstrings
       new  02d7eb4   code reuse by inheritance, more tests
       new  b79dbb0   Fast-running face decomposition example
       new  91eb864   DOC: updated narrative docs for MiniBatchSparsePCA and example
       new  1103a79   DOC: fixes and updates
       new  bb11d45   DOC: minor errors
       new  207a8fa   FIX: broken test
       new  c1c9142   Added MiniBatchSparsePCA and dict_learning_online to classes.rst
       new  10fabcd   DOC: fixed issue in MiniBatchSparsePCA docstring
       new  a6b21f9   ENH: remove references to digits + format
       new  26f7f1d   plot the original centered sample + make sparse pca a little less sparse + kmean a little less like init
       new  1eaff73   DOC: make the decomposition doc more consistent with running faces example
       new  05b3664   cosmit
       new  bfc41a5   ENH: use introspection to find the cluster components
       new  3ee301f   DOC: group SparsePCA and MiniBatchSparsePCA chapter to reduce redundancy
       new  e85c809   cosmit
       new  da7cbc6   ENH: minor style fixes in docstrings and comments
       new  a10b3ba   cosmit
       new  ff1bc4c   cosmit
       new  f999c24   FIX: removed recently introduced mistake from dict_learning_online docstring
       new  fe92da9   ENH: cleaner random number handling in tests
       new  7eb38a7   Removed default value of n_components=None in SparsePCA
       new  c7fd201   Fixed inappropriate checks for None
       new  3cca8c1   Switched dict_learning_online returns order for consistency
       new  ea8dd4a   ridge_alpha as instance parameter
       new  06bdd21   prettify face decomposition example (ft. GaelVaroquaux)
       new  82b423c   MISC: prettify the faces example
       new  872971e   add refs to example
       new  65a2d4a   COSMIT: doc and examples in decomposition
       new  24c0109   more numpy use,
       new  994241a   Improved documentation + predict/score.
       new  ab282ec   Cosmit
       new  756835f   Typo
       new  e65eee4   Typo (2)
       new  e7412c6   fine tuned some documentation
       new  3d39a32   Add TransformerMixin (back?) to preprocessing classes.
       new  547e65b   Fix plot_kmeans_digits.py.
       new  fa2e505   Carve the emmerging consensus on __init__ vs fit parameters in the contributors documentation
       new  a9eb804   cosmit
       new  ebf45f8   DOC: give some motivation for the return of self in fit
       new  c688e97   DOC: formatting mistake
       new  147e78d   DOC: more fitting doc improvements
       new  6fae2ef   typo
       new  48d1fa3   DOC: more formatting
       new  6c49694   yet another typo
       new  d68ebfd   Merge branch 'master' into sc
       new  ef7fb1c   duplicated import
       new  01e3b5a   FIX: denoise example was broken
       new  721892ed8 FIX: reconstruction test
       new  fd176ae   Added a function as_float_array in scikits.learn.utils
       new  0dd22e5   Fix : deleted a forgotten line
       new  4573adc   make tests share data
       new  8899160   clarify docstrings
       new  808a37a   added init test
       new  a50462e   partial_fit passes the test
       new  05a2672   Merge remote-tracking branch 'upstream/master'
       new  bb9f32a   Merge https://github.com/scikit-learn/scikit-learn
       new  95b17a4   Removed unused import
       new  ff14545   PROF: improved speedup thanks to ppret
       new  a2f5c44   Merge branch 'enh/tree'
       new  8065590   Merge pull request #311 from glouppe/test-coverage
       new  c72dd9f   merge with master; remove lobpcg references
       new  0c03b91   added a snazzy label propagation versus SVM decision boundary plot
       new  c3ab858   Merge branch 'master' into kmeans_transform2
       new  1505c71   added more explanation to the plot
       new  208a70e   added semi_supervised directory
       new  6a99439   removed old, useless code
       new  9d1da53   removed unused imports
       new  7a0382c   added more documentation, another doctest for LabelSpreading
       new  8e06576   minor tweaks to the overall layout of the code
       new  5071a05   reverted plot_iris accidental commit
       new  0c2196a   added unlabeled_identifier explanation to docstrings
       new  9232f88   Merge remote-tracking branch 'mblondel/kmeans_transform2' into kmeans_transform2
       new  4628d2c   Merge pull request #302 from jakevdp/manifold-doc
       new  3251031   Merge remote branch 'upstream/master' into pairwise_distance
       new  041651e   Initial changes to improve this module. pairwise_distance now uses a dict for functions.
       new  17b0a13   Working through some of the errors in testing
       new  7d1a737   Fixing twenty_newsgroups
       new  ccc181b   Fixed a few import errors
       new  0358e83   Initialise random state for examples
       new  d354e02   DOC: Added +ELLIPSIS for examples
       new  7d0042f   ENH: Support binary (-1,1) classification as well as [0,...,K) multiclass classification
       new  1fcb4f8   Merge remote-tracking branch 'upstream/master'
       new  42a6ebb   DOC: section level fix in clustering doc
       new  2cd1c83   fixed indentation problem in documentation rst
       new  2a7af09   conformance to pep8
       new  84090fc   fixed bug in tests causing gram matrix construction to not work properly (assumed casts to floats)
       new  1c123c6   FIX : corrected a bug in as_float_array and added a test function
       new  468c553   PEP8 : replaced tabulations by spaces
       new  ef55380   FIX : if X is already of the good type, we musn't modify it       Added a verification in the test
       new  a3dc4e7   added two new examples, including an active learning demo with label propagation
       new  95e1be7   Mean shift: removed obsolete reference to KD-Tree with reference to BallTree
       new  d3dfd65   Removed obsolete import of izip, made description of complexity more concise and accurate
       new  18dea8a   heavily downsampled digits examples (runtime a few seconds now) and removed supporess_warrnging bug
       new  2412daf   changed doc to remove long runningtime warning
       new  f902930   Mean shift: settled on term 'bin' and removed unnecessary references to 'bucketing' or 'discretization' from variable names and documentation
       new  3068868   Mean shift: Fixed a minor type
       new  de45027   Mean shift: Moved a test file in preparation for merge with agramfort's branch
       new  b782f7e   Merged agramfort's branch with my own
       new  50bb138   Mean shift: removed my old test script due to merge with agramfort, changed num points in plot example to ten thousand to speed it up.
       new  e77ce26   Typo.
       new  a0e0e3a   In spectral clustering example, forced the solver to be arpack (the amg solver is unstable for this example, and results are not satisfying)
       new  67a0d1a   Merge pull request #314 from emmanuelle/spectral
       new  022b5d8   added least-angle regression to dictionary learning transform
       new  097c3d2   ENH: More interesting benchmarks for OMP
       new  23e4eaf   plugged in lars instead of lasso_lars in denoising example
       new  16ac67c   Example images
       new  83b6af2   Brought my branch for mean shift modification up to date with current head on github
       new  85242e4   Mean shift: modified get_bin_seeds so that it no longer has to copy all points
       new  1bc54d4   VQ example. Not working yet - clusters aren't well formed I think. Will work on it more tomorrow.
       new  fd3d95f   FIX : if X.dtype is changed, then a copy of X is returned, even if overwrite_X  is True
       new  6c6289f   API: eps -> tol in bayes
       new  ccc6ebd   Merge branch 'master' into sc
       new  6c7711c   redesign the denoising example
       new  d402deb   FIX: BayesianRidge doctest
       new  aad2572   tweaked the example a little more
       new  67d3846   removed thresholding from denoising example
       new  d692514   completely removed thresholding from denoising example
       new  bcc63de   initial commit
       new  414a753   added compiled cython
       new  dfe292b   assure C-ordered on init
       new  2f22989   Prettify example
       new  5011d3e   fix NeighborsClassifier doctest
       new  418a32a   More work on example
       new  1074f6b   tweaking example
       new  a31e639   rennamed active learning example so it won't be run for doc compilation
       new  38de418   changed subplot titles so the plot is more clear
       new  73fc08a   Merge git://github.com/scikit-learn/scikit-learn into enh/tree
       new  d0e4757   Removed unnecessary import
       new  1bd370e   Fixed doctest example
       new  63a9a4c   Updated documentation for class interface
       new  21904ed   Minor patches to docs
       new  8811af0   Dataset loader moved to datasets.base, but not being installed
       new  1bc5e57   Fixed loader problems
       new  e4fd194   make memory allocation more efficient
       new  1ea3c95   documentation clarifications
       new  8af062f   DOC: clarified and enhanced dictionary learning narratives
       new  ecee39a   FIX: Corrected NuSVR impl type and set epsilon to None
       new  4b8e1c5   added dictionary learning to classes.rst
       new  9e0b551   Refactor tests for SVR.
       new  34c3fcc   corrected reference to omp
       new  ba3e2dc   Correct NuSVR impl in the sparse case.
       new  9c9e037   Add tests for last commit.
       new  a6912a7   DOC: fixed link to decomposition example
       new  7a8df01   Merge remote-tracking branch 'robertlayton/kmeans_transform2' into robertlayton-kmeans_transform2
       new  1028212   DOC: fix See also
       new  84312d5   DOC: fix See also in both places
       new  e73dce9   Test : lasso_lars_vs_lasso_*
       new  c13bc21   DOC: cleaner see also section
       new  ee5bfb0   DOC: improved dict learning narratives some more
       new  8674b01   ENH set_params method on BaseEstimator, deprecate estimator params to fit
       new  34aee33   set_params: update according to @GaelVaroquaux's review
       new  392cbe4   Rm k param from KMeans.fit again
       new  5a9f68a   Data centering in denoising example
       new  04f354f   Prettify structure example
       new  1216546   DOC: minor style changes
       new  ce9814b   Remove fit params from all objects.
       new  5dcc25a   Merge branch 'normalize_data'
       new  fb906b4   DOC: tweaks
       new  1aec073   Removed print in digits classification example
       new  c78908a   checkpoint style improvements for the KMeans predict
       new  bfd7cd6   track changes from upstream/master
       new  45f218c   DOC: fixed links and made examples build
       new  f0caafe   Merge branch 'clayw-label_prop' of github.com:vene/scikit-learn into clayw-label_prop
       new  67d185e   DOC: clarified example titles
       new  9ff1929   time the main operations
       new  d939914   add warning utils and use it in KMeans when data matrix is integers, boolean, complex...
       new  d1f38c1   fixed structure example
       new  2faa109   FIX : Ellipsis in least_angle.py doctests
       new  8a32927   Removed fit_params from dictionary learning objects
       new  c90b2d9   added vene's subplot adjustments
       new  e2cf62a   Merge branch 'new_lp'
       new  fa339f6   made convergence check function private
       new  47458be   checkpointing work in progress on VQ example
       new  1383385   plot the dictionary in denoising example, other one will disappear
       new  bac02e4   Optimisation: moved _find_best_split to cython.
       new  7b8074f   DOC: change classification to regression
       new  32bbe8b   completely removed the duplicated example
       new  c59fcaf   Prettify the example
       new  377899a   Merge github.com:scikit-learn/scikit-learn into enh/tree
       new  68fecf7   DOC: corrected doctest
       new  864c48c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into normalize_data
       new  6575434   Style + typos
       new  f2b0c61   API : adding proper normalize options in Lasso and ElasticNet with clean up ENH : fix doctests
       new  0e540d9   ENH : more standard import of scipy.sparse
       new  1726411   Rehauled example to show the difference
       new  02b8860   Renamed the example, bounded the difference range
       new  cb593a7   Lower the range of the difference in example for better contrast
       new  a651e8b   Implement one-vs-the-rest multiclass strategy.
       new  dfb0646   Merge pull request #317 from agramfort/normalize_data
       new  51c35bb   Added norm to titles
       new  35d9843   Fix bug in one-vs-rest when underlying estimator uses predict_proba.
       new  022c8dd   More explicit docstring in the example
       new  be023a8   Removed verbosity (example now 4s faster!), prettier output
       new  4437d22   fix output bug
       new  133f7ec   ENH: add missing inverse_transform method for Scaler
       new  cb7f94a   Merge branch 'master' into robertlayton-kmeans_transform2
       new  f50a2b1   PEP8 and style
       new  ea36a26   FIX : ELLIPSIS in least_angle.py doctests
       new  652db0f   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  c922476   Merge pull request #318 from JeanKossaifi/master
       new  3a1d5e2   fix the VQ example by switching to floats in range 0 - 1
       new  974c688   Merge branch 'master' into robertlayton-kmeans_transform2
       new  2a548ed   cosmit
       new  15872a6   use the scipy public API rather than PIL
       new  65b50b7   Merge branch 'master' into sc
       new  1900ca4   Implement one-vs-one multiclass strategy.
       new  c21d8d0   update the documentation
       new  95ca5bb   ENH: 'make test' now runs the doc doctests as well
       new  12e1c51   Merge pull request #316 from jakevdp/cython-ball-tree
       new  5ec50e7   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  32b8b7c   style
       new  e9ae6f1   Merge pull request #2 from ogrisel/robertlayton-kmeans_transform2
       new  c1227f0   Sorting parameters in BaseEstimtor.__repr__
       new  913c883   Implement error-correcting output-code multiclass strategy.
       new  c5962d2   Test grid searchability.
       new  1c1583d   Merge pull request #273 from robertlayton/kmeans_transform2
       new  929c9ff   FIX : docstest fail
       new  fffe829   X -> XA, Y -> XB. pairwise_distance back to metrics pep8'd
       new  9dcbc92   check_set_Y -> check_arrays
       new  c91a08a   Docstrings!
       new  6133a23   Add new meta module to setup.py
       new  e28c667   Merge branch 'master' into multiclass
       new  c0bc07b   Check estimator and fix syntax error.
       new  4d64530   Merge branch 'master' into sc
       new  4972454   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  c51cb3d   Merge remote-tracking branch 'JeanKossaifi/master'
       new  3bdaa9e   don't allocate a new pm for each call: 3 times faster
       new  267ba4b   Use fit_params in Pipeline
       new  67b2ff8   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9a5c0b9   Moved dict_learning stuff out of sparse_pca.py
       new  c6300d8   Merge remote-tracking branch 'JeanKossaifi/sorted_repr' into JeanKossaifi-sorted_repr
       new  3aa87a6   FIX NMF doctests
       new  32dd218   Compatibility fixes for Python 2.6 and doctest fixes
       new  7f357af   ENH: shorter doctest output
       new  f1110e0   rename eps to tol in omp code
       new  d53352f   ENH: pipeline doctest style improvements
       new  79a4a60   FIX: updating doctests in gaussian_process.rst and linear_model.rst
       new  c57950b   FIX: remaining broken doctests
       new  5da456d   Docstrings about labels + cosmit in the metrics module.
       new  5fc92a7   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  be25990   Documentation for the meta learners.
       new  a6317ee   pep8-proof.
       new  20f5eaf   Fill missing docstrings.
       new  3b9d7b5   FIX: py3k compatibility.
       new  8deeaf4   Allow one-class only in LabelBinarizer.
       new  9efe5d5   Exposed sparse_encoding, docs not updated
       new  7641255   Add a void cython module affording to check that `make` has been run.
       new  35fe5d7   Rewrite svmlight loader in pure Python for now.
       new  91bf73a   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  5eb0867   Consistent defaults
       new  00150d1   FIX : fix rounding error in test + pep8
       new  be3edab   Updated the first part of the docs
       new  5e8d845   FIX: py3k compat.
       new  01139b6   Updated the docs
       new  b6d0d2e   removed fit_transform for dict learning
       new  baa758d   Updated the narrative doc
       new  f8e578d   FIX: doctests on buildbot
       new  eaa9253   cosmit
       new  3b5b659   Tweaking the example
       new  d04a76a   Mean shift: Fixed a bug that occurs when the cluster_all argument is False
       new  378ad22   Merge remote-tracking branch 'upstream/master'
       new  b0ed0f3   Mean shift: fixed bug introduced during upstream merge
       new  2e17b36   Improved the example clarity
       new  c3825e5   ball_tree protocol 2, but paths are broken
       new  e33f336   FIX : putting back common.py
       new  60b2299   FIX : in meanshift typos, style, example
       new  85c80b0   Moved to @pprett's faster splitting code (debugged)
       new  82b731a   Added more debugging info to graphviz
       new  5d35f90   Moved to the version without a sample mask, since correctly implemented it is almost as fast
       new  8ee20e3   Merge branch 'master' into multiclass
       new  3d6ee4c   fixed spelling error with variable name (indicies -> indices)
       new  8a656a2   Merge branch 'master' into sc
       new  947735a   removed unused imports
       new  6bc7394   fixed all pyflakes warnings
       new  a019ad7   Fix mistake and docstring cosmits in SVC.
       new  74dd2d8   ENH: new example: NMF topic extraction on 20 newsgroups
       new  f393795   FIX: useless arg to argsort in NMF example
       new  6633648   Moved multiclass module to top-level module.
       new  e4ac243   Fix doc!
       new  e11684b   Fix setup!
       new  a727900   Address @agramfort and @ogrisel's comments.
       new  4ccf0ad   Merge branch 'master' into multiclass
       new  a8ea27a   More informative name for color quantization example.
       new  9c74fb2   More explanations and pep8.
       new  4a826cb   Use 256 colors and add title.
       new  31aecce   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  39ff980   Emphasize one-vs-all.
       new  afc847a   Better documentation.
       new  1032295   Fix doctest errors (hopefully!).
       new  f38f022   Document fit_ecoc.
       new  a11bf8f   Typo.
       new  3bf31ca   Fix currentmodule.
       new  f09673f   Fix bad copy-paste.
       new  bc9f375   DOC: change the name scikits.learn to scikit-learn
       new  cd88137   Merge pull request #320 from mblondel/multiclass
       new  975027d   64 colors + random codebook comparison.
       new  9039d1c   Better title + authors.
       new  5c55a50   Merge remote-tracking branch 'glouppe/master' into glouppe-master
       new  1493cce   Merge remote-tracking branch 'upstream/master'
       new  db0aeb7   PEP8
       new  e2c4946   Fixed error of splitting between identical feature vals
       new  ca9f0ed   DOC: updated comments
       new  564134e   Refactor roc_curve method. The new implementation, even if it looks very naive, reduces the computation time of fpr/tpr vectors. roc_curve computation time depends now only on the length of y_score. For comparison, here are the results between the old and the new implementation for the following vectors:
       new  d714982   Fixed memory leak in libsvm
       new  3e6ab68   Merge pull request #326 from bdholt1/fix/svm
       new  f627340   Welcome Brian Holt
       new  ba4a5ce   Improved graph visualisation
       new  06ffdfa   Ran tests and fixed a few bugs. Unit tests added.
       new  1a85097   Less verbose name
       new  cbaf294   Move initial entropy computation outside loop.
       new  67f1a39   raise ValueErrors with appropriate messages
       new  8720c6e   Merge branch 'cython-ball-tree'
       new  c9fcb82   merged upstream master
       new  a1b04dc   merged upstream master
       new  6da31e4   Standardise error messages
       new  52f76ad   move ball_tree.pyx to scikits/learn/ and write pickle test
       new  cdc3b17   Welcome to Robert and Gilles.
       new  24203c3   Copied ensemble and random forest classifiers to new branch
       new  ac0224c   Check that labels are in range for multiclass classification
       new  0970d6b   Check that labels are in range for multiclass classification
       new  1021baf   FIX: broken example
       new  a6918a4   Generate thumbnails in the example gallery
       new  a4d5760   Further clarification of error messages
       new  65f2d61   Merge branch 'upstream-master' into crossval
       new  7f78eaa   Link images to example file in new gallery
       new  f8ef06e   FIX some broken examples.
       new  0cbe2a1   Rename face_recognition so that the result is plotted.
       new  bf8af73   Merge pull request #328 from bdholt1/crossval
       new  ab5db97   Revert "Rename face_recognition so that the result is plotted."
       new  d70850f   Fixed regression bug.  Thanks @pprett
       new  0084b82   Merge branch 'upstream-master' into enh/tree2
       new  096063f   merged enh/tree2 into enh/tree
       new  3886158   Fixed doctest
       new  ece7329   Enforce 64bit and 32bit types and correct regression bug (divide by zero).
       new  165125f   Refactored construct to subsample dimensions.
       new  7cbc4b7   added NavigationToolbar to SVM gui
       new  23af548   store all tree parameters in the RF base class so that clone() will work
       new  7809dfe   Cross_val : Removed useless & tricky parameter iid
       new  713fbfa   Merge commit 'upstream-RW/master' into cython-ball-tree
       new  d095b2c   add BallTree pickle test cases
       new  4ffcd2e   Merge branch 'cython-ball-tree'
       new  66187d8   Merge pull request #331 from JeanKossaifi/master
       new  9e1ba8f   DOC: Fix doctest
       new  015b456   DOC: scikits.learn -> scikit-learn
       new  4d1bdf1   DOC: fix link
       new  4fd0a3c   optimized _build_graph with inplace methods, conform to standards with variable names
       new  67a56c2   FIX: linnerud dataset mixed variables.
       new  b40247b   Layout tweaks
       new  4f2c7d1   Cosmetic changes in example docstring
       new  cca1fcf   layout tweaks
       new  2f82370   Revert to _Fixed Doctest_ and added regression bug fix
       new  2cc13a4   update to unit test and doc test
       new  cfb269e   enforce type on storage arrays
       new  ec93ca1   enforce 64 bit types on parameters
       new  fd0d3b8   Move project directory from scikits.learn to sklearn
       new  3be3de1   Add a compatibility layer for some modules.
       new  bb6f8b3   further type enforcement
       new  41138c6   initialise variables
       new  ae48fc5   Forgot to add a blank image for the docs.
       new  9fe3908   Revert "Add a compatibility layer for some modules."
       new  68f27e3   Revert "Move project directory from scikits.learn to sklearn"
       new  ddf4b72   Move project directory from scikits.learn to sklearn
       new  30cba7d   Add a compatibility layer for some modules.
       new  403ee2b   correct imports
       new  a5538b9   refactor neighbors module
       new  3ec3153   Merge pull request #335 from fabianp/rename
       new  49c97b7   Test for tuple input. Tests now run in suite (forgot to have test_ at start of func name!)
       new  1a514f7   Add more modules for compatibility layer.
       new  be9d945   More renaming.
       new  4d1dcd2   DOC: scikits.learn -> sklearn
       new  d36c2ac   scipy.lena() no longer works on scipy's dev version.
       new  f1f2baf   XA -> X, XB -> Y
       new  ed9d1ac   Merge branch 'master' into pairwise_distance
       new  8739c91   Moved metrics file to sklearn
       new  a0eb599   removed unused import, removed unnecessary backslash
       new  5e4d984   Improved names and documentation for Leaf and Node
       new  2fa5363   Renamed K to n_classes
       new  bf6a075   renamed F to max_features
       new  38cff84   renamed features to X
       new  2b27be5   renamed labels to y
       new  03b41d7   renamed n_dims to n_features
       new  ca9d2b6   explained min_split
       new  30fb575   renamed C to predictions
       new  9878ac4   improve documentation
       new  b5d63ba   renamed K to n_classes
       new  3aa4e6d   COSMIT: improved documentation
       new  fc428b1   pairwise_kernel function (untested, for comment)
       new  116da4e   renamed pm to label_count
       new  3a661de   renamed K to n_classes
       new  a92c172   improved documentation and renamed features and labels
       new  d771289   renamed var to variance
       new  487de2c   fixed comments
       new  e6fa29c   PEP8 of metrics.py
       new  a57c2ae   more scikits.learn => sklearn updates
       new  2c8f756   import to metrics namespace for pairwise_kernels
       new  58af206   DOC: Minor scikits -> scikit
       new  42d6c56   BUG: sklearn/setup.py : learn -> sklearn
       new  e05fefe   ENH: new Makefile target to cythonize everything
       new  5994df9   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4eec40e   Updated docstrings
       new  091d92a   batch re-cythonization with version 0.15 and new package names
       new  8511b00   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into pairwise_distance
       new  b484aa4   merged upstream-master into enh/tree
       new  2f11683   doc fixes
       new  17c4a3c   merge with upstream/master
       new  769714e   Merge remote-tracking branch 'upstream/master'
       new  c090d43   Fixed examples
       new  860c1aa   new package name
       new  9b5c7cd   more renamings
       new  cb27ffa   fix typo in scikits.learn.qda
       new  cf5a889   update the Makefile test-coverage target to work with the new package layout
       new  f3173b6   Merge branch 'master' into bdholt1-enh-tree
       new  f0d88a8   one more optimization! avoids cast to numpy matrix and does in place matrix multiplications
       new  4cb9e70   fixed test cases to conform to api changes & new internal parameters
       new  9ec87e4   Merge branch 'master' into pairwise_distance
       new  56ff27d   FIX: fix variable referenced before assignement in libsvm.pyx
       new  4c18e46   trailing spaces in pyx file
       new  f0f0182   More style consistency improvements
       new  9da4e66   style: constant in capital letter on top + extract graphviz tree template
       new  481e36e   cosmit
       new  456ef97   More style improvements in _tree.pyx
       new  99b1275   ENH: cross_val module docstring and style improvements
       new  a03f836   Improved test coverage to 100%
       new  ae8f66d   ENH: more randomized cross val docstring & var naming improvements
       new  a29b1b1   Merge branch 'master' into bdholt1-enh-tree
       new  6d69591   Merge pull request #9 from ogrisel/bdholt1-enh-tree
       new  c30bc73   ENH: doctest simplification by using the cross_val_score func
       new  19a4bf9   Added RFE into the narrative documentation
       new  adc70f9   Doc: grammar
       new  515d6e3   Added n_features_ attribute to RFE
       new  453968f   He who seeks only vanity and no love for humanity shall fade away
       new  a6ef4e4   style
       new  e93a760   Moved "feature selection" section back into the "supervised learning" chapter
       new  33d4f7e   Better exception messages in SVM
       new  fde6ba1   ENH: make the cross_val_score able to use functions from the metrics module
       new  96e9f84   ENH: better docstrings for SVMs
       new  b84216d   Merge branch 'master' into cross-validation-improvements
       new  072d243   Merge branch 'enh/tree' of https://github.com/bdholt1/scikit-learn into bdholt1-enh/tree
       new  47b82fc   DOC: improvements to the cross validation doc layout + missing ref to ShuffleSplit and cross_val_score
       new  7137010   Merge remote-tracking branch 'bdholt1/enh/tree' into bdholt1-enh-tree
       new  e3f6c5e   Merge branch 'master' into bdholt1-enh-tree
       new  b87a75d   Merge remote-tracking branch 'glouppe/master' into glouppe-master
       new  66be88f   Merge branch 'master' into glouppe-master
       new  880bdf9   Add missing authorship + license info to NMF topics example
       new  1c75500   updated docs!
       new  40eb101   Merge branch 'master' into sc
       new  00c3d4d   Merge git://github.com/scikit-learn/scikit-learn
       new  38103a3   Merge branch 'master' into cross-validation-improvements
       new  38e88d7   Copied tests, fixed examples imports, enhanced see alsos
       new  59f164a   ENH: more cross_val doc for LOLO and LPLO
       new  21fc893   DOC: add info about smart CV and IC estimators
       new  7a32fbe   Merge commit 'upstream/master' into neighbors-refactor
       new  637d0c7   scikits.learn -> sklearn
       new  db24cfd   add neighbors benchmark
       new  2eba5a4   localized a variable
       new  3e83638   fixed test suite, changed module to conform to new sklearn naming scheme
       new  472cf5b   fixed examples for new naming scheme
       new  813de49   cosmit
       new  6f6b2f4   ENH: s/n_labels/n_unique_labels/g in cross_val
       new  732f18a   FIX: compat with numpy version lacking the out argument for dot
       new  a8c77e8   ENH: misc style / docstrings improvements
       new  a695a67   Merge branch 'master' into pairwise_distance
       new  5fab47a   Merge pull request #10 from ogrisel/bdholt1-enh-tree
       new  c018f60   Ensure 0.0 on diagonal elements if X is Y
       new  701438e   Merge https://github.com/bdholt1/scikit-learn into bdholt1-enh/tree
       new  c4c92d5   Merge branch 'enh/tree' of https://github.com/bdholt1/scikit-learn into bdholt1-enh/tree
       new  abefd7b   introduce reset method for Criterion and implemented linear version of MSE.
       new  41e6602   change implementation to mixin pattern
       new  9e85e61   fix: weight left and right variance by num samples in each branch cosmit: rm some comments
       new  e1a1b3e   merged upstream/master
       new  59879d4   merged ogrisel's docs & optimization, also fixed active learning example plot
       new  b8d4d4b   renamed scikits.learn to sklearn
       new  be964cd   Push coverage up to 96%, added graphviz test
       new  40fa5b0   merging
       new  891b81d   Merge pull request #11 from pprett/bdholt1-enh/tree
       new  fa1e781   added example usage of graphviz
       new  6be3921   Merge branch 'enh/tree' of github.com:bdholt1/scikit-learn into enh/tree
       new  05af797   Doc: Implementation details of euclidean_distances
       new  d93bef8   Merge pull request #341 from ogrisel/cross-validation-improvements
       new  2664d0c   Sparse matrix support in the `density` util.
       new  b93cde8   Documenting a secret feature and fixing bugs in the process.
       new  39c24bb   Use l1 penalty.
       new  6d9ca07   Merge pull request #343 from glouppe/euclidean_distances
       new  e609775   Tests working, mostly pass
       new  ca48f82   s/\bcross_val\b/cross_validation/g
       new  0f7fb46   backward compat for cross_val namespace
       new  93f598e   cosmit
       new  7b81afd   ENH: `np.fill_diagonal` replaced with more portable code. Added an explanatory comment.
       new  b435059   fixed unit test of graphviz
       new  37596d0   added trees (boston and iris datasets)
       new  91304a5   pep8
       new  1583dbb   move neighbors.py -> neighbors
       new  fbd6237   fix doctests
       new  170e995   BUG: Backward compatibility layer sklearn.externals
       new  5d08280   API: start 'API changes summary' section in doc/whats_new.rst
       new  84a2795   ENH: Add verbosity control to LinearModelCV
       new  22ea8b6   API: removal of fit parameters
       new  18ecba7   FIX: fix broken tests on ElasticNetCV
       new  78dd8b5   batch trailing spaces cleanup
       new  2999ccd   ENH: docstring cleanup
       new  73fd2b9   Mark sklearn.hmm as orphaned
       new  0718f72   merge upstream/master
       new  7a6bad9   move barycenter_weights to manifold
       new  307750b   deprecation of NeighborsClassifier and NeighborsRegressor
       new  0c47a52   Merged PR263 into this PR Basically, l1 now returns a pairwise distance matrix by default, with an option for componentwise version.
       new  63c94aa   Fixed merge conflict
       new  ad79b4d   Fixes based on ogrisel's comments
       new  499f2f5   l1_distances -> manhattan distance new l1_distances set up, which warns user to use manhattan_distance instead, and sets sum_over_features=True for compatability.
       new  04f0b2a   pep8'd and pyflakes'd
       new  8ac3ac1   scikits-learn -> sklearn
       new  272af90   FIX: make the @deprecated class decorator not break the __repr__ of estimators
       new  1215c0e   Remove l1_distances completely, updated gaussian_process update doc to make clearer
       new  97a5241   Actually removed l1_distances this time
       new  482203f   DOC improve fbeta docstring
       new  e38cb4d   Merge commit 'upstream/master' into neighbors-refactor
       new  db8e9c2   add deprecation warning to sklearn.ball_tree
       new  c10c64e   Note neighbors module changes in doc/whats_new.rst
       new  9f42fd3   added CART to covertype benchmark -> look at that error rate!
       new  eb51d07   Fix typo in SGDClassifier's docstring (via GitHub).
       new  5485569   Merge pull request #346 from npinto/patch-1
       new  efacf55   DOC : fix sgd docstring
       new  b592902   test_checks merged into test_pairwise. test_checks is empty for now. Several other small changes made based on comments received.
       new  10a551e   moved the min_split test to beginning of recursive_split
       new  9b16112   group imports by hierarchy
       new  286a80d   sed s/dimension/feature/g
       new  9c9320d   time is measured in seconds
       new  3fe8d66   print left and right child repr in graphviz
       new  9bace88   Merge branch 'enh/tree' of github.com:bdholt1/scikit-learn into enh/tree
       new  d905b58   fixed graphviz test failure
       new  6071a3e   added feature_mask to reduce fancy indexing
       new  ef64142   replaced == with 'not' operator
       new  a841381   fix typos
       new  424e447   gitignore: scikits.learn -> scikit_learn
       new  364d1df   Merge commit 'upstream-RW/master'
       new  af7d7dd   updated the decision tree docs (not done yet)
       new  db8fb53   use Fortran array layout
       new  22c0d86   ENH : better plot_img_denoising
       new  42b06b7   Removed test_checks
       new  e271afd   Fixed doctest and checked tests working - most are;
       new  c598c15   corrected feature_mask implementation
       new  f199e5c   allow for different architectures
       new  7eae6a9   pairwise callable metrics fixed test_pairwise now has a sparse test in test_pairwise_distances test_pairwise tests now pass, thanks to assert_array_almost_equal
       new  fde272e   Now tests if tuples given as input
       new  ec424eb   check_pairwise_arrays now ensures at least two dimensional arrays are returned.
       new  64a45a5   pep8'd and pyflakes'd
       new  a694b67   ENH: implementation Adjusted Rand Index for clustering evaluation
       new  a170da3   Merge remote branch 'bdholt1/enh/tree' into bdholt1-enh/tree
       new  eedab38   minor fixes in clustering metrics
       new  c9f20a2   merged upstream/master moving to sklearn
       new  ba381d7   merged enh/tree
       new  829ff69   visitor pattern for export graphviz
       new  0bae054   move neighbors examples to examples/neighbors/
       new  f084227   cosmit
       new  416bc6a   removing the undocument implementation of the unadjusted Rand index in kmeans_
       new  11bb4c8   cosmit
       new  d9abd06   missing import in the metrics namespace
       new  36f123c   DOC: narrative documentation for the ARI
       new  b5b9f8e   metrics listed in pairwise_distances and pairwise_kernels
       new  e7774b9   BUG: scikits.learn -> sklearn: backward compatibility
       new  693bc15   COSMIT: PEP08
       new  6a3103f   kwds ws being passed to squareform, instead of pdist. this has been fixed, with a test added
       new  b1b8c34   DOC: typos
       new  75f2617   FIX: fix broken document clustering example and add ARI to examples
       new  0e29205   add doctest for combinations (to document the n < k case)
       new  9997b81   Merge branch 'master' into sc
       new  eeb8585   Merge branch 'sc' of git://github.com/agramfort/scikit-learn into sc
       new  e3a83de   pairwise helper functions to give verbose knowledge of which metrics are which functions
       new  b60f7d8   Fix commenting in pairwise_distance
       new  1e3824b   check for sparse matrices for scipy metrics, and throw error. test included
       new  6844139   more tests for ARI and clustering metrics
       new  8572adc   test non consecutive integers in perfect match
       new  20421fb   FIX: use scipy's fast implementation of comb + fix tests + limit cases + faster adjustment test
       new  e0c3747   cosmit
       new  2fc866f   OPTIM: use exact comb evaluation since it's faster for the ARI case
       new  5cd048c   Added link to personal web page
       new  bc02942   Nearest Neighbors examples & documentation
       new  60a6d97   cosmit
       new  fbf839a   switch to dynamically generated docstrings
       new  0d3e198   commit dynamic doc changes
       new  863f38f   cosmit
       new  ddbf08c   DOC: add example to illustrate the concept of adjustment for chance
       new  f8dafb6   more details about ARI value range
       new  6c97f38   add weighting to classification and regression
       new  e985958   add neighbors/tools to commit
       new  b990db5   add tests for weighted regression and classification
       new  4401ea4   documentation of weighted classification and regression
       new  3675291   add graphical neighbors benchmark
       new  dfbc00d   pep8 + move weighted_mode to utils
       new  0624590   Unused import
       new  25db7c8   BUG: backward compat: scikits.learn -> sklearn
       new  5a9820f   add tests & example for weighted_mode
       new  43feed9   Brief description of kernels and distance metrics in doc
       new  c00b05a   Added a list o
       new  53be9c9   benchmark -> bar plot
       new  3305bc3   make constants uppercase
       new  5da44bd   make example script filename more explicit
       new  ed434c3   typo
       new  f96f456   Merge branch 'master' into cluster-metrics-2
       new  2292946   Little more description
       new  a6c5c18   Merge pull request #12 from pprett/bdholt1-enh/tree
       new  bccae58   return to simple docstrings
       new  625b358   Incorporated suggested changes to Graphviz exporter
       new  7908be9   visit -> export
       new  6fae82d   cosmit: added spaces
       new  5e84258   cosmit: improved documentation
       new  3908c61   Merge branch 'master' into sc
       new  0fc78f9   fixed indentation and added section on memory requirements
       new  dd30699   Included dictionary learning online in decomp example
       new  f2a8df8   Added missing dashes in doc
       new  f3d5382   increase BallTree test coverage
       new  59a0e4e   Updated documentation to include the iris svg example
       new  58bef8b   fix BallTree linkage
       new  1be34bc   Merge remote-tracking branch 'jakevdp/neighbors-refactor' into jakevdp-neighbors-refactor
       new  0d4fa9e   fix typos
       new  db4f4e4   cosmit + docstest
       new  f75ee72   Merge pull request #3 from ogrisel/jakevdp-neighbors-refactor
       new  cca5aff   cosmit: pep8 + docs
       new  a15554d   Do not import mixture from top-level sklearn.
       new  55babd7   DOC: add parameter C to docstring.
       new  c8f0936   Use LinearSVC's docstring instead of outdated one.
       new  4a8273f   scipy.lena has moved to scipy.misc.lena in scipy's dev version.
       new  c7f16d8   increase test coverage
       new  632da98   Fixed typos
       new  1ba99ed   DOC: reorg, bold important points, include adjustment plot as figure
       new  1e81d96   typo
       new  d89cee1   Use 1 / n_features as value for gamma.
       new  c315c5a   improved documentation
       new  0a50c71   Merge pull request #347 from ogrisel/cluster-metrics-2
       new  90705a5   FIX broken tests by last commit.
       new  c4e97d6   np.float64 -> DTYPE.  Set DTYPE to np.float32.
       new  aca6fa0   cosmetic changes to ari_score
       new  6ba7aab   pep8 + cosmetic changes
       new  fa50e68   rename ari_score adjusted_rand_score
       new  8cae2c1   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  0c25d43   fixes typo in roc_curve method
       new  201ce57   [refs #350] - variable renaming regarding reviewer comments
       new  9ac01aa   make sorting more efficient by transposing and sorting along last axis.
       new  da036d0   Merge remote branch 'bdholt1/enh/tree' into bdholt1-enh/tree
       new  52cc103   Removes useless (and time consuming) statement.
       new  54af74d   Merge remote-tracking branch 'jakevdp/neighbors-refactor'
       new  ff4a3f3   more enhancements, variable names and test fixes
       new  0a942f5   manhattan_distances now returns [n_samples_X * n_samples_Y, n_features_X] shape array This is instead of the [n_samples_X,  n_samples_Y, n_features_X] it was returning.
       new  d22ec2c   Doc update for manhattan_distance
       new  3f01fd0   Fixed doctest error
       new  1a06a46   changed a bunch of variable names, fixed some test cases
       new  ec149ed   all code works great, all tests pass, full coverage
       new  fba4c7c   changed a variable name to conform to scikits code
       new  1187db6   Use a sample mask instead of fancy indexing.
       new  0b613a4   Add changelog for changing gamma parameter.
       new  1831875   Merge remote branch 'bdholt1/enh/tree' into bdholt1-enh/tree
       new  6e07284   FIX example logisitic regression.
       new  4664bed   Move matrix factorization to work in progress.
       new  1b40058   Initial changelog -- to be completed.
       new  b29dea1   More changelog and .mailmap
       new  fe1305f   Why, emacs, why ??
       new  4ddb27a   Update changelog
       new  52e756d   DOC: broken link to example
       new  d0f0600   Improves signal sorting method (using numpy primitives). Adds dtype on tpr/fpr array creation.
       new  dbc7347   ENH: add control of n_init in spectral clustering
       new  431723c   correct variable names and added inline comments for active learning examples
       new  643e559   Merge pull request #350 from tinyclues/master
       new  fcfa048   STY : pep8
       new  42f8892   use hybrid sample_mask fancy indexing approach.
       new  3ee7314   added attributes text to explain named attributes
       new  27eb6c4   FIX: add test fixtures to distribution.
       new  65ce680   FIX: broken link to example
       new  f1ecd92   DOC: always generate pages for autosummary.
       new  a6d6b3a   cosmit: docs + rm comments
       new  8b561dd   added `min_density` parameter to CART
       new  74d0e29   raise ValueError for min_split and max_depth on __init__ rather than fit.
       new  67a017b   BUG: scikits.learn -> sklearn backward compat
       new  13dfebb   Edited sklearn/metrics/pairwise.py via GitHub
       new  efddf2a   STY: mostly style + avoid a zip in favor of an np.argsort
       new  3a1e98c   STY : in label_propagation.py
       new  082c873   ENH : using numpy broadcasting instead of dot_out
       new  ade5f52   Changes on the feature_selection module.
       new  db17856   Merge pull request #13 from pprett/bdholt1-enh/tree
       new  dc30c35   COSMIT: corrected comments
       new  7d11a32   made sample_mask a fit parameter
       new  9baf540   updated documentation to reflect min_density concept
       new  ebd74dc   Added items for cross validation and clustering metrics
       new  958f591   Initial Silhouette Coefficient code. no tests yet, and haven't checked it actually works yet as well
       new  5d6ddf0   Initial test. Not working yet
       new  7b15624   FIX: some sklear.test() fixes.
       new  0ebe801   Add Vlad as GSOCer
       new  c74347f   trailing spaces
       new  9a0db43   Merge remote-tracking branch 'vene/sc' into vene-sc
       new  6ec4e1b   cosmit
       new  097f26a   we grow our trees deep
       new  0a1df77   cosmit + n_samples fix
       new  1be8bf3   DOC: howto register the %lprun line_profiler magic on IPython 0.11+
       new  ca92354   Merge branch 'master' into sc
       new  5841984   Merge branch 'vene-sc' of git://github.com/ogrisel/scikit-learn into sc
       new  f0b0cfe   DOC: larger lena size in denoising example
       new  3124785   Merge pull request #313 from robertlayton/pairwise_distance
       new  200031e   Merge branch 'master' into sc
       new  531b187   Merge branch 'dictionary_learning' of git://github.com/GaelVaroquaux/scikit-learn into sc
       new  167f340   renamed MiniBatchDictionaryLearning
       new  84b8c67   layout
       new  9c54b20   Reordered dictionary learning docs
       new  2121bfc   tweaked faces decomposition and added to dict learning docs
       new  1a6f017   added dict learning face decomposition to docs
       new  5956b55   Merge pull request #14 from pprett/bdholt1-enh/tree
       new  37909f7   there is no more Leaf class
       new  ba509fa   added feature_names to GraphViz export
       new  76a2858   Merge branch 'master' into vene-sc
       new  de868c6   Merge remote-tracking branch 'vene/sc' into vene-sc
       new  b03dfa0   Fixed image display in docs
       new  cf7ae85   Merge branch 'sc' of https://github.com/vene/scikit-learn into vene-sc
       new  2359d92   simplified fit_algorithm keyword
       new  77a0ec3   s/img_denoising/image_denoising
       new  5a8c416   scikits.learn -> scikit-learn + url of Numpy
       new  3bb9186   s/scikits\.learn/sklearn/g
       new  463120b   Merge branch 'sc' of https://github.com/vene/scikit-learn into vene-sc
       new  8fee910   made sparse_encode functions visible
       new  18c6e1e   added see also refs to sparse_encode functions
       new  59e2de4   Tidied up graphviz related code
       new  639fcfb   Reordered dictionary learning docs
       new  e93612a   Merge branch 'sc' of https://github.com/vene/scikit-learn into vene-sc
       new  70ba1f5   test for improperly formed feature_names
       new  61a3e00   removed sample_mask parameter
       new  36b81e5   Stabilized and improved face decomposition example
       new  c49683c   only return values that are used
       new  d56281e   explicit seeding of olivetti faces loader
       new  565c938   Merge branch 'sc' of https://github.com/vene/scikit-learn into vene-sc
       new  4dd6f1e   Complete Changelog.
       new  4701854   FIX: import path under scipy's dev version.
       new  10c146f   Cosmit: make in-place modifications explicit
       new  241ebba   Merge branch 'vene-sc'
       new  c61ecac   LassoLarsIC/CV and metrics.roc_curve in whats_new
       new  c72711f   DOC: update whats_new.rst
       new  95d667d   Merge pull request #7 from larsmans/sklearn
       new  57b9fd3   Comment tests that depend on PIL.
       new  56a0392   Comment out tests for the current release.
       new  c8010b9   FIX typo
       new  d7d64b6   FIX: docstring for RadiusNeighborsRegressor
       new  219bc2f   BUG: ShuffleSplit: repr for random_state not number
       new  05fdfa8   non-regression test for warm-start intercept shape using a binary dataset
       new  d78bf25   letting intercept_init be of shape (), reshape to (1,) for consistency
       new  00cc485   sklearn.test() does not like doctest that don't print.
       new  ecf4b43   Doc: Print --> Issue
       new  07bf0d3   Included distance helper functions line for 0.9 release
       new  8640931   API changes in metrics/pairwise.py
       new  5526951   Cosmit.
       new  f4e2249   if __name__ == '__main__' multiprocessing protection in ex2
       new  71953c4   Safer assert_all_finite.
       new  ab4098e   pep8 sklearn/utils/__init__.py
       new  8b42048   Some more doctest fixes for sklear.test()
       new  3e3872c   Update commiters list
       new  9748a50   Start of 0.10 development cycle.
       new  9ab6be1   more details
       new  1597ef3   more details in multiprocessing comment in skeleton too
       new  4e9a91f   added hopefully more intelligible error messages.
       new  858b268   refactor linear models to call as_float_array only from _center_data
       new  1034bc5   pep8
       new  859cbd9   pep8
       new  a368a91   unconditionally call as_float_array in LinearModel._center_data
       new  e0dbd23   typo, pep8 and line continuations
       new  57eca1c   test for new error strings
       new  7c3b2e1   slight beautification (in my opinion)
       new  ac198aa   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn
       new  23fb617   Merge pull request #353 from amueller/sgd_warm_starts
       new  a40b576   refactor linear models to call as_float_array only from _center_data
       new  906b00d   unconditionally call as_float_array in LinearModel._center_data
       new  98f4861   don't test on error message, just on raise
       new  51b6f31   pep8
       new  43c5656   Some Python 2.5 fixes.
       new  a3af07a   More python2.5 fixes
       new  aed53bf   FIX: assign NaN to an integer array has no effect on old numpy
       new  18382c3   0.9 release
       new  e7f8212   Some Python 2.5 fixes.
       new  59a7b18   Comment out non-python2.5 test
       new  9c3f807   More python2.5 fixes
       new  30070c9   FIX: assign NaN to an integer array has no effect on old numpy
       new  8c4fa1b   DOC: cross validation: introduce motivation and basic usage first
       new  fd1063f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bc9ba09   Some more changelog stuff.
       new  9ee0c1e   Some more changelog stuff.
       new  45e8ac7   Update MANIFEST.in: scikit-learn --> sklearn
       new  99f6be9   Update MANIFEST.in: scikit-learn --> sklearn
       new  cc21c5c   update cluster docs (DBSCAN)
       new  48f430f   DOC: formatting examples as a topic
       new  68e6544   ENH: GridSearchCV can has predict_proba
       new  fd79a67   FIX bug introduced in 68e6544
       new  455e214   Revert "Comment out non-python2.5 test"
       new  0f7e308   Comment out test non compat with python2.5
       new  6624976   Giving due credit (last minute ChangeLog item).
       new  9955ed5   DOC: fix typos
       new  cc678ad   typo: s/accurracy/accuracy/g
       new  6182960   Merge branch 'master' of github.com:scikit-learn/scikit-learn into enh/tree
       new  7d3bc4a   Merge branch 'enh/tree' of github.com:bdholt1/scikit-learn into enh/tree
       new  ab791ef   DOC small stuff in base.py and multiclass.py
       new  adcb8e0   Add mldata loader and olivetti dataset to changelog.
       new  4a1dc2b   Add mldata loader and olivetti dataset to changelog.
       new  4f85e55   Giving due credit (last minute ChangeLog item).
       new  00370c8   DOC: fix typos
       new  3f1ea66   Comment out doctest for fetch_mldata
       new  614d12e   MSE bugfix (MSE.eval used to weight variances by n_left and n_right).
       new  3188200   take DTYPE from _tree extension module
       new  57fa068   fix: inc n_left, n_right before variance computation; hopefully the last bugfix for MSE...
       new  422a5d4   fix doctest + recompile cython code (accident)
       new  588b038   Merge pull request #16 from pprett/bdholt1-enh/tree
       new  6bd0980   use np.isfortran
       new  68a2967   use None as the default marker
       new  148f78a   compute node id's on the fly
       new  beaab34   removed leftover class_counter
       new  d5fdaf0   trees: don't use deprecated cross_val, error messages, use super
       new  4d0cddf   Merge pull request #17 from larsmans/enh/tree
       new  23c21d7   added test for pickle-ability
       new  674a25f   Merge branch 'enh/tree' of github.com:bdholt1/scikit-learn into enh/tree
       new  4c9da52   make Node an extension type + change class label indexing.
       new  dad3dea   recompile _tree.pyx
       new  0d4764a   make _tree import relative
       new  e3b2c5c   Modified package name in Easy Install section.
       new  f8435fe   make node pickleable & tidy up some rebase mistakes
       new  95e75df   remove obsolete tests
       new  124fcfdf  Merge pull request #360 from cmd-ntrf/master
       new  f7f1714   typo: threhold -> threshold
       new  18d5a9e   Merge pull request #19 from pprett/bdholt1-enh/tree
       new  851847f   fixed failing docttest
       new  b81cf2a   improved tree documentation
       new  ac989c4   DOC minor editing to naive_bayes docs
       new  b8d6cb6   Merge branch 'tmp'
       new  0668fd3   included a mathematical formulation for CART
       new  5891d98   Remove BaseLibLinear.predict_proba not implemented
       new  3ca4949   DOC: Install.rst wrong packaging info
       new  6474b87   verify that scores from pickled objects are equal to original
       new  e7e279e   pep8
       new  2aaa1fe   COSMIT
       new  c663517   scikits.learn -> scikit-learn in README
       new  71aa52d   `scikits.learn` in the README, to catch google
       new  9be71a8   Cosmit.
       new  56d26f9   rename overwrite_Foo params to copy_Foo (and inversed their meaning)
       new  8f72745   document overwrite_ -> copy_ API change in ChangeLog
       new  2da76f7   DOC: fix rst
       new  b3b507e   TEST: skip unreliable doctest
       new  c1c7070   Merge pull request #354 from amueller/liblinear_parameter_errors
       new  dec37be   DOC: minor doc ENH for trees
       new  0813b93   COSMIT: tree code simplification
       new  973e32d   COSMIT: np.random should never be called
       new  0224659   COSMIT: no seeding of the global RNG
       new  7addd70   ENH: move parameter checking to fit
       new  2f1ce85   COSMIT: y is a vector, not a matrix
       new  c9b0913   Cosmit, PEP8
       new  185f7ab   DOC: doc and example cosmetics for trees
       new  716b1e7   DOC: improve spectral clustering docs
       new  27fe56e   API: spectral clustering uses arpack by default
       new  c4f32dc   DOC: proper docstring for load_sample_image
       new  ee0aae7   Merge branch 'silhouette' of https://github.com/robertlayton/scikit-learn into silhouette
       new  6065f24   Test working, pep8'd and pyflakes'd
       new  3841f43   Sparse matrix testing
       new  6ebefad   Merge pull request #20 from GaelVaroquaux/tree
       new  6e7ec46   COSMIT: +SKIP on classification doctest
       new  f1113fc   BUG LinearSVC.predict would choke on 1-d input (+ regression test)
       new  2e4bb7e   ENH: Cleaned setup.py
       new  c5ad2c5   rewrote GraphvizExporter into a function export_graphviz
       new  edbdf48   Swapped y, D to distance, labels
       new  b6e691a   more helpful error message in SGDClassifier.predict_proba with wrong loss
       new  be2683a   Merge remote-tracking branch 'bdholt1/enh/tree' into tree
       new  b029a5b   removed duplicate tests (already in fit)
       new  cfc3b8b   DOC: Some docstrings have been rewritten + small cosmetic changes
       new  170db7b   Merge remote-tracking branch 'bdholt1/enh/tree' into tree
       new  e5d4228   DOC: Improved documentation + cosmit changes
       new  bc45db9   COSMIT: GraphViz exporter cleaned up
       new  2474a32   ENH: Made apply_tree_sample slightly more efficient + various cosmits
       new  eaa2a22   Regenerated _tree.c
       new  6f84f45   Merge pull request #357 from larsmans/overwrite-to-copy
       new  c83cae6   Merge pull request #21 from glouppe/tree
       new  41fd435   classes can be any integer values
       new  458cc72   fix doctest failures in linear_models docs
       new  9fb3257   refactor and simplify naive_bayes
       new  02ec8e4   prevent some copying in sparse SGD
       new  d85aca5   API: default in spectral clustering: auto
       new  e6eff45   ENH: no need for L2 norm on input in doc clustering
       new  cfe6b64   BUG adapt text feature grid search example to new 20news loader
       new  a95af4d   BUG fixed and cosmetics in CountVectorizer
       new  5a4d10e   BUG + optimization in GaussianNB
       new  4ef3db8   refactor common code of NB estimators into BaseNB class
       new  4dfbe74   Refactor/simplify naive Bayes tests
       new  597015b   API change: 1-d output from BaseNB.predict_(log_)proba in binary case
       new  d8c6e47   ENH SGD error messages better still
       new  d626e7c   FIX embarrassing SyntaxError in linear_model.base
       new  e53a63d   BUG multiclass.predict_binary still relied on old MultinomialNB.predict_proba
       new  5fef41a   DOC prob_predict -> predict_proba in SVM docstrings
       new  338f4d0   Revert "BUG multiclass.predict_binary still relied on old MultinomialNB.predict_proba"
       new  cc686f0   Revert "API change: 1-d output from BaseNB.predict_(log_)proba in binary case"
       new  edd47de   Add dump_svmlight_file.
       new  f3d3fd1   ENH: make load_files use a fixed shuffling of the samples
       new  0e516d0   Export data option in SVG gui.
       new  d6b2f8a   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  31e9a54   refactor SVMlight reader and writer
       new  4d0793f   silhouette_coefficient -> silhouette_score
       new  1e9402c   Restructured metrics/cluster into a folder with supervised and unsupervised modules
       new  dbc83ca   Narrative documentation
       new  e9ec8c0   require that the next_sample_larger_than is greater than the previous by at least 1.e-7
       new  5e3ae08   regenerate cython
       new  414ce61   DOC: better svmlight_loader / dumper docstrings
       new  712b12f   ENH: 30% speed improvements in load_svmlight_file
       new  f90c8ab   ENH: remove useless call to strip while staying robust to empty lines
       new  5d8fcdf   ENH: make MiniBatchKMeans display more info in verbose mode
       new  2fb124a   API change in SVMlight reader: handle multiple files with svmlight_load_files
       new  569e446   Merge pull request #373 from larsmans/svmlight
       new  178c6e2   check if y.shape[0] == X.shape[0]; this is especially troublesome for svm.sparse because most people are not aware of the sparse matrix - KFold troubles..
       new  8a20cb9   unified predict for sparse and dense SGD.
       new  1b7b031   cosmit
       new  3b34162   fix: use None as default value for class_weight and sample_weight for sparse OneClassSVM; ample_weight -> sample_weight
       new  93f7bd9   cosmit: pep8
       new  0d9c3fb   added y.shape[0] == X.shape[0] check to DiscreteNB
       new  b8b655f   added X.shape[0] == y.shape[0] check to ElasitcNet
       new  34a6bd5   Merge remote branch 'upstream/master'
       new  de6e930   Revert "BUG fixed and cosmetics in CountVectorizer"
       new  6588a6e   ENH: make it possible to skip label assignements in MiniBatchKMeans
       new  fda113f   documented changes in whats_new
       new  0ebbef3   Implements a robust covariance estimator: Rousseeuw's MCD.
       new  1cdcbd4   Faster tests for coordinate_descent.
       new  57beda2   Add changelog entry.
       new  e5c8699   Retry "BUG fixed and cosmetics in CountVectorizer"
       new  4e84ff7   Integrate Fabian's comments on Minimum Covariance Determinant.
       new  246903d   CountVectorizer.fit_transformer refactoring, part N
       new  85c680c   thanks to @larsmans, TFIDF is now always positive :)
       new  d6783c0   Micro-optimize NMF for memory usage: topic spotting example down by ~17%
       new  86c0df6   Merge pull request #376 from fabianp/fast_tests
       new  334bbd1   if threshold is indistinguishable from a, choose b
       new  bd5677b   Replace two more flatten()s in NMF with ravel()s
       new  4078d6b   FIX broken doctests in NMF + pep8
       new  1e3f1dc   Allow sparse input to NMF
       new  f24d6dd   NMF: cosmit
       new  8b57f1e   Merge remote branch 'upstream/master'
       new  c0881f7   Fixed issue #378 on the RFE module
       new  430fb43   Merge remote-tracking branch 'upstream/master' into silhouette
       new  cfb5230   "whats_new" updated
       new  91f7abc   Example updated, which required fixing a backwards compatability bug (adjusted_rand_score not imported in metrics/cluster/__init__.py)
       new  90c8ee7   Silhouette added to AP example
       new  90366b2   Using pairwise_distances in the Silhouette Coefficient. Updates to docs, code, tests and examples
       new  7ff7e1f   Silhouette calcualted for all forms of k-means in example
       new  86cd33a   fix: trailing-spaces option fixed to be executed
       new  c85e93a   doc: updated docstring for permutation_test_score to reflect nature of p-value given the type of score_func
       new  9d68886   sty: ran make trailing-spaces
       new  013b66e   Faster version by removing inner loop comprehension
       new  1cf24e6   doc: fixed spelling
       new  3ee51c0   doc: updated docstring based on feedback
       new  461046b   Merge remote-tracking branch 'bdholt1/enh/tree' into bdholt1-enh-tree
       new  3fc9e7a   Merge pull request #381 from satra/doc/permutation
       new  730bdc7   Sampling to improve SC speed
       new  42714e5   sampling added to silhouette_score, examples updated to match
       new  a1f273d   pep8 and pyflakes
       new  7e7af61   Updated doc with new API
       new  e2888ed   Removed unneeded line from doc
       new  4462347   FIX: compat with numpy 1.5.1 and earlier in NMF
       new  cb9e8e6   Implements a robust covariance estimator: Rousseeuw's MCD.
       new  9a54ba9   Integrate Fabian's comments on Minimum Covariance Determinant.
       new  80a5baa   Merge branch 'mcd' of github.com:VirgileFritsch/scikit-learn into mcd
       new  2f1bc44   Merge pull request #375 from VirgileFritsch/mcd
       new  edfcd5d   modified threshold comparison from < to <=
       new  918eb56   Merge remote-tracking branch 'bdholt1/enh/tree' into bdholt1-enh-tree
       new  c9d142c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into enh/tree
       new  401aef4   Added tree module to whats_new
       new  dd9de04   release sv_coef memory
       new  051a4bf   ENH: add doc target to Makefile
       new  be48c63   Merge pull request #377 from larsmans/sparse-nmf
       new  8206495   tree construction depends on n_features
       new  3ce1d69   Merge pull request #383 from bdholt1/svm-mem-leak2
       new  fb7a048   Add Brian's name to the Changelog.
       new  92815c1   STY: imports in covariance + pep8
       new  fb69f69   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  1ac53fc   pep8
       new  1aea8d2   pep8
       new  bcecd31   OPTIM: inplace max in distances computation
       new  ab882df   OPTIM: avoid unnecessary repeted memory allocations in minibatch k-means
       new  d6b9138   Merge remote-tracking branch 'bdholt1/enh/tree' into bdholt1-enh-tree
       new  c41fe29   cosmit: pep8 and trailing spaces
       new  a5b8a84   merge master
       new  4873dd6   DOC: fix broken links + various cosmits
       new  0f3d93f   Merge pull request #22 from ogrisel/bdholt1-enh-tree
       new  ddcde05   Merge commit '0.9' (theirs) into releases
       new  543bc78   FIX: keywords {precompute, Xy} where implemented and documented but unused ...
       new  bc9ad28   Merge pull request #364 from robertlayton/silhouette
       new  6fad3d5   FIX: remove non-ASCII char from silhouette docstrigs
       new  8dfe631   Merge branch 'master' into tree
       new  ed8ebee   Minor cosmit
       new  f202342   Updated changelog.
       new  bf68b53   Some clarification of the memory copy issues.
       new  1cff3d3   Added a fix to prevent tree splits on samples that are not in the sample mask. The variable "a" should not necessarily be zero when it is passed into smalles_sample_larger_than(...).
       new  0610e2b   Added person webpage
       new  365dadd   added trailing underscore
       new  ae49d5b   Removed exception from _find_best_split to avoid code bloat.
       new  c49789b   Merge branch 'master' of github.com:scikit-learn/scikit-learn into enh/ensemble
       new  0f4b1f0   Removed unnecessary variables
       new  4a76a34   Merge branch 'fix-split-sample-mask' of https://github.com/TimSC/scikit-learn into TimSC-fix-split-sample-mask
       new  70316d3   Refactor ensemble learning code
       new  afc4a03   Merge pull request #23 from larsmans/enh/ensemble
       new  dd4613a   compute threshold as t = low + (high - low) / 2.0
       new  52555de   scikits.learn -> sklearn
       new  a1c29c9   update parameter names
       new  0b41a2d   Merge branch 'master' of github.com:scikit-learn/scikit-learn into enh/ensemble
       new  1853583   remove enforcement of return type
       new  8f5c2ef   DOC: use random_state in KMeans
       new  9323a22   DOC: improve silhouette coefficient docs
       new  ea59b84   MISC: better check_build error reporting
       new  8782723   MISC: even better check_build error reporting
       new  74ed8f4   OPTIM: inplace dense minibatch updates and better variable names
       new  b1b2cc1   cosmit
       new  87c8664   FIX Issue 379 and use the opportunity to refactor libsvm code
       new  efd6b06   cosmit: better variable name in MiniBatchKMeans
       new  f6c3969   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4e78abd   ENH: make it possible to control the add variance caused by Randomized SVD
       new  fd425f6   Added a numerical stability test to decision trees
       new  8a2c166   Added a numerical stability test to decision trees
       new  97f0c3a   Revert "Added a numerical stability test to decision trees"
       new  b9087e5   ENH: document clustering example simplification
       new  d3b4e4f   Merge remote-tracking branch 'upstream/master'
       new  4964ae7   FIX broken doctests on buildbot + pep257
       new  8878774   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  3f4664d   PEP08 names in graph_shortest_path
       new  c55c1f4   add warning flag to balltree + tests
       new  7bf8a6b   warning_flag doc
       new  bc563c2   add warning messages to KNeighbors
       new  1ccb8eb   fixes for tests
       new  755ba4f   attempt to address warnings catcher
       new  5d1d437   hack to fix warning test
       new  f1d981b   change warning message
       new  3d93ce8   simplify warning test; remove assert_warns from utils
       new  86a78a5   COSMIT
       new  61af784   first stab at nearest center in cython (+30% perf, need check correctness)
       new  6ade898   factorized label assignement as a reusable python func for the predict method
       new  ff9bd5b   use direct blas ddot call and reuse _assign_labels in predict
       new  85d6de9   Merge remote-tracking branch 'upstream/master'
       new  fe0ab94   DOC copy-edit naive bayes doc, with an emphasis on the formulas
       new  2b04cff   FIX: broken test cause by the use of todense which return a matrix instance instead of a regular numpy array
       new  b5cb57d   COSMIT in chi² feature selection
       new  2f7e997   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  0ac341a   added support for sparse KNN graphs and tests
       new  5e7d3c9   finishing up sparse additions (need to complete todo)
       new  8d65e48   DOC ported latexpdf target from Sphinx 1.0.7-generated Makefile
       new  24ac2f5   sparse KNN graphs now work
       new  9aabbdb   WIP on simpler cython impl of the center update (still buggy)
       new  ae445be   DOC: added Gaussian Processes to class reference
       new  aa13538   compute inertia + remove code :)
       new  6cd6c30   update renamed function call
       new  2eb9833   Trying to fix NaN errors, but its not working. Pushing to work on it later.
       new  f366d52   DOC typos in Ward tree docstring
       new  fbad80e   replaced ratio r with sampling with replacement
       new  87a6f5b   factorize dot product and bootstrap implementation for the dense case
       new  b7fe3bc   use cpdef + less array overhead in ddot
       new  5576ecf   started kmeans test suite refactoring
       new  c3d3906   Mutual information now works (tested!)
       new  6168d9c   more code factorization
       new  b2a8956   refactored the kmeans tests
       new  5f8d554   test and fix input checks for various dypes
       new  6d1dda8   much cheaper yet stable stopping criterion for the minibatch kmeans
       new  c2696a0   Fixed bug in updating structure matrix in ward_tree algorithm.
       new  bda7451   Added test case that reproduces crashes in old version of ward_tree algorithm.
       new  18f6e62   Performance tweaking in ward_tree.
       new  0cd593f   TEST: simplify test case
       new  4a23653   SPEED tree: 2X in Gini criteria
       new  8b2ca2a   Cosmetic changes in LARS
       new  fe4c258   FIX: Py3k compatibility.
       new  e0c217d   FIX: missing relative import marker
       new  d9acd6c   COSMIT little things in hierarchical.py
       new  0ecd9fc   Merge branch 'releases' into dfsg
       new  ec7c63f   Merge branch 'dfsg' into debian
       new  7d2c23a   removed deprecated up_workaround_numpy_cython_issue589652
       new  3a4d5a0   Refreshed deb_use_system_joblib patch to reflect renaming of the module
       new  8b80b52   added patch up_release_sv_coef_memory to "cherry-pick" 59006f248f24b22f7a8a21ada85c8558b5a1d1b6 (release sv_coef memory)
       new  c2c2403   First wave of changes for scikits.learn -> sklearn transition
       new  ed59b51   Initial changelog for 0.9.0-1
       new  4b77953   updating deb_use_system_joblib again
       new  bf3c61b   more on scikits.learn -> sklearn -- running tests
       new  527ed34   Merge remote-tracking branch 'upstream/master'
       new  6bccd7d   DOC: Added load_boston in classes.rst
       new  4a3975f   BUG NMF topic spotting example would output n_top_words-1 terms
       new  7465296   DOCS: Image is aligned to the right...
       new  395c2e8   Merge pull request #400 from amueller/docs_typo
       new  8eaaa75   Enabled all unittests and ignore failures on doctests: https://github.com/scikit-learn/scikit-learn/issues/401
       new  2ad830f   to please recent gbp -- we are building from the branch, not tags
       new  c4bb0b4   BF: fixing movement of scikits.learn into compat package
       new  b88e21b   added Enhances field to reference upcoming mvpa2 and mdp
       new  deec957   Adjusted debian/copyright to be fresh DEP5-compliant (+minor reformatting of changelog)
       new  5308cad   Removed ipython from build-depends and suggested it for the binary
       new  bfd36aa   fixing bashism in custom 'move around' loop
       new  9f05c11   AMI now works, and has been tested against the matlab code (test based on this to come!)
       new  e473f70   Remove phantom double v-measure !?
       new  2d8677f   Added tests. There are two errors, but I'm going to bed. I'll fix them in the morning.
       new  96746bc   DOC explain multiclass behavior in LogisticRegression
       new  10d78a7   DOC: LogisticRegression is a wrapper for liblinear.
       new  3c02ca8   MISC: mk roc_curve work on lists
       new  ff50c21   FIX: keep track of index swapping in OMP
       new  50249f6   Merge branch 'master' into omp_bug
       new  296909f   pacify lintian and add ${python:Depends} for python-scikits-learn
       new  f7a245a   DOC Added documentation for important attributes of GridSearchCV
       new  1a9319b   specify dict type
       new  5a8c45f   Merge pull request #404 from amueller/grid_search_doc
       new  45d05ab   COSMIT pep8 feature_extraction.text
       new  7451033   DOC some stuff on input validation
       new  8b3c656   Merge branch 'omp-bug-test' into omp_bug
       new  a316327   Testing for swapped regressors in OMP
       new  a7fda94   Merge branch 'omp-bug-test' into omp_bug
       new  0d42b51   DOCS: Typo in url
       new  7c5f6c0   Merge pull request #407 from amueller/sgd_url_typo
       new  0240e19   PEP8
       new  1718399   Merge branch 'master' into omp_bug
       new  797041a   ENH Cython version of SVMlight loader
       new  af91582   Merge branch 'master' into ami
       new  770020f   Merge branch 'ami' of github.com:robertlayton/scikit-learn into ami
       new  193ac02   - AMI in the cluster examples - See Also sections updated - mutual_information -> mutual_information_score (and updating subsequent imports)
       new  867ec2f   Higher level import for ami_score
       new  18b9a78   There is an overflow problem. It can be reproduced with the plot_adjusted_for_chance_measures.py example Commiting to get help on error
       new  a509d4c   Merge pull request #408 from vene/omp_bug
       new  aaf5c23   Narrative doc, and I think I fixed the overflow issue (more tests to come)
       new  fb0c3f8   Fixed logs to match the matlab code results. This fixes the tests (which needed a bit of updating)
       new  db845fa   Test now tests a much larger array
       new  77ec530   Test actually does what I meant it to do, and works sufficiently
       new  52fc4c5   Fixed this example. Tested the others (they worked!)
       new  c021b59   pep8 and pyflakes
       new  3bde2bf   ENH accept matrix input throughout
       new  51d4310   COSMIT rename safe_asanyarray to safe_asarray to prevent confusion
       new  4eebf9b   fix: permutation test score averages across folds
       new  afefe0b   initial checkin of gradient boosting
       new  210a11c   GBRT benchmark from ELSII Example 10.2
       new  eb8a2c9   added GBRT regressor + classifier classes; added shrinkage
       new  96fef35   use super in DecisionTree subclasses
       new  18bdc1c   first work on various loss functions for gradient boosting.
       new  62d9af7   added store_sample_mask flag to build_tree
       new  25ab085   implemented lad and binomial deviance - still a bug in binomial deviance -> mapping to {-1,1} or {0,1} ?
       new  2200d97   updated benchmark script for gbrt.
       new  d075567   some debug stmts
       new  0de4ab6   new benchmarks for gbrt classification
       new  900d0b5   fix: MSE criterion was wrong (don't weight variance!)
       new  6bc49ce   more benchmarks
       new  55fcd10   binomial deviance now works!!!!!
       new  a6a59f4   add gradient boosting to covtype benchmark
       new  9168add   add documentation to GB
       new  39da308   timeit stmts in boosting procedure.
       new  b53bac3   add previously rm c code
       new  e4318a8   updated tree
       new  c00d86f   hopefully the last bugfix in MSE
       new  014f1a1   new params in gbrt benchmark and comment out debug output
       new  6e370b7   make Node an extension type + change class label indexing.
       new  7ce5163   predict_proba now returns an array w/ as many cols as classes.
       new  aac158a   cosmit: tidyed up RegressionCriterion
       new  8a57c4e   added VariableImportance visitor and variable_importance property
       new  aadc505   minor changes to benchmark scripts
       new  191fffb   use `np.take` if possible, added monitor object to `fit` method for algorithm introspection.
       new  1c669a4   cosmit
       new  01ae6b2   choose left branch if smaller or equal to threshold; add epsilon to find_larger_than.
       new  9359b98   compiled changes for last commit
       new  92c773a   cosmit
       new  638a65e   some tweaks and debug msg in tree to spot numerical difficulties.
       new  eccc0f0   added TimSC tree fix
       new  73682fa   changed from node.error to node.initial_error in graphviz exporter
       new  5ce1739   recompiled cython code after rebase
       new  7ffa230   fix: _tree.Node
       new  a4e4203   comment out HuberLoss and comment in benchmarks
       new  d517ced   STY: pep8 + naming
       new  cbab862   DOC: prettify plot_permutation_test_for_classification.py
       new  5c18598   DOC : adding permutation_test_score to changelog
       new  0dfb8f0   DOC correct Google URL
       new  c06ebe7   MISC: __version__ in scikits.learn
       new  15b4047   ENH: Adds more verbosity to grid_search. verbose > 2 gives scores while running grid.
       new  fdffade   Merge pull request #414 from amueller/grid_search_verbosity
       new  aa69d85   FIX #401: update tutorial doctests to reflect recent changes and add them to
       new  bb46a32   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  92e3ec2   DOC: new scikit-learn.org URLs and mention license in README.md
       new  3052291   Merge remote-tracking branch 'robertlayton/ami' into robertlayton-ami
       new  b7d7642   measure runtimes for various clustering metrics in adjusted for chance example
       new  3845fd8   FIX warnings by avoiding 0.0 values in the log + cosmit
       new  b89b6d0   Merge pull request #3 from ogrisel/robertlayton-ami
       new  77ff098   Optimising the expected mutual information code Still a long way to go, but I've halved the time from approx 30 seconds to 16 seconds for an example
       new  6c177a1   pep8 grid_search.py
       new  4ab62e7   changed from y in {-1,1} to {0,1}
       new  5fe9b26   cosmit: beautified RegressionCriterion (sum and sq_sum instead of mean).
       new  d63f39c   rename node.sample_mask to node.terminal_region renamed BernoulliLoss to BinomialLoss and rm old BinomialLoss
       new  a14778e   Merge branch 'master' into minibatch-kmeans-optim
       new  634cf71   Delete benchmarks/bench_svm.py
       new  26b5896   Delette benchmarks/bench_neighbors.py
       new  71d5243   FIX replace np.atleast_2d with new utils.array2d
       new  59f1970   DOC correct and clean up empirical covariance docstrings
       new  1f2549d   ENH test input validation code on memmap arrays
       new  0c9cb49   Merge pull request #410 from larsmans/accept-matrix-input
       new  a03b104   ENH sample_weight argument in discrete NB estimators
       new  c6aaef8   MISC: More meaningful names for lapack functions in least_angle.
       new  b5429ab   Removed unused parameters in least_angle
       new  f161fab   BUG handle two-class multilabel case in LabelBinarizer
       new  7367155   TEST better test for binary multilabel case in LabelBinarizer
       new  de280c8   fix: Node.__reduce__
       new  c68a368   unused import
       new  3ef20fe   fix init predictor for binomial loss
       new  96daa4f   ENH multilabel learning in OneVsRestClassifier
       new  82696d5   BUG: Use threshold in LabelBinarizer in multi-label case.
       new  894747f   ENH support decision_function in multi-label classification
       new  025cf80   DOC OneVsRestClassifier multilabel stuff
       new  6f232b9   ENH multilabel support in SVMlight loader
       new  688d2dd   DOC multilabel classification in narrative docs
       new  3fa52bf   FIX Python 2.5 compat in utils/tests
       new  94f98dc   Cosmit: used named parameter.
       new  8a487b6   ENH Label indicator matrix support in LabelBinarizer and OVRClassifier
       new  498f293   COSMIT multiclass.predict_ovr
       new  b260880   DOC expand Naive Bayes narrative doc (BernoulliNB formula)
       new  f997740   DOC: add IterGrid in reference
       new  55c20cb   performance enh: update predictions during update_terminal_regions
       new  2e42f89   COSMIT in naive_bayes
       new  13a48df   Adding old version of EMI, as I'm about to change it
       new  baae4df   fix: samplemask cosmit: pep8
       new  983556e   low memory computation of the square diff
       new  054b682   be more consistent with the usual behavior of fitted attributes
       new  b453e1e   ENH prevent copy in sparse.LogisticRegression
       new  e7c02a3   base convergence detection on EWA inertia monitoring
       new  e6d5ed0   Revert "ENH prevent copy in sparse.LogisticRegression"
       new  0bfc957   DOC: minor typo "precom[p]uted"
       new  8ea63e9   COSMIT: no import as
       new  942b354   Merge remote-tracking branch 'upstream/master'
       new  8170fb5   various cython cleanups
       new  744072e   working in progress to make it possible to use a speedy version based on smoothed inertial only
       new  8c65edf   ENH: more informative error messages when input has invalid shapes
       new  95d497f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  eab5877   Simplified tree module API.
       new  1c00a45   Added some comments
       new  40afcd9   ENH: more informative error message when shape mismatch in TF IDF transformer
       new  76e6197   merge master
       new  e8ddec5   preparing new stopping criterion impl
       new  1e4e601   ENH: make it possible to pass class_weight='auto' as constructor param for SGDClassifier
       new  562fcae   Merge branch 'master' into minibatch-kmeans-optim
       new  28fc441   Allow for max_depth to be set to None
       new  7813be0   work in progress (broken tests) on early stopping with both tol and inertia lack of improvement
       new  a753733   MISC: Warn for integers in scaling/normalize
       new  1a1b9f4   MISC: better warning message
       new  ccc3cad   This version doesn't work either. I am uploading for historical sake. I'll be undoing these changes with the gammaln function to see how that goes.
       new  e13b3f0   make min_dist test more explicit
       new  77c9663   fixed broken test
       new  fb6fe36   Initial usage of gammaln. Not yet tested
       new  a786bdc   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into ami
       new  ba52695   Simplified the tree code
       new  c50067b   Added k_features argument to build randomized trees.
       new  7769275   DOC: Document "cache size" argument of SVR
       new  8f602bc   COSMIT: remove unused error string.
       new  b991096   First draft at find_best_random_split (not yet tested)
       new  e072eb0   COSMIT: remove unused error string.
       new  5bf49f9   Renamed k_features to max_features
       new  8735632   Added some explanatory comments into the code logic
       new  d17fba0   optimize label assignment for dense minibatch and new test
       new  a2d136f   fix tests
       new  c10964d   fix tests
       new  8579119   start with zero counts in tests
       new  d117ec5   fix bug: x_squared_norms should follow the shuffle...
       new  77db343   ensure that the sparse and dense variant of the minibatch update compute the same thing
       new  82c3f62   better default value and parameter handling for max_no_improvement
       new  a0f2598   switch to lazy sampling with explicit index to divide memory usage almost by 2 and decrease code complexity with no measurable impact on the run time
       new  28b4d88   more code simplification
       new  9943f40   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  895094d   ENH: removed kernel cache from fit method of DenseLibSVM, added to __init__ of BaseLibSVM
       new  8505e23   Added kernel cache argument to init of all SVC and SVR classes. For the moment the conservative 100MB default.
       new  82561e6   BUG cache_size instead of cache as paramter name
       new  d961d79   BUG: cache_size also for sparse SVMs
       new  627bf41   DOC typos and style in linear_model docs
       new  25782ef   added timing info
       new  da847fb   cosmit: get rid of gcc warning (q_data_ptr was not initialized)
       new  b654278   fix: overflow of `offset` variable if X.shape[0] * X.shape[1] > 250M
       new  aea7b5a   COSMIT: never use np.linalg, but scipy.linalg
       new  4170b24   Re-extended the _build_tree API
       new  39c086d   Factored is_classification
       new  e35d7ea   Added ExtraTreeClassifier and ExtraTreeRegressor
       new  b4e7ce5   Typo
       new  cb3eff0   BUG: ProbabilisticPCA.score work with pipeline
       new  928fb05   First draft at forest of random trees (work in progress)
       new  d812573   Convert to scipy doc convention + add missing options
       new  94f67f0   Added some tests
       new  b4faf00   Cosmit
       new  4e74650   ENH: SVM cache_size default value changed to 200 mb
       new  8878ebe   ENH Sparse SVM: removed cache_size parameter from fit method. Is now part of constructur.
       new  60e4c80   DOC fixed doctests for cache_size parameter
       new  776b534   DOC slight reformatting of kernel cache note in module docs.
       new  3e04da6   Fixed bugs in forest + first test
       new  07b4121   BUG: minor mistake in earlier commit.
       new  bdc69b6   DOC: fogot doctests in python files.
       new  b73fd80   Check X is a fortran-array and y is contiguous
       new  87f4cf6   DOC: another doctest.
       new  711345a   Fixed bugs
       new  73178f6   Added tests of the forest module (work in progress)
       new  9b62d90   Default value of n_trees=10
       new  62b9bdf   bootstrap=False for extra-trees
       new  f220d9e   Set random_state=1 in tests
       new  0fa97be   Added documentation in the forest module (work in progress)
       new  5b6a88a   Cosmit
       new  669ecbe   COSMIT cleanup sgd Cython code
       new  75bf5b0   Still overflows, but the closest so far. Using gammaln
       new  01ae29e   It works! Still have some optimisation to do, but it works for larger arrays
       new  10a21f7   ENH: in Scaler, warn if fit or transform called with integer data.
       new  e625ba4   Merge pull request #425 from amueller/svm_cache_size
       new  9224538   started example to check the convergence stability in various settings
       new  32fb601   FIX: buggy usage of for / else for k-means n_init loop
       new  231057b   DOC: update what's new
       new  161430c   tracking changes from master
       new  433a381   FIX: broken HMM tests caused by KMeans convergence in one step
       new  113d394   merge master
       new  0e9b771   ENH parameter "epsilon" in SVR error messages is given correct name.
       new  033e657   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  87a28cd   ENH: use integer indexing instead of boolean masks by default for CV
       new  f407372   Moved start and finish outside of loop
       new  37df796   implemented n_init for MiniBatchKMeans
       new  4f9f32c   Merge branch 'master' into minibatch-kmeans-optim
       new  48bb4f3   DOC update cross validation docstrings for default indices=True
       new  3d58c49   refactored the init logic for MiniBatchKMeans
       new  a68c85f   Merge branch 'master' into minibatch-kmeans-optim
       new  8858a80   fix stability and warning in tests
       new  9fbe559   make k-means++ work on sparse input and use it as default for MB k-means
       new  768e471   add version info in deprecation message
       new  6af7996   factorized out the early stopping logic in a dedicated method
       new  5645473   MISC: remove links to sourceforge URL
       new  822b4ee   BUG handle broken estimators in grid search by cloning them
       new  e3dc529   DOC: fix links in mixture
       new  982359c   first stab at a reinit strategy that work on low dim data only
       new  7e02960   MISC: add citation information
       new  f87248d   new example to emphasize issues with current naive reinit scheme on sparse data
       new  8118f6e   BUG: vectorizer.inverse_transform on arrays
       new  3f1901c   second experiment on reinit that does not work on high dim sparse data either
       new  951b997   Completed documentation
       new  5df1a83   ENH don't require numeric class labels in SGDClassifier
       new  88adf3b   comments, pep8 and pyflakes
       new  9f6ca7f   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into ami
       new  ce7e009   Added some tests
       new  f0b519d   Added predict_log_proba
       new  2ab37e5   Added some more tests
       new  9e8b973   Removed old random forest files
       new  2fa1052   ami_score -> adjusted_mutual_info_score Warning about slowness of function
       new  e51e4d6   ami_score -> adjusted_mutual_info_score This time, in examples and docs!
       new  8ac39e0   Added some more tests
       new  38f1309   Cosmit
       new  19f813e   Regenerate _tree.c
       new  e0ec089   BUG fix SGD doctests
       new  a35936d   Fixed a small bug
       new  b23a53e   Cosmit
       new  021ae0e   FIX: array2d was did not return contiguous arrays with order='C' ...
       new  09677e0   FIX: do not use reshape in libsvm sparse bindings.
       new  8686d40   * doc/conf.py: added required latex packages (bm and morefloats) * doc/images/iris.png: added image based on iris.svg * doc/modules/tree.rst: changed reference to iris image
       new  4ad73b5   DOC: pdf compilation
       new  954b098   bug: mode='LM' -> mode='LA'
       new  a42ed3c   BUG fix Naive Bayes test + refactor module
       new  7375d50   DOC typo
       new  66f7a0b   "What's new?" AMI!
       new  7184e39   Use super()
       new  362a986   Use take instead of __get_item__
       new  feeee90   Rewrote some comments
       new  030afba   Cosmit
       new  2be7b0e   Revert changes on conf.py (mistake on my part)
       new  f0d40dc   Added random_state parameter to _find_split functions
       new  f4b80fa   Factored out changes on the ensemble module
       new  b7df59f   Merge remote-tracking branch 'origin/master' into tree
       new  7e7da4e   Fixing conflicts
       new  ae6c76f   Merge remote-tracking branch 'upstream/master'
       new  81bff12   ENH: Easier debugging in check_build
       new  f20c0d0   ENH check_build: better error msg for local imports
       new  beace0d   DOC Made reference to "Getting started" in "Datasets" section a link.
       new  dcb38d7   DOC: inline example for precomputed svm kernel
       new  3485737   Removed extra-trees (for now)
       new  5975f4c   Removed extra-trees from __init__
       new  116de7e   Removed extra-trees (again!)
       new  85aee9d   Merge pull request #432 from glouppe/tree
       new  1657df4   fix: broken doctest with precomputed kernel
       new  b6f4de5   changed Decision Tree representation to struct of arrays instead of composite structure.
       new  729cb50   fix: use tree.predict instead of functor
       new  4916ec2   Graphviz visitor now works on array repr. each node now has best_error and init_error; best error is the error of the split, init error is the error of the node. For leaves we compute the error using a helper function and best_error == init_error.
       new  1c7f089   cosmit: doc strings
       new  d6cd84d   DOC: turn off generation of index pages
       new  b9685ff   Merge branch 'ami' of https://github.com/robertlayton/scikit-learn into ami
       new  f26ed76   mutual_information_score -> mutual_info_score
       new  17ee6c4   and in plot_adjusted example (mutual_info_score)
       new  ff0dbf9   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into ami
       new  83c7445   Use centralized directory for generated files.
       new  4eb4caa   Merge remote-tracking branch 'upstream/master'
       new  f36ec61   ENH in preprocessing.Scaler, raise warning also if given unsigned int
       new  118e8bd   cosmit
       new  70b188b   ENH support array-like y (lists, tuples) in GridSearchCV
       new  2e1c5e8   ENH: Capture stdout in executed examples
       new  d96b154   ENH support arbitrary labels in metrics module
       new  6263124   Merge pull request #402 from robertlayton/ami
       new  fbe8aed   COSMIT: layout in plot_kmeans_digits example
       new  f45a269   DOC: minor fix to AMI docs
       new  f2697cb   ENH: First sketch of glasso
       new  519ab77   ENH: example for l1 covariance estimator
       new  65fa358   ENH: Add cd solver to glasso
       new  3d1d982   COSMIT glasso: docstring and cleanup
       new  b9e6d54   ENH: the GLasso estimator
       new  0a29d06   DOC: Better glasso example
       new  3dd76fd   TEST: test GLasso
       new  53d49bd   ENH Glasso: don't penalize the diagonal
       new  71d76eb   ENH: Add a GLassoCV
       new  a01bd34   ENH GLassoCV: iteratively-refined Grid search
       new  bc9176b   ENH GLasso: stability on correlated data
       new  db60df4   ENH GLassoCV: better parameter optimization
       new  6bf2bdc   TEST GLasso: increase test coverage
       new  17e2462   DOC: narrative documentation for GLasso
       new  46fdc66   COSMIT: @agramfort's comments
       new  b349812   DOC: add sparse inverse covariance in whats_new
       new  b00504d   PEP8
       new  17a1d5f   DOC: rmks on structure recovery
       new  da97008   DOC: better stock_market example (WIP)
       new  17361c7   COSMIT: address most of @ogrisel's comments
       new  779bc9a   ENH: don't echo convergence warning on CV grid
       new  84f0d1b   DOC GraphLasso: be explicit about which algorithm
       new  aae84b3   DOC GraphLasso: notes on algorithms and recovery
       new  f6d28c3   DOC: docstring in stock market example
       new  dfd4689   DOC/API: integrate make_sparse_spd_matrix
       new  218767c   PEP8 + various cosmits
       new  57cab00   pep8 in sparse covariance example
       new  b7e2491   PEP8 + PEP257 in samples_generator
       new  bdaf459   Typo
       new  825d033   MISC: address @larsman's comments
       new  1668a92   API: g_lasso.py -> graph_lasso_.py
       new  b7fa45f   DOC: GLasso -> GraphLasso
       new  04ae951   MISC: @VirgileFritsch and @mblondel's comments
       new  4716c25   MISC: silence stdout in GraphLassoCV tests
       new  e951c37   PEP257 - docstring style
       new  a5007b5   ENH GraphLasso: Silence warning
       new  9eec795   ENH: graph_lasso works on empirical covariance
       new  cfeab0e   BUG: update tests to changes in graph_lasso
       new  ac803a5   Fixed values in Adjusted Mutual Information doctests
       new  a8e45bd   use safe_sparse_dot instead of np.dot
       new  2782c53   BUG: fix layout in examples
       new  91d0333   MISC: fix rst bug
       new  e108215   DOC: put class reference in the banner
       new  47fd425   DOC/Website: Changed link on "support" page to scikit-learn.org, added 0.9 release doc link
       new  dba2538   COSMIT: prettify plot_oneclass
       new  d363d08   DOC: rework front page
       new  4fa9b07   DOC: Add 'up' relative link
       new  1689ecb   changed int64 to int32 in tree repr; cosmit in _tree.pyx
       new  4007610   Merge branch 'tree-array-repr'
       new  86801ba   changed for `for i in 0 <= i < n` to `for i in xrange(n)`.
       new  69803e0   Merge branch 'tree-array-repr'
       new  6def64c   ENH : adding support for scaling by n_samples of the C parameter in libsvm and liblinear models
       new  6cf3040   COSMIT rm comment in coord descent code about np.dot
       new  0f15066   COSMIT no need for csr_matrix "cast" in coord descent
       new  541528c   Remove C from NuSVR.
       new  cbeb674   DOC: title for the user guide content file
       new  329e89d   DOC: don't display empty tocs
       new  3db4c67   MISC: scikits.learn -> sklearn
       new  0961424   DOC: proper link structure in examples
       new  7897b8c   DOC: title to relative links
       new  ad9f7b5   FIX : removing param nu from sparse.SVR, C from NuSVR + pep8
       new  5504034   $Merge branch 'master' into n_samples_scaling
       new  71aa6f3   typo
       new  fdbd584   s/C_scale_n_samples/scale_C
       new  8fb7c84   Merge remote-tracking branch 'upstream/master'
       new  7bd3f88   Rebase of @bdholt1's ensemble branch
       new  dfd924c   DOC: Added module descriptions
       new  46abcb0   PEP8: tree.py, forest.py
       new  e980d3f   DOC: EPD ships a recent version, but not latest
       new  f9161f6   changed tree.left and tree.right to tree.children (similar to cluster.hierachical)
       new  75faece   DOC: state clearly the version number
       new  2e55ca3   Merge remote-tracking branch 'upstream/master' into ensemble-rebased
       new  fca7dda   MISC: plot_stock_market cluster on learned covariance
       new  08b5069   DOC: Added warning and see also
       new  1bae431   DOC fixed whitespace in GridSearchCV doc string so that html doc is generated correctly.
       new  16ab921   ENH: Modified forest API to make it possible to grid-search the parameters of the underlying trees
       new  4b8619b   use new tree repr; adapt gradient boosting for new tree repr.
       new  136d2ba   Merge branch 'master' into gradient_boosting
       new  e2fa33d   cythonized tree (still broken)
       new  b0d5f93   clear tree.py
       new  44b778b   updated _tree.c
       new  492442c   updated GradientBoosting with current master
       new  d9aa128   Merge branch 'master' into minibatch-kmeans-optim
       new  0a3d896   fix: update variable importance
       new  a8c7d60   l1_distances was renamed to manhattan_distances.
       new  0d54dfa   added gradient boosting regression example
       new  1569b73   added test deviance to GBRT example
       new  090037f   added avg_f1_score
       new  e546621   tst: added tests
       new  df6010b   enh: added matthew's correlation coefficient
       new  d7fe7a9   updated TODO in module doc
       new  baf69e6   BF: index out of bound in GraphLassoCV grid refinement.
       new  ee5451e   sty: pep8 + doc
       new  8870d53   ENH prevent copy in PCA if not necessary
       new  a63931e   Merge remote-tracking branch 'upstream/master' into ensemble-rebased
       new  15f9522   ENH: Check that base_tree is an estimator
       new  87cf1b3   fix: sgd module clone issue w/ rho parameter
       new  fe55429   Merge remote branch 'upstream/master'
       new  fee7e8a   FIX use super consistently in SVMs
       new  fb07d61   Description for logo: font, color, etc.
       new  8de628e   ENH: Make forest derive from BaseEnsemble
       new  02f2751   Removed Bagging and Boosting modules from this PR
       new  a06504b   ENH: Make the Forest's API coherent with BaseEnsemble's API
       new  f117f31   FIX: Don't clone estimators at instantiation
       new  fa3dc81   TEST: Added test case for grid-searching over the base tree parameters
       new  b637ea1   ENH: Cosmit
       new  21288e9   EXAMPLES: Improved plot_tree_regression
       new  a8030a7   Typo
       new  7df65e5   EXAMPLES: Improved plot_iris
       new  1fb6dde   EXAMPLES: Added plot_forest_iris
       new  58eeb71   FIX: Trees couldn't be cloned properly
       new  199bf94   ENH: Added __init__.py into ensemble/tests/
       new  303b554   BUG: fix score() with GraphLasso
       new  892dd6b   Compatibility with numpy 1.1
       new  cdb15d3   DOC: Improved documentation in the examples
       new  95c159e   PEP8
       new  5bf4b8c   TEST: Added tests of BaseEnsemble
       new  d434ed4   TEST: Improved test coverage
       new  8829cab   EXAMPLES: Fixed a bug in plot_forest_iris
       new  ca63fd1   Skip tests in OMP that fail on old Python versions
       new  a86b444   Fix one-dimensional y in Gram OMP estimator
       new  82e3b17   BUG GraphLassoCV: score() needs a store_precision attribute
       new  dee14ae   Merge branch 'master' into gradient_boosting
       new  48fda63   ENH incrementally build arrays in SVMlight loader to reduce memory usage
       new  21932a0   DOC: Cosmitis in the narrative documentation of the tree module
       new  742c427   DOC: Improved narrative documentation of the tree module
       new  ffd9c04   DOC: Added ensemble methods to TOC
       new  62d79fa   DOC: Added ensemble methods to the class reference
       new  6b7a9ce   Merge pull request #446 from larsmans/svmlight-loader-memory-use
       new  09701e4   DOC: First draft at the narrative documentation of the ensemble module
       new  e300668   DOC: Narrative doc of the ensemble module (work in progress)
       new  fd38239   DOC: Completed the narrative documentation (work in progress) + What's new
       new  8d30a9d   DOC: Fixed What's new
       new  0d664de   DOC: Last details on the narrative documentation
       new  f071368   DOC: Added a last example in the narrative doc
       new  b277f54   FIX: make the doctests outcome deterministic
       new  ccc860f   DOC: better toplevel docstring
       new  902dbdc   DOC: add simple descriptions in the concrete class docstrings
       new  50d8ac2   FIX: workaround what looks like a numerical instability in doctest
       new  c3cd700   Merge pull request #1 from ogrisel/glouppe-ensemble-rebased
       new  820d5d0   DOC: Address @vene and @satra comments
       new  cdc5954   TEST: Added test_base_estimator
       new  b83882a   DOC: Cosmit
       new  a0a876e   COSMIT removed unused import, pep8
       new  d137da1   COSMIT pep8 in cluster module, removed unused import
       new  ff3f814   COSMIT pep8 whitespace
       new  772d338   COSMIT removed emacs modeline
       new  638ba28   COSMIT: pep8 whitespaces instead of aligned decimal points
       new  6524496   COSMIT indentation
       new  88dc827   COSMIT ugly line break for pep8
       new  911eb2b   COSMIT reindented for pep8
       new  77495c7   COSMIT pep8 whitespaces
       new  c6e6893   Merge pull request #447 from amueller/pep8
       new  ce57ec4   DOC: Move practical info into its section and delete duplicates.
       new  6fd239d   Style: webpage tweaks
       new  63887c2   Style update in documentation.
       new  6d5cb36   Doc: minor fixes
       new  72c5f01   Minor update and fixes to linear_model documentation
       new  7bd3d84   Minor update and fixes to linear_model documentation
       new  27f89c7   Move implementation details into RST doc.
       new  306985a   ENH: Simplified RandomForest and ExtraTrees API
       new  868d4af   Docstring conventions.
       new  6514438   DOC: rename n -> p
       new  63005c0   Web page layout tweaks.
       new  d7f41e4   ENH: Use trailing _ for private attributes
       new  3ffbe9c   DOC: Added warning in make_estimator
       new  8c1426f   DOC: Removed 'default'
       new  2d057f7   Revert "Remove C from NuSVR."
       new  16f628e   Revert "FIX : removing param nu from sparse.SVR, C from NuSVR + pep8"
       new  9b587f5   FIX: Bug with bootstrapping
       new  101b75c   FIX: Bug with bootstrapping (2)
       new  574b272   FIX: Bug in plot_forest_iris
       new  4623214   Merge remote-tracking branch 'upstream/master'
       new  d5c5188   DOC: Use ELLIPSIS in doc-test
       new  344cca2   Cosmit
       new  f72634d   fix: make GradientBoostingBase clonable.
       new  6ebaede   fix: learning rate schedule doc.
       new  afcf12c   Merge remote branch 'upstream/master'
       new  bb5ddcb   fix: rm `nu` argument from sparse.SVR (taken from dense SVR).
       new  f6d7344   ENH: Address @agramfort comments
       new  aaeaa14   added unit tests for gradient boosting (coverage ~95%)
       new  a0c8414   Benchmark: Added random forests and extra-trees to bench_sgd_covertype.py
       new  62aa5d6   Merge pull request #439 from glouppe/ensemble-rebased
       new  54bc641   Merge branch 'master' into enh/metrics
       new  e6558db   Merge remote-tracking branch 'upstream/master'
       new  227770f   better test coverage
       new  87c5317   store loss object in estimator
       new  e0b78b9   don't use dict comprehensions (support python 2.5 and 2.6).
       new  cf1233e   fix: tree doctests + ensemble doctests
       new  7b13fc9   Merge branch 'master' into gradient_boosting
       new  6d90b2b   stub for gradient boosting documentation
       new  3b17eaf   STY: pep8 + pyflakes
       new  209ecb0   Merge remote-tracking branch 'upstream/master'
       new  d98e02d   FIX: Use random_state in _find_best_random_split
       new  af611cb   Merge remote-tracking branch 'upstream/master'
       new  d679e8d   ENH: in sgd classifier, check that parameter alpha is greater than zero
       new  404152d   Small comment on the dual parameter in LinearSVC.
       new  65155d8   restore original bench_tree.py
       new  a23f6e2   Merge branch 'master' into gradient_boosting
       new  a51424b   Small comment on the dual parameter
       new  81f3d12   Refactor MCD robust covariance estimator: it is easier to regularize.
       new  ae075ae   Merge with Gael's glasso changes.
       new  545a332   Make the design even more modular for MinCovDet.
       new  47fa3bd   Make the "robustness parameter" accessible through the API.
       new  c12fa6b   Integrate Gael's minor comments + Magnify examples + 1D data case.
       new  f43c524   Remove `correction` and `reweighting` parameters from the API.
       new  4a56f3e   Merge pull request #396 from VirgileFritsch/refactor_mcd
       new  1955ffd   COSMIT some pep8
       new  741dfd3   some pyflakes
       new  78fa279   COSMIT more pep8
       new  b62b97a   COSMIT more pyflakes
       new  b0d10fc   COSMIT: more pep8. enough for today...
       new  ae3b943   Re-ran the tests and found that the GaussianNB error was much lower.
       new  0074331   Update svmlight loader documentation.
       new  d02cfe4   Fixed typo
       new  510b1a1   Merge remote-tracking branch 'upstream/master'
       new  6709ed5   First draft at Reference rewrite
       new  ea7d943   DOC: "the scikit-learn" -> "scikit-learn"
       new  6d83701   DOC: References to user guide sections
       new  18fb39d   min_density now works with store_terminal_regions (however, this only matters if you learn deep trees max_depth >> 5 which rarely happens).
       new  c40dbf7   DOC: Standardize the module documentation format (work in progress)
       new  b25c574   DOC: Standardized the module documentation format (2)
       new  05dd1ec   DOC: Fixed graph_lasso reference
       new  aa5af20   DOC: "Class Reference" -> "Reference"
       new  3573dbc   DOC: Fixed warning
       new  f7e8c18   DOC: fix name for line_profiler_ext.py extension
       new  fbe6e97   Merge pull request #453 from yarikoptic/master
       new  c2e73e7   pep8
       new  4188212   Merge pull request #452 from glouppe/doc
       new  1519739   DOC: enhancement for Debian installation + fixed various typos
       new  b68226b   cosmit
       new  61c39f1   DOC: Changed sections titles in the reference
       new  560352c   DOC: restore 'This page' in sidebar
       new  dfa2d08   added input type and shape test
       new  9254379   fix: added ensemble to setup.
       new  5916017   Added SparseCoder estimator
       new  52bbc80   Basic testing
       new  2e3316a   DOC: add missing split_sign in docstrings
       new  9656faf   FIX: 10% of features should be at least 1
       new  2752c0f   DOC typos in ensemble.forest
       new  a7764a9   PEP257 :)
       new  c1beb20   restore typo
       new  6b018a6   Added SparseCoder to init and class index
       new  fced530   initial work on docs
       new  20ba4fa   implement noop fit in SparseCoder
       new  826df8d   clean up test
       new  7fba5f5   Fixed doc links
       new  e9a6567   Fixed lena in example
       new  28d9dce   Fixed lena import in denoising example
       new  b4fe032   Merge branch 'master' into sparse-coder
       new  3fac035   cleaned up imports in test
       new  e2bd616   Use M.dot instead of np.dot on sparse matrices
       new  2aeec69   FIX: LLE mode='auto' for small matrices and tuples.
       new  cc1ba3b   FIX: use .toarray() instead of .todense()
       new  9576c9c   COSMETIC: more readable syntax for mult. of sparse matrices.
       new  49e7b0d   OPT: (minor) remove useless determinant computation in FastMCD.
       new  1a0f575   drop Python 2.5; no more with statements from the __future__
       new  06de436   drop Python 2.5; no more need for utils.fixes.product
       new  106648f   drop Python 2.5; document and rm some workarounds for kwargs quirks
       new  f34db3d   PEP257 cosmit
       new  7d62369   Check for consistent input in Logistic Regression.
       new  75ceb43   COSMIT rm some SciPy pre-0.7 compat code
       new  98ead8a   raise TypeError instead of ValueError in check_arrays
       new  1f28be9   COSMIT docstring fix + US spelling in K-means code
       new  ac93562   cosmit
       new  70f9c9f   Update README.txt dependencies info to match the configuration tested on jenkins
       new  e0b6cd4   DOC I don't think Ubuntu 10.04 will be the last LTS release
       new  b88191b   test @deprecated using warnings.catch_warnings
       new  05006c0   COSMIT use utils.deprecated as a class decorator
       new  4c8b15e   don't use assert_in, not supported by nose on buildbot
       new  bc0c26e   Revert "FIX: more python2.5 SyntaxError"
       new  c7373e1   Revert "FIX: python2.5 SyntaxError"
       new  19705d4   COSMIT use urlretrieve and "with" syntax in LFW module
       new  e013191   COSMIT use ABCMeta in naive_bayes
       new  1ec01bc   COSMIT a few more easy cases of with open syntax
       new  8e802fe   rm Py2.5 compat factorial and combinations from utils.extmath
       new  3797a75   use cPickle in spectral clustering tests
       new  1de57fa   COSMIT use Python 2.6 except-as syntax
       new  582c431   Merge remote branch 'upstream/master' into gradient_boosting
       new  b1d269f   n_samples > min_split instead of >= no division through zero in predict_proba various cosmits
       new  528da65   cosmits (cleanup after profiling)
       new  2e37f0e   repeat decorator now with arguments
       new  f28ee57   modified 'gid' to 'git'
       new  5aa7fc9   Merge pull request #461 from Balu-Varanasi/bug_in_rst_file
       new  7017eb7   Fix svmlight loader doc.
       new  5af9a05   DOC rm Methods section from KMeans docstring
       new  a66c0e9   Add arXiv link to Halko et al. 2009 paper.
       new  3e78064   Merge pull request #463 from npinto/patch-2
       new  772bef6   FIX - error in the bibtex entry - extra comma that makes bibtex fail
       new  c450034   Merge pull request #464 from NelleV/FIX_bibtex
       new  3b5e6a6   Merge branch 'master' into enh/metrics
       new  940f34a   enh: added support for weighted metrics closes #83
       new  bbd8b71   doc: added description for matthew's corrcoef from wikipedia
       new  493e20e   sty: pep8 fixes
       new  e69c24d   sty: pep8 on test file
       new  8c052b4   doc: removed strange character
       new  584a2e6   fix: updated tests to reflect that micro shows the same precision and recall
       new  cfcf42e   Merge branch 'master' into n_samples_scaling
       new  d2b50d8   STY: prettify doctest
       new  3ade680   fix: average with elif
       new  9a6d934   doc: improved description of average
       new  a7afa5b   ENH : adding scale_C in NuSVR
       new  ef76dba   ENH: fastica returns whitening matrix "None" when whitening=False
       new  e8bc26a   TEST non-regression test for issue 238, FastICA failing with whiten="False"
       new  d0854d5   COSMIT pep8
       new  04db540   COSMIT pyflakes
       new  25c93f4   COSMIT: pep8
       new  f29a652   COSMIT pep8 in backported sparsetools...
       new  292da79   DOC Added Gael's explanations about the memory usage in grid_search / joblib to the docs (Hope I understood everything correctly).
       new  9af02a5   DOC: Auto example digit classification plot without interpolation and axis.
       new  2056428   FIX: typo in with statement
       new  07c7182   Example for random dataset function.
       new  2bd00c2   Random dataset example: make figure look nice on the web
       new  e9be1c8   DOC: Added random dataset plot to doc.
       new  289b0d7   COSMIT: random dataset plot prettified
       new  da16658   DOC Added comment about equivalence of nu-SVM and C-SVM to the docs
       new  57b4581   Examples: Replaced NuSVM by rbf SVM in example. RBF-SVMs are really important, NuSVMs not so much imho.
       new  c8afec0   pep8. whoops..
       new  e0f8a9d   COSMIT: pep8
       new  5dc49fc   FIX: Return "None" fist.
       new  8b0e18b   Example for finding the hyperparameters in a RBF SVM
       new  7fb3793   Examples: Make SVM parameter estimation look good on the web.
       new  ca3602d   DOC: Fixed legend in iris svm example
       new  488cbea   Merge pull request #466 from amueller/svm_iris_example
       new  36d0adb   pep8 compliant
       new  3578396   DOC Nonlinear SVM example changed to satisfy my sense of aesthetics. Hope you like it.
       new  a95ace9   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  5ad5207   COSMIT pep8
       new  034ca3f   Example illustrating parameters of an RBF SVM
       new  81cc312   COSMIT removed unused import math in utils/extmath.py
       new  b086f9f   FIX: make kmeans test not raise warning when init is passed.
       new  8bae14d   ENH : more contrasted colormap
       new  28e4918   MISC: typos + subplot adjust
       new  dd0ce5b   Merge pull request #467 from Balu-Varanasi/pep8-compliant
       new  dd6a56c   cosmit
       new  b9d0a76   MISC: update joblib
       new  6c4c2b1   BUG: fix joblib doctest
       new  37c904d   BUG: make the tests pass with numpy 2
       new  e385dae   COSMIT
       new  deb136c   FIX: make kmeans test not raise warning when init is passed.
       new  0343627   DOC Description of the basic dataset API
       new  ad019b6   DOC: Corrections and additions to the dataset docs. Also more detailed docstrings for the functions loading the data.
       new  fa3c9dc   COSMIT: prettify datasets docs
       new  b244f4b   DOC test fix. Set printoptions to get rid of epison.
       new  82d7783   FIX: whitespace after ..
       new  0fb38ad   DOC test fix finally....
       new  d353a96   Merge pull request #469 from amueller/preprocessing_epsilon_doctest
       new  264cec5   DOC fixed fastica docstring: if whiten=False K=None
       new  9592a98   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  8ec89a9   ENH : C scaling of sparse models
       new  75672f1   DOC: start to merge statistical learning tutorial
       new  04aa47c   ENH linnerud dataset interface adjusted to be consistent with the others
       new  7f544b1   FIX: typo in diabetes docs
       new  db74b53   Merge remote-tracking branch 'origin/master' into n_samples_scaling
       new  4298760   ENH : adding missing scale_C in docstring
       new  f1d3374   Merge pull request #471 from amueller/linnerud_renaming
       new  3baab22   Merge pull request #465 from amueller/fastica_wowhiten
       new  f119d22   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9519c69   STY: PEP 257 in ridge.py
       new  de6e8a0   DOC RST field lists don't behave as I want them to:(
       new  c174764   COSMIT datasets doc using rst tables
       new  5fd08bc   Merge pull request #473 from amueller/dataset_whitespace
       new  e4a2916   DOC: Fixed broken reference to user guide
       new  badce0c   Merge remote-tracking branch 'upstream/master'
       new  8ec25b8   FIX This should fix the doctests in the datasets dir. They take quite long, I think it's because of the svmlight loaders. So I didn't include them in the standard make target
       new  8a586e5   DOC: explicit the __init__ convention
       new  1db82e3   Cosmit
       new  96fe959   Cosmit
       new  901e124   fast_svd: factor out the randomized range finder (more generally useful)
       new  9c4325b   Cosmit on randomized range finder
       new  76ce924   Merge pull request #475 from amueller/datasets_doctests
       new  2c3c2b3   BUG: fix RandomizePCA: renaming of fast_svd args
       new  0f24736   DOC: scikit.learn -> sklearn
       new  4378117   BUG: casting with numpy 2.0
       new  a33485f   Remove useless benchmark.
       new  2d734a1   FIX: broken benchmark
       new  515a8e3   BUG: API change in fast_svd
       new  ae7e901   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  74251b9   COSMIT rst formatting
       new  e1325f2   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  e22429f   Merge branch 'master' into sparse-coder
       new  b4ee207   DOC: Added missing rst label
       new  50af0c3   track changes from master
       new  bb3e069   FIX: objective functions in Lasso linear model docs
       new  695ce94   pep8
       new  666c92b   remove unused return_log keyword in GMM
       new  9cef7eb   Merge pull request #477 from jakevdp/gmm-fix
       new  325d5ae   DOC: correct ordering of returns in dict_learning_online
       new  08df0d5   FIX RST references
       new  ae9a516   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  976c16e   ENH : avoid global seeding in plot_polynomial_interpolation.py
       new  05b4025   DOC: clarified dimensions in _update_dict
       new  f0bef20   ENH : clean up plot_feature_selection.py
       new  0ed1fad   FIX rst errors in docs
       new  ae26835   fix k_means docstring to better match the scikit naming conventions
       new  2842f48   FIX doc rst references
       new  ca9472c   ENH: Added feature importances to decision trees and to forests
       new  dcc2228   DOC Added link for Satrajit Gosh, removed dead link for Robert Layton since I couldn't find his website.
       new  5d49086   Merge branch 'master' into n_samples_scaling
       new  a1b40bd   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  7980067   TEST: Added test on feature importances
       new  1f42e15   MISC: FutureWarning on C scaling
       new  0a4591b   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  3a51078   BUG typo in NB error msg
       new  e66fdff   BUG/DOC: address manifold singularity issue
       new  d766775   EXAMPLE: Added examples for feature importances using trees
       new  a7bd476   DOC Robert Layton again.
       new  171dae5   COSMIT: rfe examples
       new  c5b8fb8   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  e50c600   EXAMPLE: Improved plot_forest_importances.py plot
       new  41e0c40   COSMIT: tree examples
       new  6333184   DOC: Fixed links to modules in the example gallery
       new  a8c5a5a   DOC: Fixed broken links
       new  64adbca   Implement mean_variance_axis0.
       new  6db1ff8   WIP: n_init refactoring
       new  8311b1b   Fix bug with sparse matrices.
       new  b4838e7   merge master
       new  13a0507   Fix the API and the scaling inside dict_learning
       new  0d927f0   Cosmit.
       new  ac1c9aa   Test edge case.
       new  e57aa2a   tmp -> diff
       new  ab7b857   DOC: specify scaling in linear_model.rst
       new  270fc7f   EXAMPLE: Moved to the Olivetti dataset
       new  da4fb43   ENH: Accelerate ensemble of trees by precomputing X_argsorted
       new  ce59395   Merge pull request #481 from mblondel/mean_var2
       new  e5628f7   peping8 examples
       new  a232c23   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  3371f9e   FIX: bootstrap=False by default with extra-trees
       new  70c0aa1   Merge branch 'master' into minibatch-kmeans-optim
       new  6759068   peping8 examples/applications
       new  63f06a1   work on failing tests
       new  3e75241   DOC fix datasets.load_digits example
       new  38d8444   scale tolerance of minibatch kmeans on CSR input variance
       new  5ef4a4d   Merge branch 'master' into sparse-coder
       new  10a16a2   EXAMPLES: Removed useless import
       new  29fa29f   delete broken example
       new  5d06bee   skip tests that were wrongly passing before
       new  8636e32   ENH: Use extra-trees instead of rf
       new  697653e   Test for almost equal instead of equal in sparse_encode_error
       new  e78ae4a   COSMIT: examples
       new  5d6d9f4   Merge pull request #482 from DraXus/master
       new  f9aacba   example script is not meant to be executed when building the doc as it is slow
       new  c6613e8   DOC fix datasets.load_digits example, second attempt
       new  66796ca   FIX: slices generation
       new  fde9807   Add score method to KMeans.
       new  96618d1   DOC: add utility information for developers
       new  96ea318   Move graph_shortest_path to utils/graph.py
       new  0690438   remove duplicative utils.fixes.arpack_eigsh
       new  eb248b7   Move validation utils to their own submodule
       new  384f4b7   Hide sparse_encode -- redundant
       new  3aa5b12   DOC: add optimization objective to lasso and enet docstrings
       new  e9f432a   DOC: make docstrings as good as I could
       new  3a68704   Warnings and deprecation
       new  ae1785c   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  4c8aec3   typo: accross => across
       new  af05e76   Merge branch 'master' into minibatch-kmeans-optim
       new  2f3cca5   typo: accross => across
       new  c696ad4   Use int for indptr and indices.
       new  a2bb3ff   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  da38d74   Use python int for indices and indptr of scipy sparse matrices to ensure cross platform support
       new  dc72cf6   Move uninteresting example to docstring
       new  adfae9b   Make init less expensive by default on MinibatchKMeans to avoid dominating computation on large scale datasets
       new  154fcc7   Added links and various cosmits
       new  0c38141   BUG: example plot compatibility with older matplotlib versions
       new  a08d54e   Merge branch 'example-fix'
       new  51f74dd   DOC: better cross refs and docstrings
       new  5e0f0e9   DOC: Added fetch_olivetti_faces to Reference
       new  794b7b1   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  5d47ce2   Sparse matrix support in KMeans.
       new  382bea2   Merge pull request #4 from glouppe/dev-doc
       new  cf1c1f1   randomized_range_finder -> randomized_power_iteration
       new  bb03e8a   Adapted examples for alpha scaling
       new  093a2d8   DOC: Cosmits on the Support page
       new  31582d3   FIX docstring
       new  0572c56   Merge branch 'master' into sparse-coder
       new  213e886   PEP8
       new  b448fe0   Vectorized news20 dataset loader.
       new  43e9934   Change logsum to logsumexp for comparability with scipy
       new  e93482e   Fix broken duplicated / tests and more practical init
       new  ea27e11   BUG: fix scale_C bug in svm
       new  48a6500   ENH prettify kmeans vs minibatch kmeans example
       new  8d5fd79   ENH adjust subplots to look good on the web.
       new  b2666d2   TESTS: remove deprecated NeighborsClassifier calls
       new  c3aa2d7   Merge multilabel branch with master.
       new  621417e   ENH: Parallel fit/predict/predic_proba/feature_importances in forest
       new  1425e29   Check that LabelBinarizer was fitted.
       new  a835c8c   FIX: Ensure random random_states
       new  5885d0b   Separate correction and reweighting steps from raw MCD computation.
       new  00a9911   Add a set of tools and a new object for outliers detection (+ example).
       new  281f5d6   Add tools to perform outlier detection with sklearn + documentation.
       new  ed0087a   Clean working directory
       new  745f981   Integrate AlexG's comments on doc and examples + add tests.
       new  d102afb   Magnify novelty and outlier detection examples again + minor fixes.
       new  26548f9   DOC: Move Parameters section outside objects __init__ method.
       new  29c46c1   Example on real data (outlier detection on boston housing data set).
       new  a9acfca   Fix bugs + adjust OCSVM parameter in outlier detection example.
       new  d898877   Cosmit: address Olivier's comments on examples naming.
       new  361bfde   COSMIT: beautify the plot_oneclass example
       new  46cfacb   DOC: outlier detection improve docs and examples
       new  3b61fee   DOC: improve outlier detection docs
       new  5dc60ba   API: h -> support_fraction
       new  f4b6f45   Cosmit
       new  8975e8c   ENH: use controled random numbers
       new  494fb02   consolidating all cython utils for sparse CSR in the same file under utils
       new  66d462f   ENH: use pre_dispatch
       new  3e6fd74   FIX minor typo
       new  04612b5   FIX errors in doc
       new  aac994b   FIX minor docfixes
       new  7fdd8df   added sparse coding example
       new  6e7fd21   s/threhold/threshold
       new  a646e9a   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into sparse-coder
       new  9a43b03   Merge pull request #456 from vene/sparse-coder
       new  c4d40ad   DOC: Return->Returns
       new  1648c21   Multilabel classification dataset generator.
       new  56b3a3c   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  9c0e7b6   Add SparseCoder example
       new  51fd2bb   Remove duplicate definition in RST
       new  c1003f4   BUG: follow API change in outlier_detection
       new  117be7e   Test multilabel classifier on random dataset.
       new  e4c41ce   STY : pep8 and add print __doc__ in plot_sparse_coding.py
       new  3f7cec3   added kernel approximation using monte carlo approximation of fourier transform
       new  233d3f2   ENH pipeline compatible interface to fit, transform and fit_transform
       new  250a11f   DOC example comparing linear classification, kernel svm and kernel approximation with explicit mapping.
       new  94b4db8   kernel approximation example
       new  158a897   DOC beautified kernel approximation example plot
       new  84c775f   better docs, remove unimplemented kernel approximations
       new  ab51317   COSMIT pep8
       new  52063dc   DOC added kernel_approximation module to docs
       new  258176c   DOC: placeholder entry in user guide
       new  c6ee39b   ENH: renamed D to n_components for consitency
       new  f4fb0ed   DOC approximate kernel functions narrative docs
       new  e6c9606   DOC: more narrative documentation for kernel_approximation
       new  97cf930   DOC: references for approximate feature maps
       new  7619894   COSMIT: pep8 in kernel approximation test
       new  e3a694c   DOC approximate kernels: added formular for skewed chi squared kernel
       new  d111fe2   COSMIT removed commented out import
       new  8380710   ENH: additive chi squared kernel implemented and tested
       new  a831a89   pep8
       new  bd5ffe0   DOC: added AdditiveChi2Sampler to doc modules
       new  f1c543f   ENH: Default value for n in AdditiveChi2Sampler
       new  4548fa0   DOC narrative doc for additive chi squared kernel
       new  933c370   ENH: sensible defaults for RBFSampler and SkewedChi2Sampler
       new  5bac125   ENH Added AdditiveChi2Sampler to feature_extraction __init__
       new  a6655c5   BUG: AdditiveChi2Sampler fit method should return self
       new  3df0815   ENH: in Chi2Samplers, check if input inside inside desired range.
       new  31a47b2   FIX: Renaming of RBFSampler argument
       new  5ab3afc   DOC: Move kernel approximation to be a "plot" example.
       new  1dede23   Don't test as strictly so not to fail randomly..
       new  e40cd69   Example of decision surface of approximate kernel svm
       new  972de6d   Moving kernel_approximation to the top level
       new  7d5c746   ENH: Restructuring User Guide: kernel_approximation, preprocessing and feature_extraction are under a common chapter, " "Dataset Transformations"
       new  3e3136d   DOC: finetuning the narrative docs for kernel_approximation
       new  43b02ed   DOC: kernel_approx make examples show correctly
       new  2955c26   DOC rst
       new  7525923   ENH Addressing some of Gael's comments, mainly naming and docstrings
       new  48be642   ENH better testing
       new  6f3263f   ENH fixed location of the legend in kernel_approximation example
       new  eca01e8   DOC more discussion in docstring
       new  44e9bae   ENH timing results in approx kernel example
       new  5959e7a   ENH kernel approximation: More specific references and example referencing the narrative docs.
       new  761eb3e   FIX: use safe_sparse_dot in kernel_approx transform
       new  0de5a1a   DOC minor doc improvements, different example
       new  7a8e3d8   NONSENSE improve the example that i'll remove in a sec
       new  e9982cd   BUG import ...
       new  23f12cb   COSMIT + SPELL
       new  4c8658c   DOC added reference to the user guide in kernel_approximation module
       new  79ad277   FIX path in plot
       new  a25d611   FIX typo that cost me half a day of sprinting...
       new  39771fa   ENH Remove redundant example
       new  e33d955   FIX fix module links, figure split into two
       new  486cd55   COSMIT pep8
       new  13e6917   MISC: update whats_new
       new  ad19fac   DOC: cosmit in kernel approximation
       new  f30fafe   DOC: removing dangling link
       new  d530fa1   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  a14d46d   FIX: Kernel approximation module in references in alphabetical order.
       new  3f330f2   STY : pep8
       new  a2774a2   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  4761dc3   COSMIT rename load_vectorized_20newsgroups + DOC + pep8
       new  1f3262d   scale_C will be True in scikit-learn 0.11.
       new  fdd73cf   Merge pull request #8 from larsmans/news20_loader
       new  28f947d   Merge pull request #2 from mblondel/multilabel
       new  e51ceb1   WIP: scaling CSRs
       new  7388915   Merge branch 'master' into minibatch-kmeans-optim
       new  0a0d95f   DOC: Cosmit on the reference
       new  cec10b7   ENH: Improved _parallel_predict_proba
       new  d32a220   FIX compat for errorbar legend for old matplotlib versions
       new  b8085fe   BUG only handle labels specially in SVMlight loader + multilabel
       new  de41406   DOC trying to clarify the kernel_approx documentation.
       new  9315187   DOC: add n_jobs to specs
       new  c4decfb   Replace unmaintainable test
       new  7dc1ea0   More robust test for lars_path
       new  2ac0b02   Return bunch object.
       new  51b6dea   Rehauled SparseCoder example
       new  6628bc8   Merge pull request #493 from amueller/kernel_approximation_doc
       new  c8fbfa6   Merge branch 'master' into sc-example
       new  0b0d8d1   Add to class reference.
       new  ef6a563   BUG fix off-by-one error in SVMlight format loader
       new  3fc6f14   Typo in example. Thanks Virgile for the cool example.
       new  f609c35   DOC multilabel learning: note that it's experimental + @mueller's remark
       new  a18fff7   DOC document svmlight file loader changes in changelog
       new  9944b0d   Added @vene's work to the changelog
       new  07f8c6f   DOC FIX typo
       new  560c7e0   FIX docstring errors...
       new  763af61   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  b5a6191   sparse coding transform is now a mixin
       new  a91b89a   slight optim: remove useless assignment from the inner loop
       new  86f909d   FIX: numerical instability caused by collapsed allocation of bad clusters to the center of mass
       new  ff7e187   Add precompute_distances option back and export it.
       new  2654a4e   Merge branch 'minibatch-kmeans-optim' of https://github.com/ogrisel/scikit-learn into minibatch-kmeans-optim
       new  2998a7c   COSMIT reorganise utils tests
       new  ff4ca1c   TST add test for sklearn.utils.extmath.logsum
       new  9784dfd   Address @ogrisel and @amueller's comments.
       new  8197ca3   ENH: Assign chunk of trees to jobs
       new  ea2daf3   species datasets commit
       new  2cbcc1e   example tweaks
       new  07b3690   FIX: missing import
       new  cb629ec   Cosmit in metrics
       new  c57fae8   BUG: fix bug introduced in fd8c
       new  d819ebf   ENH: Store the full cv_scores in grid_search
       new  b1a41dc   clean up species distribution example
       new  4683779   DOC: add alpha_ to attributes of LassoLarsIC
       new  627bdab   COSMIT: utils.fixes: document versions
       new  16a2f6d   misc
       new  4a78dea   FIX: fixed link to Virgile in whats_new
       new  ea5d8e8   randomized_power_iteration -> randomized_range_finder
       new  0a8b6e5   api: changed pos_label to None for metrics
       new  7f5a9ba   STY: s/grid_points/cv_scores
       new  680ea1e   DOC copyedit kernel approximations docstring
       new  fdfd7b0   DOC kernel approximations, some last bits
       new  00f2af5   Merge pull request #495 from vene/sc-mixin
       new  30cdbb8   fix text position in example
       new  91ae12b   typo in fastica doc
       new  ac7cb28   DOC unbreak kernel approx docstrings (UTF-8 + s/References/Notes/g)
       new  e2a4a98   its
       new  dbbdbe4   Better doc for the 20newsgroup dataset loader.
       new  e65bca1   better documentation for the convergence stability example
       new  8c396d8   Merge branch 'master' into minibatch-kmeans-optim
       new  0083db3   Revert code that I erroneously changed
       new  16ef900   Merge commit 'upstream/master' into util-docs
       new  924c856   Merge commit 'upstream/master' into util-docs
       new  c85c890   simplify stability evaluation example
       new  d353331   enable the kmeans stability as an auto examples as the speed is now fast enough
       new  377cc1e   DOC: add toc for developers resources
       new  5ebcd59   EHN: multilabel samples generator can create different number of labels per instance
       new  959dee2   docstring in cython funcs + better var name: with_sqrt
       new  e2685b6   DOC: add warning that utils should only be used internally
       new  0d8efb7   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into minibatch-kmeans-optim
       new  b791792   pyflakes test_multiclass
       new  5c7b3f2   Do not use joblib's memoizer.
       new  4b8b9fd   Merge pull request #486 from jakevdp/util-docs
       new  ad53ecd   Merge branch 'master' into multilabel
       new  9970fe2   Add the samples generator to the references
       new  8f7cfdf   cosmit
       new  572a7e3   use joblib for saving species data
       new  df4351d   Merge commit 'upstream/master' into dataset-fix
       new  75c30c6   merge master
       new  3d2ea0d   Use int16 for more compactness.
       new  65a47a9   readd dtype and ccontiguous checks removed by mistake during last conflict resolution
       new  8ad077d   fix logsum test
       new  d7d99d9   ENH add multilabel_ property to OvR and raise NotImplementedError in score
       new  3530efa   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into minibatch-kmeans-optim
       new  17afd07   Merge pull request #490 from mblondel/news20_loader
       new  9dac9c9   Merge branch 'master' into sparse-kmeans
       new  0763ee8   Merge with master.
       new  e4a7eaf   One more test.
       new  856c56d   Fix test.
       new  71599e5   Merge pull request #488 from mblondel/sparse-kmeans
       new  8f3c2c8   Remove old API change warning
       new  e705d94   Change depreciated behavior in feature agglomeration example
       new  cad0ee0   HACK: sphinx/prevent proliferation of build images in doc
       new  0ce0f32   FIX: Added DBSCAN to references
       new  f29b29a   FIX: typo in docs
       new  216a012   ENH demo sparse KMeans on 20news set (it's slow!)
       new  5832946   Cosmit in MiniBatchKMeans.
       new  fbab710   ENH: Added the synthetic example
       new  400093f   ENH: Really added the synthetic example
       new  18a70f7   Merge remote-tracking branch 'vene/lars_multilabel' into multilabel
       new  ba475b1   BUG forget a return keyword in OvR classifier
       new  f41352e   DOC describe test_ovr_multilabel better
       new  6e619e2   Optimize for high dimensional data.
       new  c90571b   TST extra test for LabelBinarizer's multilabel behavior
       new  e3163ed   COSMIT set union in LabelBinarizer
       new  ac5b122   simplify removal of _images dir
       new  b322539   remove unneeded import
       new  fa74a6c   DOC: add multiclass to class reference
       new  f2b282d   DOC: add example to multiclass.rst
       new  8c6cee2   Merge pull request #504 from jakevdp/sphinx-images
       new  2da4474   BallTree -> NearestNeighbors in Isomap
       new  42b47ac   DOC: really add example to multiclass.rst
       new  6c6d9e3   DOC: add image to narrative doc
       new  97bf1ad   Added missing space in PIL warning
       new  87a85c1   Merge pull request #417 from larsmans/multilabel
       new  d38e3a7   ENH improve stoplist handling in feature_extraction.text
       new  3359032   DOC: isomap fixes
       new  ed34ad6   convert LLE neighbors to NearestNeighbors object
       new  aba3020   BallTree -> NearestNeighbors in mean_shift
       new  ff517dd   DOC rm References sections in docstrings
       new  6155c97   pep8
       new  8cd34b1   DOC update changelog
       new  38f78a0   Variable renaming and dostring fixing
       new  2108fcf   merged with master logsum -> logsumexp
       new  6a37e47       ENH: renaming estimated variables from self._variable to self.variable_
       new  b4d4a9f   removed the decode
       new  2826ac7   removed the decode in dpgmm and removed return_log in eval
       new  412970a   ENH: Cleaned after rebase and compatibility with hmm
       new  aad9a42   ENH: Removed X and z varaibles from dpmm cladd (should not ship the data)
       new  a965318   ENH:aviod initializaing GMM means with zeros
       new  a830ca7   ENH: more snsible initialization in case of divergence
       new  54a11a0   BF: Mended the tied covariance estimator
       new  9e21156   ENH: added multiple initialization to the GMM -- untested
       new  0ced7a6   FIX: fixed collateral dammages in hmm
       new  ec3b839   DOC I broke the docs and I liked it
       new  d3954a3   added some tests to ensure that GMMs work in about all conditions
       new  3fff256   COSMIT minor ticks
       new  1a78610   FIX getting rid of some more sphinx problems
       new  5be18b4   FIX: SO EINE SCHEISSE!
       new  af93a59   COSMIT fixing indents in balltree
       new  1aa10ea   Merge pull request #510 from amueller/aaarrrgghhh
       new  09d8090   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  089c02c   COSMIT "Examples" instead of "Example"
       new  de202d7   ENH: renaming cv_type and posterior to more explicit name + tested multiple init
       new  6abd1c3   avoid changing the covariance when computing the Gaussian density
       new  2ff74a3   Merge pull request #507 from jakevdp/neighbors-check
       new  a255ed9   COSMIT Addressing @agramford's comments about whitespace and a minor fix in pipeline.
       new  9420b60   Add Andy to the author list
       new  d0ac763   Allow unlabeled samples in multilabel ex, collab between @vene and @mblondel on the plane
       new  512c911   FIX typo that broke the test
       new  1382b26   Use CCA as well in multilabel example.
       new  355a51c   FIX: Section Returns not Return
       new  e087daf   COSMIT: Class docs don't have a 'Methods' section. It is autogenerated.
       new  ccb7493   COSMIT Examples not Example
       new  6c17d0d   COSMIT make 'References' bold and minor other fixes.
       new  26a7934   COSMIT underscore fixes
       new  cd161e2   COSMIT "Optional Parameters" Section removed
       new  61c2cee   ENH make example more expressive
       new  955fdf8   Change seed to make example behave better
       new  ae143a7   fix: xmin -> X.min() cosmit: unused imports and vars
       new  122e8b4   EXAMPLE: renamed Frankenstein, set cmap in matshow
       new  873b13d   ENH: Forest -> BaseForest
       new  c730edc   Merge pull request #512 from amueller/doc_consitency
       new  ce00c68   Merge pull request #501 from jakevdp/dataset-fix
       new  0cf16dc   merge master
       new  3a67aa0   Removed unused imports in species dataset
       new  75af061   remove useless dependency on pylab
       new  18f4e56   fixed conflict in import resolution
       new  1725ecb   remove unused import
       new  1006a65   FIX: Fixed a buf I introduced in dpgmm
       new  7f53024   remove unused imports
       new  5c6ea0c   ENH: Added AIC/BIC + tests. Seems to work
       new  96d6060   pep8
       new  3a6e490   COSMIT pep8
       new  4e46115   Merge commit 'upstream/master'
       new  6bdc904   COSMIT: pep8
       new  60971cd   FIX developers rst malformed
       new  5b492d3   remove unused link
       new  874e2bd   COSMIT remove unused malformed tag
       new  bd13d21   FIX indentation and string literals
       new  0ebf87b   FIX backtics for members_, spaces around colon (not cologne)
       new  b5be2ce   COSMIT minor docstring stuff
       new  116dc6e   COSMIT remove Methods section
       new  c9aecfb   FIX: rename complexity section into notes section
       new  b5d2282   FIX docstring variable names
       new  b03be4b   FIX rename "Details" into "Notes"
       new  511c370   FIX: validation is a relative package
       new  2cf1ac7   FIX: py3k - more relative imports
       new  5fecb53   FIX: py3k: string.letters is locale dependent and absent in py3k
       new  57a90f8   Merge branch 'master' into sparse-scaler
       new  1d01534   FIX remove infinite recursion
       new  1fdbbfb   WIP: feature scaling for CSR input (lacks some tests)
       new  16f0864   COSMIT: Make references link and show up correctly, parameters of __init__ documented in Class, not in function.
       new  760692e   COSMIT make formulars show up correctly, use reference formatting for references
       new  7cf021b   COSMIT make references use reference formatting
       new  6af9b01   COSMIT format references and dict stuff...
       new  ea4db67   COSMIT Indentation of formulars
       new  1dab0e1   Cosmit in dpgmm
       new  74b95bb   merged with master
       new  f1f6fee   COSMIT make BaseLibSVM an abstract base class
       new  872e70d   BUG input validation in kernel approximations + pep8
       new  fc91c53   BUG fix Vectorizer to play nicely with Pipeline
       new  617d731   BUG Disallow negative tf-idf weight
       new  f116314   Revert "BUG Disallow negative tf-idf weight"
       new  0e74917   PY3K fix in datasets.samples_generator
       new  f85e7be   fix scaling, more tests and docstrings
       new  b0f221a   Merge branch 'master' into sparse-scaler
       new  2e0857a   wording
       new  708b61a   FIX: py3k integer division in robust covariance estimation
       new  cd27709   FIX: py3k integer division in samples generator
       new  8232d49   FIX: in py3k svmlight files must be explicitly opened in binary mode
       new  826443e   ENH: update joblib
       new  68b7ca3   FIX: py3k bytes split in svmlight format parser
       new  9fe13b3   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bc18af3   FIX: py3k need explicit bytes buffers for svmlight format serialization
       new  29c092d   FIX: py3k need output file in binary mode for svmlight format serialization
       new  8226a98   FIX: py3k: string formatting is not supported on byte strings
       new  0f8b3c2   FIX: fix test: integers are valid file descriptors in py3k
       new  33aca24   Changed the shape of spherical covariance matrices to be equal to disgonal covariance matrix, in order to avoir handling the dimension in particular
       new  90a9ae6   Merge branch 'master' of github.com:scikit-learn/scikit-learn into gmm-fixes
       new  86559e5   detail fixed in an example
       new  03a4e3b   Add missing reference.
       new  9ad004f   Break down fit_transform into parts.
       new  59cdcc7   Merge branch 'master' into sparse-scaler
       new  b2dac63   FIX: unused cython variable
       new  671f304   More checks when transforming sparse matrices with centering scalers + typo
       new  c4f38fa   DOC: update narrative documentation
       new  cd09d87   DOC: Added reference for feature importance
       new  6707fec   ENH: Revisited importances API
       new  ab23e22   Merge remote-tracking branch 'upstream/master' into tree
       new  83f693f   EXAMPLE: Fixed API changes
       new  df8adf0   ENH: Missing default value for feature_importances_
       new  7a8b4f2   FIX removed duplicate explicit linke for Vlad
       new  600f685   FIX: RST indentation and blank lines
       new  b623aa0   FIX RST and references
       new  f57d254   FIX minor rst
       new  432e7a4   FIX workarounds for docutils bug
       new  e5fb9e9   FIX whitespace where rst demands it...
       new  d6dc36b   FIX workaround for table problem
       new  78f64c5   FIX two more underscores
       new  8a8c6eb   optim: avoid useless memory copy when input is non CSR
       new  51a1768   DOC: typo / wording
       new  3efe8d3   DOC: document sparsefuncs cython routines in developer section.
       new  235f274   MISC: improve copy_joblib script
       new  7719b37   ENH: Integrate joblib 0.5.7
       new  db01ca9   DOC: wording
       new  9c5a37b   ENH: Added SelectorMixin
       new  222dd20   DOC: wording
       new  2f5b368   TEST: Added tests of transform
       new  6ccbaa5   ENH: Simplified API
       new  5ecb141   Hopefully clarified notations in dpgmm
       new  951dffe   DOC: Tree-based feature selection
       new  ba71108   ENH: don't sum if coef_ is 1-d
       new  01f71f7   ENH: Inherits from TransformerMixin
       new  986c43e   PEP 257
       new  228d91c   PEP 257 (bis)
       new  b0d2771   ENH: address @ogrisel comments
       new  f3cb852   FIX: Used np.abs instead of ** 2
       new  b47666f   DOC: formatting
       new  2bab89b   Merge pull request #515 from ogrisel/sparse-scaler
       new  7808e63   update what's new for sparse scaling
       new  37be658   Merge remote-tracking branch 'upstream/master' into tree
       new  eb760c2   ENH: Smart thresholds
       new  f9cbfb9   Cosmit
       new  aceaf72   PEP8
       new  6c7496f   Fix the docstring of the univariate feature selection module to match the scikit conventions
       new  48c193d   cosmit
       new  f6006f8   typo
       new  45c5482   cosmit
       new  bb52a45   FIX: None and int comparison not authorized in py3k (in PCA)
       new  a234019   FIX: dicts no longer have the has_key method in py3k: test for the method we actually use instead
       new  1bd911e   FIX: make feature extraction work with the new py3k string API too
       new  6a973dd   FIX: py3k's zip is not subscriptable
       new  17e06cc   FIX: handle py3k exception API
       new  1994ae6   FIX: previous fix for py3k str API in feature extraction was a bug in python 2
       new  bdfc019   FIX: pervasive use of unicode in feature extraction for py3k compat
       new  a3aa805   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into doc_underscores_for_real
       new  169df41   COSMIT docs hmm
       new  bf0327a   FIX: don't use latex in rst
       new  56c9dc8   FIX + COSMIT rst warnings
       new  c8bc552   Many corrections in dpgmm to remove en-necessary loops (significant speed-up) + renaming
       new  2e1040e   COSMIT docs
       new  bed5f4c   FIX: fix again errors in NMF after merge
       new  e56a242   DOC: :mod: link
       new  5e5d797   FIX: Document properties in a way that the docstring actually shows
       new  7177b3f   FIX: rst errors in ball_tree
       new  ed7de9f   FIX: Notes instead of Note in preprocessing init
       new  5c997d1   FIX remove handles for references as they are not used anywhere and raise warnings if doubled.
       new  c8208ea   ENH: no predispatch with chunk strategy
       new  0c926b5   Fixed an example that happened to fail
       new  6cd4ff6   Several details outlined by Jake
       new  244dcab   FIX: Address Gael comments
       new  886a004   Merge pull request #478 from glouppe/tree
       new  f967f7e   DOC: doctest bug
       new  97c2533   Cosmit: example prettier without colorbar
       new  aaf9435   DOC: add links to examples.
       new  fdb98a0   DOC: improve univariate feature selection docs
       new  a595754   MISC: move SelectorMixin outside of __init__.py
       new  c78b1d2   OPTIM: minor optimization
       new  c1becdc   MISC: better error message
       new  d73b24f   COSMIT
       new  8361178   TST: fix doctest
       new  e692a32   TST: fix touchy doctest
       new  84a4d39   Merge remote-tracking branch 'upstream/master' into parallel-forest
       new  2c510c4   ENH: Simplified parallelization
       new  7799d3a   PEP8
       new  a5db7f1   ENH: Simplified code
       new  79cdf46   DOC: Quick docstrings for private functions
       new  48a63de   FIX: Revert changes
       new  da927fb   DOC: What's new
       new  1d1348e   Update random forest face example to use several cores
       new  7aedf4b   Merge pull request #2 from ogrisel/glouppe-parallel-forest
       new  0e646ef   ENH: make ShuffleSplit able to subsample the data
       new  ad3de7f   handled the eval on Null data
       new  3a5476f   FIX: Address @ogrisel comments (1)
       new  a6027a5   Merge branch 'parallel-forest' of github.com:glouppe/scikit-learn into parallel-forest
       new  12d2913   COSMIT: avoid set_cmap and pcolormesh in example
       new  9e9bd1c   FIX: ensure fetch_20newsgroups_vectorized outputs CSR matrices to work with cross validators
       new  9a46722   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  8d654ff   TEST: Added tests of parallel computation
       new  e0c3eb3   DOC: Parallel computations in forest
       new  96b4a12   TEST: Improved coverage of the ensemble package to 100%
       new  ab5fea0   DOC: Renamed example (+Parallel)
       new  91b5bec   Merge pull request #519 from ogrisel/subsampling-shufflesplit
       new  13a1114   Merge remote-tracking branch 'upstream/master' into parallel-forest
       new  8bb13d2   Merge pull request #491 from glouppe/parallel-forest
       new  ae564d7   DOC: Added missing BSD3 licenses
       new  73e65c2   PEP257: docstring cosmits in utils.extmath
       new  0f6c525   ENH: renamed fast_svd to randomized_svd + related improvements
       new  4b60c02   Merge pull request #513 from amueller/doc_underscores_for_real
       new  ee4c204   FIX: incomplete test for inverse_transform in text feature extraction
       new  83750ba   improved computational speed by calling fast scipy build-in function and replaceing double loop
       new  281e1d4   fixed some pep8 warnings
       new  51024b9   Merge remote branch 'upstream/master'
       new  1bb76e6   Merge pull request #521 from lucidfrontier45/master
       new  a174522   Cosmit in docs
       new  124594d   ENH: Better default values to trees and forests
       new  64ea90b   TEST: Added tests of max_features values
       new  b92ad62   DOC: Review of the narrative doc wrt max_features
       new  00102c1   DOC: Added warning to default values
       new  885a43a   DOC: typo
       new  0f2e443   Merge pull request #523 from glouppe/tree-doc
       new  4946c62   DOC: fix broken doctest
       new  4cc04d0   FIX: max_features=None by default on single DT
       new  6b05ec8   pep8 in forest.py
       new  26662bd   pep8 in tree.py
       new  10d1207   pep8 in kmeans tests
       new  7ca1756   more pep8
       new  f09db8d   more pep8
       new  5443e19   FIX: heisen doctest
       new  23909c7   DOC: readibility: make colon after 'Parameters' stay on the same line in reference documentation
       new  8735140   FIX: Boston is a regression dataset
       new  c71dbdf   oops, the last test is about classification, not regression
       new  c4b49bb   COSMIT docs underscore fixes (again)
       new  73b37a0   COSMIT fixing doc errors and making html docs pretty
       new  d1c8a39   COSMIT Minor beautifications and RST error fixes
       new  bb6e07c   MISC: fix bibtex
       new  0e7c16b   Cosmit
       new  3fb1bc6   FIX doctest errors + cosmit
       new  dfe3a2f   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  d2c3419   DOC use SVC in grid search instead of SVR. Iris is a classification dataset as pointed out by @agramfort here: https://github.com/scikit-learn/scikit-learn/issues/171
       new  5ed5da5   DOC score returns accuracy, not error
       new  d8ae74f   FIX for doctest I just broke :-/
       new  b917b13   More tests for nuSVR.
       new  bff621e   DOC uncommenting doctests in balltree.pyx, addind doctest: +SKIP this is not terribly pretty in the source but looks good in the docs. I think this is a reasonable compromise.
       new  208c2d7   COSMIT a little less skips...
       new  e286125   merged the master repo
       new  c2169fa   FIX broken links to Rubinstein's K-SVD paper.
       new  0e4886e   Merge pull request #527 from otizonaizit/master
       new  d61ed89   FIX: Add reference (stop words)
       new  d4c4c6f   Merge remote-tracking branch 'upstream/master' into issue349
       new  edfa3e9   Merge pull request #528 from glouppe/issue349
       new  64ea6e5   fixed the function definition of cross_val_score
       new  2a3c8f5   changed cross_val_score doc again
       new  841d679   Merge pull request #529 from eickenberg/doc_fix
       new  25c8d8d   DOC: pep8, add quotations, and fix typos
       new  e33d8fc   ENH Add underscores to estimated attributes in GridSearchCV and deprecation warnings.
       new  567338f   fix for doc math issue
       new  3fa8271   TYPO: generate all images
       new  96dcaf7   ENH: Make LinearRegression work with sparse
       new  68a8b8b   ENH renamed best_estimator and best_score in examples and tests.
       new  b9af16b   DOC: update LinearRegression docstring
       new  788b96e   FIX: sparse LinearRegression with scipy 0.7.0
       new  ed0584f   Merge pull request #532 from amueller/grid_search_attributes
       new  e4c44e8   COSMIT typo
       new  2882b97   ENH in GaussianNB, let estimated parameters have underscores.
       new  6cebdb5   added a cython module to the hmm
       new  a703048   replaced (T, N) -> (n_samples, n_states)
       new  c057dab   eliminate `compute_importances` fit parameter - make `feature_importances_` a property that will be computed on demand.
       new  f8b8c2d   Added an example with model selection
       new  e77ee68   Oups: really added an example with model selection
       new  0faf870   ENH add label propagation algorithm
       new  1543ba2   FIX: docstring
       new  fbc1bac   scikits.learn -> sklearn migration in label propagation
       new  4c44755   BUG don't pass estimator params to fit in label propagation
       new  df80486   initial_error -> init_error
       new  12df639   max_features bug in _tree.pyx (check if < 0 and assume all features!)
       new  4a70636   Merge branch 'tree-feature-importance' into old-gradient-boosting
       new  83bedfd   merge with master finally resolved!
       new  dac2e4f   - renamed (n_samples, n_states) -> (n_observations, n_components) in hmm.py
       new  0a92a94   ENH: mark coef_ as immutable for linear SVM models trained in the dual
       new  54fcfc2   ENH: update joblib
       new  1a62567   immutable coef for the sparse SVM variant too
       new  d7c313c   mark liblinear coef as immutable too
       new  3277b9e   MISC: tag explicitely a dependency
       new  1694a8f   ENH: use joblib compression in datasets
       new  58b2b57   MISC: tune test verbosity
       new  b9340e4   ENH: update joblib
       new  bbf65fc   DOC Reworking Bayesian regression documentation
       new  bf8cc20   DOC mentioning sparsity of ARD, reblocking text
       new  a42c94f   COSMIT typo, thanks @vmichel for pointing it out.
       new  d424973   DOC added reference for sparsity of ARD
       new  436fcbe   document the fact that coef_ is readonly for LogisticRegression and LinearSVC
       new  fda55b7   DOC: restore index on pages
       new  a24486b   Merge pull request #526 from amueller/ball_tree_skip_doctests
       new  d869bf3   Merge pull request #537 from amueller/gaussian_nb_underscore
       new  da5d1d8   avoid a memory copy in coef_ property
       new  26632b6   Merge pull request #541 from ogrisel/immutable-readonly-coef
       new  6e250d3   DOC cosmetics in SVM docstring
       new  2f64f0d   ENH : reformatting hmm_stock_analysis.py examples
       new  c5c62e5   MISC : typos in hmm_stock_analysis.py
       new  0b84a03   ENH : rename hmm_stock_analysis.py so it appears as a figure in the doc
       new  7e70d19   Merge pull request #1 from agramfort/hmmc
       new  d58f2cf   enh: performance enhancement by removing redundant computation of values - we use the state of `criterion` instead.
       new  9ae9844   COSMIT reintroduce ABCMeta into BaseSGD*
       new  df19607   dropped "_c" suffix
       new  5b00f44   BUG refactor SGD classes to not store sample_weight
       new  e85edc8   started work on gradient boosting docs
       new  746d686   remove obsolete `sparse_coef_` doc string
       new  954b390   remove reference to obsolete `sparse_coef_` parameter.
       new  069bc0b   COSMIT rm unused svm.base.dot
       new  722fef6   BUG use ValueError in BaseLibSVM.coef_
       new  070007f   COSMIT pep8
       new  5745b2e   BUG update test for SVMs raising ValueError for coef_
       new  5729aa2   DOC fix linking to load_sample_images and load_sample_image in docs
       new  6e67a82   DOC underscores in DeprecationWarnings... shame on me for forgetting that....
       new  35c1222   DOCs workaround for docutils bug (column alignment problem)
       new  3e4b6df   DOC external references go under "references" not "see also". "See also" can only handle internal references
       new  a6e7178   COSMIT remove superfluous imports in svm/sparse/base.py
       new  a04d701   FIX: broken link in SVM doc
       new  60118bf   FIX: issue #540, make omp robust to empty solution
       new  d335640   Merge branch 'omp-zerofix'
       new  0a256ae   DOC: Removed performance and utilities from toctree (they were appearing twice)
       new  6cadc54   set coef_ to fortran layout after fit - this will enhance the test time performance for predicting singe data points.
       new  9a98f73   added to whats new
       new  2da6464   DOC: Fixed 'See also' in tree/forest
       new  4ddf69e   cosmit: more detailed doc string for why fortran style arrays
       new  79f949e   Merge branch 'sgd-fortran-layout'
       new  06c48ff   ENH: Removal of properties from GMM -- unfinished
       new  b43aa5b   debugged _hmmc.pyx
       new  552bc9d   DOC: exaple for sklearn.test()
       new  ec72437   FIX: convert lena to float32 (originally it's ints)
       new  2108f11   FIX: doctest
       new  a072187   Still some tweaks for the sklearn.test() example
       new  dfc7c5c   Remove pylab code from docstring and +SKIP those that requie PIL
       new  f4eced3   BUG don't use deprecated attributes in GaussianNB.predict
       new  0372ba7   FIX: explicit conversion to float64 in ElasticNet
       new  f977346   FIX: bug in elasticnet with precompute not being updated correctly.
       new  f2f8caa   fixed proble of _accumulate_sufficient_statictics in hmm.py
       new  f4da029   Merge branch 'master' into old-gradient-boosting
       new  d0d595d   ENH liblinear: cythonized sign switch for n_class<=2
       new  e52da1e   removed feature_importances_ property in tree module
       new  8c65e21   work in progress on GBRT docs
       new  73a7f77   added script to bench sklearn gbrt against R's gbm package.
       new  0777c6c   cosmit: pep8 + comments
       new  b10226b   fix: undo compte_importances property merge in forest module and examples
       new  5427f81   ENH liblinear: get rid of n_class sign by switching class signs in liblinar implementation.
       new  6a9a449   removed properties from dpgmm
       new  8db0338   replace log_weights_ by weights_, which makes the API more consistent
       new  68444c0   finalized KNN work, all tests pass properly
       new  fdfb531   Merge branch 'larsmans-label-propagation'
       new  bcd5e33   removed extra semisupervised folder
       new  04a0ee2   polished the lp & test code
       new  af0feb9   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  691ac41   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into label-propagation
       new  e3ed8e0   Use rbf_kernel.
       new  68236ab   Add decision_function to ElasticNet.
       new  d581006   wip: narrative doc
       new  ded1e17   fix: table layout
       new  9d638ec   DOC: complete docstring for regression score function
       new  c0ed2af   DOC: restructure docstring of ElasticNet.
       new  895dc9c   Changelog
       new  2f08660   change version number to 0.10
       new  4533aa3   Start of 0.11 development cycle.
       new  a2925ef   Mailmap alias
       new  5f5d994   And the winner is ...
       new  187164a   Mailmap alias
       new  7c25674   And the winner is ...
       new  446e781   DOC: links for people that have webpage.
       new  acefa36   restore original
       new  7273da3   restored original version
       new  ffaeac9   restored original version
       new  24c00b5   restored original version
       new  4f47792   restored original version
       new  50227d9   DOC: links for people that have webpage.
       new  d2aa4c6   ENHanced the multilabel example aspect
       new  7b3d819   MISC: species distribution example plotted
       new  a3df4fb   MISC: species distribution example plotted
       new  2f79cb2   ENHanced the multilabel example aspect
       new  b5e928f   Merge branch '0.10.X' into releases
       new  f6d4094   ENH: we are packaging releases so dfsg rules works on releases branch
       new  a8cef47   fresh changelog for upcoming 0.10.0
       new  3207240   Adjusted dfsg and other rules to operate on sklearn directory
       new  f774f82   Merge releases into DFSG (pruning externals: joblib)
       new  9ae5dd9   Merge branch 'dfsg' into debian
       new  26ee247   Dropping cherry-picked fix up_release_sv_coef_memory
       new  4b0ee77   refreshed the joblib patch
       new  2729f06   Getting rid of properties in hmm, gmm, dpgmm
       new  97cc3f9   DOC: typo
       new  3d255e9   COSMIT typo
       new  43887f1   COSMIT typo
       new  049d4e4   2011 -> 2012
       new  937cf13   whatsnew: gave myself some credit
       new  8b11f3e   whatsnew: gave myself some credit
       new  420eef1   FIX: support for regressors in multiclass module.
       new  34334f5   DOC: some documentation fixes.
       new  f371edd   DOC: some documentation fixes.
       new  094e2d0   ENH: better error messages
       new  b6f98fe   remove deprecated Neighbors{Classifier,Regressor}
       new  a43a187   ENH raise ValueError in metrics instead of AssertionError
       new  fabb996   Python 2.5 compatibility dropped
       new  05bd43e   Operate on 'requested' not all supported Python versions
       new  477bcf0   fix an error in naive bayes docs
       new  32f611e   Merge pull request #551 from fannix/master
       new  f748b23   Support for coef_ in OneVsRestClassifier.
       new  a7e5cc2   Mention multi-variate resgression support in Ridge.
       new  b2cba9f   ENH intercept_ on linear OvR clf + change exception to AttributeError
       new  55e3226   Merge branch '0.10.X' into releases
       new  81599bb   Merge releases into DFSG (pruning externals: joblib)
       new  d38004d   Merge branch 'dfsg' into debian
       new  79749fd   Added 'set -e' to composite cmdline constructs in debian/rules to prevent swallowing errors. Thanks to Jakub Wilk for citing me the relevant exerpt from Debian policy ;-)
       new  05c42cc   Add safe_mask utility.
       new  1a8e6e0   coef_ and intercept_ in LinearSVC are now writable.
       new  c988ad7   Add safe_mask to developer doc.
       new  812b41c   Typos.
       new  63fea8b   ENH : make metrics.auc work with 2 samples + add test
       new  388e99d   Create partial_fit and call partial_fit from fit.
       new  1e534ce   Add partial_fit to SGDRegressor.
       new  944ff6e   Partial tests + fix bugs.
       new  7b62b29   Fix a few more bugs.
       new  ee5beef   Use proper assertions.
       new  5048670   Fix more bugs + tests.
       new  fec3557   Add decision_function to SGDRegressor.
       new  869a02b   DOC pep257, or "sentences end with a full stop"
       new  789827a   Multiclass tests.
       new  0d48f49   Merge dense and sparse SGD implementations.
       new  52a076b   Re-enable sparse tests.
       new  f9701d3   Add deprecation warning.
       new  ab71276   Update docstrings.
       new  00dceab   What's new.
       new  e5c7ea7   Removed needless line.
       new  ffc51fe   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  f95e5c9   Merge branch 'master' into svm_coef_sign
       new  b804f52   FIX adjust _set_coef_ and _set_intercept_ to sign switch
       new  6588974   ENH DenseBaseLibSVM.coef_ correct. test simplified.
       new  60a5053   DOC try to document layout of dual_coef_ in multiclass libsvm
       new  8852786   DOC fixed errors in load_images doc and SKIP'ed load_image doctest as was already the case for load_images
       new  7912b65   DOC: OCD and added image loader to class reference
       new  7ba791d   DOC Trying to enhance the tree/forest docs. Headlines in tree, added reference, hopefully better description of 'min_density'.
       new  c716902   - removed unnecessary **kwargs specification in fit and _do_mstep methods - addad trailing underscores to estimated quantities.   e.g. self.startprob_, self.transmat_, self.emissionprob_ etc
       new  3228a53   Use only one epoch in partial_fit.
       new  8717916   Use named parameters.
       new  36c2dcb   Updat examples.
       new  fb04a33   Update doc.
       new  2c6866b   Use only epoch SGDRegressor.partial_fit.
       new  8ec3395   Save iteration number.
       new  0117b60   More tests + fixes.
       new  aca8445   Fix bug when fit is called mutiple times.
       new  d712b03   Fix "what's new".
       new  9478ac0   ENH input validation in DBSCAN
       new  d40666a   DOC rm confusing line in BernoulliNB docstring
       new  7599eff   replaced deprecated "rvs" to "sample"
       new  63ce5a5   Example on tomography reconstruction with Lasso for the gallery.
       new  7414e37   COSMIT: PEP08
       new  f43236b   Tomography example: PEP08, typos...
       new  e957a75   Reference to tomography example in narrative doc
       new  6cdf7dd   MISC: shorten a bit the description
       new  810e7c0   DOC: fix image
       new  0b026db   FIX small stuff in new tomography example
       new  c978177   DOC: layout
       new  1ac5fd1   factor out some common code in dense/sparse SGD
       new  22aba1f   prevent a copy in SGD regressor fitting
       new  3bf7e92   refactor SGD, part 2: simplify parameter passing
       new  f9a7307   DOC: random selection of frontpage images
       new  ac14c74   DOC: compress a bit the layout
       new  ee8f828   DOC: shorten a bit the front page
       new  19756dd   DOC: avoid imgs taking 2 lines
       new  cb59fa2   Merge pull request #10 from larsmans/sgd_partial_fit
       new  8aeb7a1   DOC: Add a few images to the banner
       new  3a12976   DOC layout of dual_coef_ in 1vs1 svm in user guide, example
       new  eccb452   DOC fixed indices in dual_coef_ example
       new  d257e71   COSMIT factor out 1vs1 coef construction in libSVM, PEP8
       new  bbeec6d   FIX: make sklearn.base.clone robust to empty params
       new  65b7e4e   DOC: fix wrong link
       new  bf6c7a4   DOC: avoid line return
       new  6f66978   first stab at trying to wrap MurmurHash3
       new  ec20311   ENH: get the murmurhash to build properly
       new  e50415c   Address @ogrisel and @larsmans 's comments.
       new  5d18b0d   pep8!
       new  e135962   Merge pull request #3 from GaelVaroquaux/murmurhash
       new  7136ff0   FIX: y should be np.float64.
       new  94717db   Add filter_params option to pairwise_kernels.
       new  1917b0b   DOC: prettify ensemble docs
       new  ef92060   Precomputed kernel can actually be non-squared.
       new  b7ffb6b   Use pairwise_kernels in KernelPCA.
       new  778ed5d   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bb6ee7d   refactor SGD, part 3: factor out more sparse/dense common code
       new  6a2f824   COSMIT rm no-op conversion in SGDRegressor
       new  76035f2   BUG restore symbolic class label support in SGD + test it
       new  72fc229   Merge pull request #11 from larsmans/sgd_partial_fit
       new  4fb1dc4   More technically correct description.
       new  ffcc6ff   small simplification in LDA
       new  9438f99   DOC added RidgeClassifier to References
       new  207cc74   DOC fixes in Multiclass docs. Didn't show correctly on web.
       new  263d6c0   ENH merge dense/sparse LinearSVC, part 1: no more SparseBaseLibLinear
       new  915cf0c   DOC multi-class narrative: added links to the references, made citation clickable
       new  a9470d1   ENH merge dense/sparse LinearSVC, part 2: no more sparse.CoefSelectTransformer
       new  498ed0e   DOC: docstring update for dump_svmlight_file
       new  9b0fbab   ENH merge dense/sparse LinearSVC, part 3: deprecate sparse.LinearSVC
       new  0b0bcfc   ENH merge dense/sparse LinearSVC, part 4: deprecate sparse.LogisticRegression
       new  cd131ee   DOC reference for logistic regression training with liblinear
       new  278fe52   COSMIT refactor liblinear bindings
       new  db7271f   TST merge dense and sparse LogisticRegression tests
       new  152d225   implementation & test for the murmurhash wrapper module
       new  c80083c   Export some public cython API
       new  8b0fe8c   DOC: add entry for murmurhash in the developer utilities section
       new  8126e5f   Rename _get_params() to get_params().
       new  85032d4   Merge branch 'sgd_partial_fit'
       new  d5f02da   Merge branch 'master' into merge-linearsvcs
       new  b223933   COSMIT fix ugly import, left over from LinearSVC refactoring
       new  1908176   DOC put merged LinearSVC and LR in changelog + explain @mblondel's work
       new  f9d1efe   ENH: add the ability to hash int arrays
       new  4b6d1f0   Better docstring
       new  6bbdd4b   Use classes_.
       new  db2b556   BUG fix SGD doctest
       new  a713982   Merge pull request #561 from larsmans/merge-linearsvcs
       new  4ac7686   Better title in README.rst.
       new  c6f17c0   Shorter cpdef function names + missing docstrings
       new  e33d41e   DOC: give usage example
       new  9c5ef62   test developers utilities as well
       new  5c8d4e4   OPTIM: avoid unlikely np.int32 test upfront
       new  558d727   Merge pull request #564 from ogrisel/murmurhash
       new  98ec09b   More intuitive warm-restart in SGD.
       new  691e445   Fix doctests.
       new  a438087   warm_restart -> warm_start
       new  438b3a6   More intuitive warm-start in ElasticNet.
       new  9b27f68   Fix doctests.
       new  caae5c3   Copy in user-land.
       new  67a72eb   BUG promote type-safety in murmurhash
       new  44abb43   Missing docstring in ElasticNet and Lasso.
       new  cf074b4   Fix failure in `test_bad_input`.
       new  eb55ec7   ENH trees in random forests save the indices of the training data used in bootstrap sample
       new  b5dd4a3   BUG: restore score functionality in grid_search
       new  0844118   ENH: refit now works in the GridSearchCV
       new  0973860   ENH Add function to predict on left part of training set
       new  a83bc7e   ENH use self.classes_, check input on predict_oob, add test
       new  1afd7cc   DOC Out of bag error estimates in grid_search module
       new  ba39918   COSMIT @glouppe says this is more pythonic :)
       new  cb272e1   DOC reformulation out of bag error
       new  de16cb4   BUG make coef_ 1-d in Naive Bayes for binary case
       new  eac110f   COSMIT in doc: @ogrisel's remarks
       new  cb0a9d7   ENH oob score as attribute, not separate function.
       new  287f6e3   Revert change on svm.base.
       new  c4be24d   BUG replace assert by custom exceptions
       new  7e33fff   ENH: added oob_score_ and oob_prediction_ to regression ensembles
       new  d21ef4f   Remove if statement.
       new  835439f   Suppress deprecation warnings.
       new  6c07fc9   Merge branch 'master' into gradient_boosting
       new  fecea5f   FIX copy/paste error. guess it was to late
       new  f8ebc29   Merge branch 'warm_start' of github.com:mblondel/scikit-learn into warm_start
       new  0ae85d2   Make sure order="C".
       new  c912ab0   COSMIT refactor SGD code further
       new  8e83572   Revert "COSMIT refactor SGD code further"
       new  58b348e   Merge branch 'warm_start'
       new  11f4c07   Fix doctest.
       new  1b81d3b   Mutual Information docstring incorrectly said it was the adjusted mutual information
       new  015da64   ENH made oob_score an ``__init__`` param as suggested by @agramfort
       new  06ba0b3   Fixed data.filenames consistency issue when 'all' specified for 'subset'.
       new  7d1782c   Added basic test for filenames consistency when all specified.
       new  dca709b   add old version warning
       new  3604565   add newline at file end
       new  cc567ee   DOC what's new, minor doc improvements
       new  9a31dac   Merge pull request #571 from amueller/tree_indices
       new  b28e6e4   ENH: Replace asserts by appropriate errors. Fixes the rest of issue #570.
       new  330c1f3   COSMIT how I love these sphinx errors
       new  3785b2a   DOC complicated objects as parameters confuse sphinx and the reader. Fixes issue #567.
       new  2ce7a79   ENH: Default in Vectorizer "None" as @ogrisel suggested
       new  ebb42e7   preprocessing/__init__.py -> preprocessing/preprocessing.py
       new  c9d27f3   Move preprocessing.py to sklearn/.
       new  54f5425   fixed a doctest
       new  aeb868e   made `sample` also return the sequence of internal hidden states
       new  0022655   FIX: broken build / tests
       new  e4bbaf6   added doc for hmm
       new  e778f7b   variable name changes, using premade functions, doc fixes as per @larsman
       new  58433a7   variable name changes, doc corrections (@amueller's requests)
       new  794a31c   removed unlabeled_identifier, updated tests and examples to reflect this
       new  3fb76f3   corrected example that still refered to unlabeled_identifier
       new  4fa5e29   optimization that stores the spatial index when using knn graphs
       new  4d53343   Merge remote-tracking branch 'larsmans/typesafe-murmurhash'
       new  bd78f63   DOC website: added link to 0.10 docs under support.
       new  2695c7f   DOC added required versions of python, numpy and scipy to install documentation. Closes issue #579
       new  2d6fde0   updated rst docs with kernel information
       new  213f0a2   shuffled digits example, added sensible point colors to plot chart, removed fix for unsupported numpy api
       new  7e837b0   docs describe the different kernels available in techniques
       new  268aed6   TL directory change to push label propagation code into semi_supervised
       new  32efba5   FIX: MurmurHash3 compilation on older GCC
       new  0d54938   Merge branch 'master' into gradient_boosting
       new  8b7b620   test_oob_score_regression oob_score below 0.8 if n_estimators < 50 removed raise exception on underflow error in gradient boosting - random forest test breaks.
       new  cb00e97   COSMIT pep8
       new  59db66e   Refactor in KFold.
       new  24477d4   Cosmit: remove unused imports
       new  5c2a869   COSMIT removed unused imports
       new  ad288a3   changed ``n_iter`` to ``n_estimators`` and attribute ``trees`` to ``estimators``.
       new  16c5ed0   added artificial dataset generator from Hastie et al. 2009, Example 10.2
       new  91a36a3   wip: narrative doc for gradient boosting.
       new  2d6e484   fix: wrong assertion
       new  8648cfe   renamed estimators to estimators_
       new  46ff908   wip: narative documentatio for gradient boosting.
       new  3077ace   fix: import numpy in doctest
       new  81dca03   ENH: Some cleaning in the examples
       new  cbdcb2b   ENH:pep8
       new  77d83b6   Remove CoefSelectTransformerMixin and use SelectorMixin instead.
       new  362852a   Better default threshold for L1-regularized models.
       new  ddc74e5   euclidian_distances is to be deprecated in v0.11.
       new  3b8f54e   Add n_jobs option to pairwise_distances and pairwise_kernels.
       new  e14fb73   MISC: fix bibtex
       new  c22e719   ENH: enforcing skls conventions
       new  9687df4   A pass on the docs
       new  88f4849   turn off old version warning
       new  dcb0d2f   add random_state to LocallyLinearEmbedding
       new  89ad0c7   initialize indices and distances in balltree
       new  bcb129b   check random state in _fit_transform
       new  889075a   Address Issue #590 : use relative path link to about.html
       new  d1dd073   Merge commit 'upstream/master'
       new  6e51321   ball_tree: more efficient array initialization
       new  62f8e03   add info about valgrind to dev documents
       new  423cb6b   Merge pull request #587 from jakevdp/arpack-init
       new  bc35aa3   corrected the doc for dpgmm
       new  7d1e581   Set the download link to PYPI.
       new  f548fce   Merge branch 'enh/metrics' of https://github.com/satra/scikit-learn into metrics
       new  2fa7a09   Backward compatibility in precision, recall and f1-score.
       new  f62398b   Factor some code.
       new  c68f4f5   Merge remote-tracking branch 'upstream/master' into enh/metrics
       new  28a40df   Merge remote-tracking branch 'mblondel/metrics' into enh/metrics
       new  c34bf09   Added a note to the install documentation
       new  689931c   Merge pull request #443 from satra/enh/metrics
       new  1fb883e   Merge pull request #591 from jaquesgrobler/doc_update
       new  de88672   More what's new items.
       new  bc6255e   ENH merge sparse and dense SVMs, part 1
       new  72707ab   ENH merge sparse and dense SVMs, part 2
       new  41ba56f   ENH merge sparse and dense SVMs, part 3: adapt sparse tests
       new  3d12a74   DOC merge sparse and dense SVMs, part 4
       new  6062a2b   Added a note to the contributers documentation
       new  8d066b9   Merge pull request #576 from larsmans/merge-svms
       new  e8a3cb0   Shorted the long line
       new  d582395   Merge pull request #593 from jaquesgrobler/doc_update
       new  a47d184   DOC improve intro to Git in the developers' documentation
       new  65402be   Fix what's news.
       new  e786ac5   Merge branch 'master' into svm_coef_sign
       new  90369a8   DOC comment in linear.cpp
       new  bd4fabc   DOC @ogrisel's suggestion: putting a link to pull request in liblinear.cpp
       new  77f3fec   COSMIT pep8
       new  a1ba062   Merge pull request #588 from jakevdp/balltree-fix
       new  bc776db   DOC fixed doc errors in metrics module
       new  85e986c   COSMIT removed unused imports
       new  5e3ea55   Add Perceptron.
       new  bb1be4e   Add Perceptron to document classification example.
       new  037b641   cosmit in memory debugging doc
       new  6ee1dd4   Minimal documentation.
       new  6fb7243   DOC rm unused param from sparse.ElasticNet docstring
       new  c5aed1d   added __init__.py file to semi_supervised folder
       new  8f0326d   Current version -> Latest version
       new  60cdaa7   Merge commit 'upstream/master' into old-version-warning
       new  16d7857   set warning margins to zero
       new  ca8334a   Updated docs for label propagation, added more technical details about regularization procedure
       new  e585485   specific fine tuning to the label propagation docs
       new  a6fdecd   Merge pull request #546 from amueller/svm_coef_sign
       new  46aa934   Add references and implementation details.
       new  d41f89c   Propagate parameters.
       new  bb7ed64   Expose more parameters.
       new  4af253b   Explain parameter in Hinge loss.
       new  61fd058   Don't rescale coef if not necessary.
       new  1f205e3   Quick note on sparsity.
       new  fe5c85b   ENH: make LassoLarsIC more reproductible
       new  900c97f   doc updates & tweaks
       new  729a4c2   fixed typo in test code
       new  c07a041   added AISTAT ref to docs
       new  a8a4948   added AISTAT ref to rst doc
       new  d23e653   BUG: fix test_precision_recall_curve
       new  ee4dd2e   fixed bug causing error on sparse input data
       new  5464a0b   Don't break API in precision_recall_fscore_support.
       new  ea86b7f   Pep8!
       new  6d8b8a4   Fix scale_C warning.
       new  17ae146   Merge branch 'perceptron' of github.com:mblondel/scikit-learn into perceptron
       new  77ca09e   t -> threshold
       new  b0c119c   ENH: Add randomized lasso
       new  11aaf98   ENH: randomized_lasso example: multiple alpha
       new  f7ca7b3   Better randomized_lasso
       new  6522b82   Jacknife in randomized_lasso
       new  e4919e6   Add a randomized logistic
       new  5bc3f51   COSMIT: pep08
       new  f3dbcb4   ENH: Add pre_dispath to RandomizedLinearModel
       new  3746eb6   ENH: RandomizedLinearModels transformers + memory
       new  b1f5975   fix with new as_float_array
       new  777bb07   STY: pep8
       new  ab4c418   mv randomized_lasso.py randomized_l1.py
       new  0dbd7ad   ENH : some doc + renaming in RandomizedLasso
       new  7ba93da   ENH : better plot_randomized_lasso.py with score path
       new  9c7deea   ENH : prettify plot_randomized_lasso.py
       new  f3423f8   ENH : creating lasso_stability_path func + adding tests on randomized_l1
       new  640400d   ENH : add docstring to RandomizedLogistic
       new  bcf80be   FIX: fix test_randomized_logistic
       new  52e9896   BUG: fix broken merge
       new  c2341b0   s/jacknife/jackknife
       new  0af96c0   STY: s/a/scaling + adding docstring
       new  19331b5   DOC : adding doc for Randomized sparse linear models + fix test
       new  b275030   ENH : adding sample_fraction to lasso_stability_path + add to doc
       new  815725e   typos
       new  66d7b26   cosmit in doc + pep8
       new  53b003d   cosmit in doc
       new  331aa5b   ENH : addressing @ogrisel comments (PEP257, naming, see also)
       new  89a43ce   DOC: rephrase rand linear model doc
       new  45ae9a9   ENH : fix docstrings + add func missing reference
       new  c3a2faa   ENH : center y too in _randomized_lasso
       new  b1ab2ce   ENH : adding support for multiple regularization parameters in RandomizedLinearModel
       new  6fccfef   MISC: removing one XXX
       new  d99e27a   MISC: inherit from BaseClassifier
       new  e1dacc6   BUG: parameter was not set right
       new  95e5350   DOC: Improve feature selection docs
       new  5e840ac   DOC: try to improve randomized lasso example
       new  a47773f   ENH: numerical stability in LassoLarsCV
       new  56f5931   DOC: update dostring
       new  2b1df1f   ENH: grid in terms of alpha/alpha_max
       new  b0b3406   DOC: nicer path
       new  c33756b   DOC: beautify feature_selection docs
       new  41a4863   ENH : early stopping in lasso_stability_path (faster)
       new  afb6dea   ENH : fix legeng of plot_randomized_lasso.py
       new  6ada8de   DOC: cross-reference linear_model and randomized_lasso
       new  2b08a0d   DOC: enrich example docstring.
       new  023d8e3   pep8
       new  51cf742   DOC: better example for randomized lasso
       new  2350402   MISC: make sure two figures hold on a line
       new  db1a212   DOC: example and docs for randomized-lasso
       new  a882bed   MISC: address @ogrisel and @mblondel's comments
       new  5e0f82b   Cosmit
       new  209d23d   Merge remote branch 'upstream/master' into gradient_boosting
       new  42114d9   FIX RandomizedLogisticRegression test import
       new  06d1d3b   COSMIT removed unused import
       new  a77f9ed   DOC fix sphinx errors
       new  2588a3c   DOC more fixes in Docs
       new  9e036a6   MISC: add randomized linear models to what's new
       new  2588335   DOC cluster metrics: fixed see also sections, errors in references section.
       new  b4201fe   BF: Avoid two consecutive centering of the data in outlier_detection.
       new  8d8496a   Add mean_squared_error and deprecate mean_square_error.
       new  8e686be   Don't raise warning when passing explicit scale_C=False.
       new  19fcb70   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  af5037a   DOC: scaling regression targets.
       new  d079dde   FIX: bug in DenseBaseLibSVM when subclasses implement new params
       new  beda3f4   FIX: inheritance in DenseBaseSVM
       new  af34030   Added a small note about the use of an upstream remote in the Contributions documentation
       new  95aa2f2   Shortened a line in the code
       new  860a862   allow for multiple nuggets in gaussian process
       new  c5f36de   example + documentation of gaussian processes on noisy data
       new  83935b3   Merge pull request #602 from jaquesgrobler/doc_remotes_note
       new  613ca44   Merge remote branch 'upstream/master' into gradient_boosting
       new  e79552d   use mean_squared_error
       new  c2eb12f   added new mean_squared_error to metric imports
       new  ee2186f   Merge remote branch 'upstream/master'
       new  7792b36   Merge branch 'master' into gradient_boosting
       new  88d9b09   ENH: use linear gradient cmap for more readable hyperparam heatmap
       new  125aa04   polished narrative documentation. fixed doctest.
       new  e2e7d51   cosmit: fix doc format
       new  6208198   cosmit: fix doc format
       new  aa3a161   COSMIT pep8
       new  3a172f3   FIX SGD loss example for new hinge loss.
       new  46db486   FIX lasso_dense_vs_sparse_data.py example needed update.
       new  1cf592e   BUG: make clone work on 2D arrays
       new  f686cd2   TST: add a test for bug fixed in previous commit
       new  86a3de7   corrected the documentation and add semi-supervised section to the user guide
       new  e3fe334   placed semi-supervised under supervised learning techniques in user guide
       new  dcfc124   COSMIT: make the plot landscape
       new  e4dc6a7   DOC: improve the label_propagation docs
       new  adc5201   COSMIT: authorship and licensing info
       new  010b5d1   Cosmits
       new  7f62f8c   DOC: minor rmk on label_propagation
       new  b774b27   TEST: assert -> nose.tools.assert_equal
       new  b36beb4   Merge branch 'label-propagation'
       new  649bbce   BUG: fix typo in tests
       new  6c6197d   DOC: update whats_new
       new  b1e82a0   BUG: fix tests under numpy 1.5
       new  23f61ae   docstring cosmits and typos in label_propagation.py
       new  3242c6a   removed get_means, set_means, get_weights, set_weights
       new  e1938f9   ENH: renamed plot_gmm_model_selection.py to plot_gmm_selection.py
       new  78e1b6a   fix error in unwhitened case
       new  4f55d52   TEST: add a test for whitening in ICA
       new  fa67019   PEP8
       new  62430eb   ENH: control random state in ICA
       new  b4717b6   Merge commit 'upstream/master' into GPML-fixes
       new  35f9dfc   DOC: expand nugget explanation; combine two GPML examples
       new  8bdf267   COSMIT pep8
       new  02650c4   Merge branch 'WIP_tut', remote-tracking branch 'gaelVaroqueux/stat_tutorial' into WIP_tut
       new  8c490c7   COSMIT abstract base class in univariate feature selection
       new  3acc19e   Fixed the doctests in hmm
       new  a167b36   COSMIT:pep8 in hmm
       new  c217766   DOC add cross_val_score to references, OCD.
       new  19f17b3   BUG: SVM raw_coef_ must be fortran ordered.
       new  1ace675   BUG: missing subpackage svm/sparse on setup.py. fixes issue #559
       new  ed6164d   MISC: cosmit: use subpackage setup.py
       new  0177e7d   Corrected the docs
       new  7c5518b   ENH: changes in the code to fulfill Gaels requirements
       new  91164d0   Merge branch 'master' of github.com:scikit-learn/scikit-learn into gmm-fixes
       new  907f897   ENH:Added back rvs as deprectaed and updated whatsnew.rst
       new  81c3f9f   Merge branch 'master' into gradient_boosting
       new  dc16faf   FIX bug in text feature extraction, issue #606
       new  5e92beb   ENH sublinear tf scaling in TfidfTransformer
       new  ae83ab7   ENH: fixed the GMM docs
       new  b6ad55d   ENF changed INF_EPS to EPS in hmm too.
       new  e61e910   DOC s/with dense data// in merged SGD module
       new  3922816   refactor SGD regression input validation + doc fixes
       new  a87936d   - fixed typo in hmm.rst - pep8 on hmm.py and plot_hmm_sampling
       new  ad50075   DOC: reorganize GMM docs
       new  ac62e91   DOC: reorganize GMM docs
       new  b187b9e   DOC: more examples for DPGMM
       new  7c4f418   Cosmit
       new  4faf20e   MISC: remove custom __repr__
       new  531dcf2   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  c8eb019   BUG: fix doctests
       new  b8d134d   Removed single k-means run to it's own function to enable optional parallelisation later.
       new  1520ad3   Parallel version of k-means.
       new  1f98e35   pep8 and pyflakes tested
       new  b26f03f   Not doing a full sort for getting the best results Fixed docs slightly.
       new  54361dc   Updating random_state inbetween iterations of k-means fixes some issues
       new  e9db8c2   useless imports
       new  c0bb78d   simpler random seeding scheme for parallel kmeans
       new  ee32fda   less hacksih parallel random state seeding
       new  44afb23   Doc updates
       new  0726912   Fixed author reference (removed link as it wasn't working)
       new  1d1c1a5   Added my twitter account as homepage.
       new  20d85cc   ENH: optim hierarchical: heapq in tree traversal
       new  545f321   ENH: hierarchical: speedups in tree cut
       new  08a8252   MISC: clean up old c file
       new  48021c6   MISC: assert -> raise ValueError
       new  0220e18   BUG: typo
       new  71fb671   MISC: fix broken link to example
       new  4aa627a   made `sample` also return the sequence of internal hidden states
       new  3b2338c   rebased to the master and fixed conflicts
       new  eff6a32   bug fixed
       new  ec16c9e   fixed _do_viterbi_pass built-in `reversed` does not return list but and iterator.
       new  2722fea   ENH: parallel in lasso_stability_path
       new  c3d8cd7   - Further integrated tutorial.rst (Section 2 in Userguide) with links to   supervised and unsupervised learning. - Added a tutorial menu page to the user guide as Section 2, making the old   Section 2 a sub-header thereof. - Added the Scikit-learn tutorial of Gael Varoquaux to the tutorial   menu page   NOTE: the scikit-learn tutorial will, once fully integrated and linked up   move down to become point 2.3 as Oliviers tutorial will be merged in   as point 2.1 - NOTE [...]
       new  7fa3485   API univariate_selection: _scores -> scores_
       new  ea5f0f3   COSMIT pep8
       new  4ef2a86   DOC fix sphinx errors
       new  0e83123   ENH: moved class_weight parameter in svms from fit to ``__init__``.
       new  45f189d   fixed doc
       new  cefd76d   fixed typo
       new  b7be62a   factored out weight vector class; dense SGD now uses ``WeightVector`` instead of explicit ndarray and wscale.
       new  36fa40a   MISC Adjusted class_weight param in examples, fixed legend in unbalanced dataset examples.
       new  62c5420   enh: performance of WeightVector now comparable to explicit weight vector. some cosmits in dense sgd extension module.
       new  a620e55   wip: sparse sgd now uses WeightVector - there are some broken tests tough.
       new  81db2e1   Add Satra to the AUTHORS list.
       new  00fbd25   ENH changed naive bayes' self._classes attr to self.classes_
       new  28f9b4c   wip: still hunter sparse sgd bug
       new  1bf502b   fix: forgot to scale by wscale at the end of dot_sparse. All tests are green again!
       new  654a39c   ENH more generic dict-like test in CountVectorizer
       new  94dc280   ENH: update joblib to release 0.6.2: bugfix
       new  d2ee677   added new sgd dataset abstraction to unify sparse and dense implementations.
       new  cc8489b   Merge branch 'master' into sgd-refactoring
       new  f09b881   DOC typos in whats_new
       new  36a4c37   major refactoring of sgd module::
       new  a9b8cc6   Fixed docstring to reflect current code in precision_recall_curve. 1. It has n + 1 values for precision and recall 2. It adds a "fake" point with precision = 1. and recall = 0. Signed-off-by: Brandyn A. White <bwhite at dappervision.com>
       new  15bdcdf   Merge pull request #613 from bwhite/patch-1
       new  231f4e2   MISC: remove joblib from .gitignore
       new  2417504   BUG: add missing file in joblib
       new  8306c1e   DOC typos.
       new  0376895   replaced function call of decode to predict
       new  2cb344c   round numerical results for better stability of the doc tests
       new  efc5bfe   use Py_ssize_t where appropriate; cosmit
       new  24ea7e7   MISC reinserted class_weight as fit parameter, added deprecation warning.
       new  16727b4   MISC cleanup
       new  2c79bdf   DOC margin for old warning wrapper fixed
       new  d63f03f   Merge pull request #6 from amueller/old-version-warning
       new  a5d07cf   fix link in warning
       new  f6deb73   removed pure python codes and beam pruning options
       new  b35c19a   Added change history to what's new
       new  a86afb7   updated author and pep8
       new  ea5d4d5   latest version -> latest stable version
       new  33dc483   modified phrases in what's new
       new  e42bdc1   MISC Deprecated class weights in SGDClassifier
       new  5d363de   Merge pull request #578 from jakevdp/old-version-warning
       new  c385655   pep8
       new  04af4cb   COSMIT pep8
       new  cf2f3b4   use internal dataset fetch + switch to SGDClassifier for faster execution
       new  1d45c77   COSMIT get rid of warning in nosetests for equidistant neighbors. it's intentional.
       new  55bb67c   typo + cosmit
       new  b504ae4   avoid pl.set_cmap and align colors of colormesh with scatter
       new  c7eaf3c   started work on utility function for quick train test split
       new  8666a46   more doctest
       new  1f9f3b3   moved tutorial files into separete folder within main tutorial folder. added folder for section2 tutorial. fixed some links.removed savefigure from plot_cv_diabetes.py
       new  0e5e0e3   Merge remote-tracking branch 'origin/master' into WIP_tut
       new  6f86062   DOC typos
       new  b981f56   DOC typo
       new  bf1cb6b   MISC more sensible NMF test.
       new  9c330c6   COSMIT pep8 wooops thanks @ogrisel
       new  df2a786   MISC forest tests: boston faster, probability test faster and no warning.
       new  275cc33   add parameters in docstring
       new  9e8a904   MISC decision tree test faster and no warning
       new  0fe3117   COSMIT simplified error message checking, remove deprecation warning.
       new  24e2736   DOC: narrative doc for train_test_split
       new  106a6bf   add tests for invalid argument + fixed a type error
       new  1973e3b   more tests
       new  637b22d   MISC more iterations for test_lasso_path. Still runs in <.1s, gives no warning and more accuracy.
       new  2cc634b   MISC more iterations also for test_enet_path, same runtime as before, no warning.
       new  b424c58   typo
       new  14b0759   reworked nested grid search example for better doc and output, use train_test_split and add more cross links
       new  8abda3a   DOC: related improvement in GridSearchCV doc
       new  ba69aba   DOC: more cross references
       new  77b9bdc   cosmit
       new  3de41cf   DOC: what's new
       new  ca2f9ec   Merge pull request #618 from ogrisel/train_test_split
       new  29a302a   COSMIT pep8
       new  5f4ca06   FIX: make LFW data shapes consistent with Olivetti faces
       new  1be1f71   ENH: more informative exception message
       new  f0dc611   API: set scale_C to True by default in libsvm/liblinear models
       new  67c2518   update what's new
       new  d41768a   DOC : add warning in docstrings for scale_C gone in 0.12
       new  ed90bc5   DOC: indent pb
       new  366ceb2   DOC: update scale_C docstrings + add notes to svm.rst
       new  734eaae   ENH : use not(scale_C)
       new  fa26e89   remaining docstring to be updated
       new  7441db8   update docstring with WARNING
       new  d0f7e8a   Merge pull request #601 from agramfort/scale_C_true
       new  ad8e491   TST: use assert_true instead of assert + remove some relative imports
       new  da4651f   COSMIT pep8
       new  93213fe   FIX added missing import
       new  4d652d1   MISC added warning to coordinate descent if alpha=0, don't call cd with alpha=0 in tests.
       new  be62145   MISC replaced deprecated mean_square_error in test.
       new  ed34b09   use Perceptron and quick splitter
       new  33d60cc   BUG: follow API change in example
       new  4017ade   Merge remote branch 'upstream/master' into sgd-refactoring
       new  6f914f3   cosmit: better docstrings for SGD
       new  a71dd04   MISC test for warnings as @ogrisel suggested.
       new  78b2665   DOC: fix a few incoherencies in ridge.py
       new  559dcbb   - added decoder selection - some doc fix
       new  ad5b966   Merge pull request #623 from npinto/ridge-docfix
       new  b5993ee   updgrade to new scikit-learn API
       new  2598f66   make sklearn version dependency explicit
       new  f7fd043   C is now scaled
       new  af521b5   Merge remote branch 'upstream/master' into sgd-refactoring
       new  d1a87f9   COSMIT refactor SGD with Dataset factory function
       new  a04681b   COSMIT rename _mkdataset function in SGD
       new  dfd6bea   fixed some typo, doctest and pep8
       new  88f35b4   added comment on decoder algorithm in the rst doc.
       new  1be8865   DOC: improved SVM docstrings
       new  467f6ab   Merge pull request #620 from amueller/coordinate_decent_alpha_warning
       new  dd1d636   add min_leaf (minimum size of leaf node) to decision tree
       new  614f657   ENH min_leaf for ExtraTree
       new  5d40322   ENH added test for "min_leaf"
       new  c4cf4a9   ENH set min_split if min_leaf is set.
       new  ba7f9a0   typo
       new  3c4ece6   FIX: typo for default init_size in MiniBatchKMeans
       new  e93fbbb   Added tests to check for the correct value of init_size
       new  df17c4e   Merge pull request #628 from daien/master
       new  fb34320   DOC add load_svmlight_file to references
       new  b2eb6b3   DOC minor fixes and typos
       new  c6c011d   DOC more rst fixes....
       new  365f674   DOC typo in whatsnew
       new  868eac1   Merge branch 'master' into svm_class_weights
       new  3647f1f   DOC renamed duplicate label
       new  5e33d10   Merge branch 'master' into WIP_tut
       new  d429b68   feature_extraction/text.py: 'ignore' removed as a default, class param instead with a default of 'strict'
       new  d961bc2   Added a test (that doesn't work yet)
       new  b2a873f   Removed savefig from tutorial plot files. unnested tutorial.rst (section 2) in the tutorial folder
       new  15173d7   ENH: update joblib
       new  71017ff   WeightVector now keeps track of its squared norm.
       new  2d1431d   move WeightVector and Dataset abstraction to new module
       new  d9bdece   moved WeightVector and dataset abstraction to new module
       new  6e4bb8e   updated Dataset imports
       new  ab04eae   no need for sgd_fast header anymore.
       new  add2416   added largescale ext module to setup.py
       new  6297939   fix: declare extension type attributes
       new  64662c3   Test now works, testing both the Word and Char analyzers
       new  eb682ae   decode_error -> charset_error
       new  aa03d66   comment in forest classes for covertype benchmark
       new  435328c   FIX flip sign in decision function of LibSVM in binary case.
       new  4946294   MISC renamed min_split and min_leaf to min_samples_split, min_samples_leaf, added them to the ensemble classifiers and documented them....
       new  ccfba5b   FIX OneClassSVM decision function sign.
       new  0ca3452   ENH more elaborate one class svm testing....
       new  d808c07   docstring update
       new  b8ba71b   MISC address @mbondels comments
       new  8969659   MISC simplified test
       new  c87c168   FIX one class test, added more decision function tests.
       new  0083154   COSMIT pep8 + "leafs" typo.
       new  4e78ea4   Updated tutorial folders in doc with placeholders for other tutorials. updated index.rst for the tutorial menu accordingly
       new  0def213   DOC Added changes to decision functions and coef_ to whatsnew
       new  f4ed07c   Merge pull request #627 from amueller/min_leaf_cherrypick
       new  a7925e9   added an html page for plot_digits_first_image.py
       new  361a058   MISC don't use deprecated mean_square_error
       new  1a7c5ed   Added links to some keywords. Added a small note after the Training set.. topic - linking to the other basic tutorial Fixed a typo in k-Nearest neighbors classifier heading.
       new  10493cf   cosmit
       new  bf77ed0   cosmit undoing (was testing)
       new  c60a1b2   pep8
       new  67af6c7   cosmit to docstrings
       new  4f99b95   Merge pull request #633 from robertlayton/ig
       new  edd7772   Merge pull request #603 from jakevdp/GPML-fixes
       new  4d638a1   Fix a test case
       new  c521c86   Merge pull request #637 from fannix/fix
       new  10e105d   Merge pull request #634 from amueller/svm_decision_function_dirty_fix
       new  9c6e044   FIX #614: raise ValueError at KernelPCA init if fit_inverse_transform and precomputed kernel
       new  448dcd2   Links, image resize and updated ipython code in tutorial Few links added into supervised_learning.rst image in settings.rst now square so doesnt shrink adjacent code box updated ipython code regarding nearest neighbours(after API change in 0.10)
       new  1e06904   Merge branch 'master' into svm_class_weights
       new  4f90efc   Merge pull request #610 from amueller/svm_class_weights
       new  aaa6611   COSMIT pep8
       new  c219d09   FIX whooops sorry
       new  e23908d   Added a dataset image, some links and 'import sklearn' updates Added the iris dataset image to supervised_learning.rst Added html for iris dataset image. Updated plot_iris_classifiers.py to use KNearestNeighbors plot_iris_dataset_3class.py is for the image of the iris dataset - Didn't want to overwrite plot_iris_dataset.py just yet.
       new  dc74f81   DOC Insert hidden toctree, mv "included" files from rst to txt
       new  331eaf3   MISC Issue #639. Remove unused member types in linear_model CVs
       new  d24337b   DOCs change extension from txt to inc, add inc as doctest extension to makefile
       new  a2108ea   ENH: optim in ward_tree
       new  a734166   Cosmit
       new  43acd6a   Added Knn classification example image&html
       new  79ec933   BUG: ShuffleSplit should give reproducible splits
       new  b5a1de7   MISC verbosity parameter for forests: better control over tree building.
       new  ded4327   Merge pull request #641 from amueller/doc_fixes
       new  9c44867   FIX dataset docs: changed suffixes in include to match rename.
       new  19b5e33   DOC fixed inconsistent titles. sphinx didn't like them and didn't show these sections.
       new  a0ce56e   MISC @ogrisels comment about human-parsable counting
       new  0718280   Merge pull request #643 from amueller/forest_logging
       new  07b9208   changed colours of plots, added links
       new  6e59f6c   Change exception text when multiple input features have the same value from: "Multiple X are not allowed" to: "Multiple input features cannot have the same value"
       new  9787396   Fixed link typo
       new  4551288   Merge branch 'master' into WIP_tut
       new  5a7f07a   FIX : fix SVM examples with new scale_C=True
       new  f8904bd   DOC rudimentary docstring to deprecated.__init__ describing "extra"
       new  37aaf41   Simple linear regression example added to tut
       new  b383983   DOC C is pretty large now...
       new  bf83809   MISC class_weights constructor parameter in RidgeCV
       new  0d6e66f   Fixed spelling error,import lines,figures and html links for shrinkage section
       new  5ae4fbc   FIX: make GridSearchCV work with precomputed kernels
       new  d685f20   raise ValueError when given a kernel_function or a non-square kernel matrix + some tests
       new  1dff36a   Fixed a small typo
       new  5f43ada   Added links, images and docstrings to some plot files
       new  d434a39   ENH: small speedups in coordinate descent
       new  10f18e6   Revert "ENH: small speedups in coordinate descent"
       new  51f5680   FIX : Fixed bug in single_source_shortest_path_length in sklearn.utils.graph
       new  4395694   ENH/FIX: in graph shortest path
       new  0cfd283   BUG: fibonacci heap implementation
       new  ae1dea6   TEST: non-regression test for fibonacci heap bug fix
       new  80743ce   Generate c-code with cython 0.15.1
       new  b81b25e   Faster hierarchical cluster for very dense trees
       new  580afa8   ENH: Add the ability to set rho by cross-val
       new  ff10d10   ENH: store the path for rho in ENet
       new  e9eb1bb   BUG: fix tests and reorganize code
       new  5b9318a   ENH: draft of parallel CV in elastic net
       new  8dbb8de   TEST: setting rho with ElasticNetCV
       new  de85be2   DOC: document ElasticNetCV
       new  fcec0ff   DOC: formatting improvement to ensemble.rst
       new  81816ac   FIX : fix ward benchmark
       new  f7645f4   MISC: cosmit to please @agramfort
       new  a055f97   BUG: Same MSE scaling for LassoLarsCV and LassoCV
       new  1dbf4c3   TEST: better tests of LassoCV and LassoLarsCV
       new  ae7ab21   working on tutorial exercices skeleton generation
       new  392db25   fixed plots to have class coloured datapoints
       new  0ba9048   Merge pull request #654 from GaelVaroquaux/enet_cv
       new  f3ec32b   Fixed some figures, added links & corrected SVM Param C explanation -all figures in section 2 of StatInference tut added/fixed -fixed datapoints gettin cut off -all figures link to page with code
       new  ddccc6e   Set label encoding in LabelBinarizer.
       new  5d242ab   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  f9cd715   Guess threshold if not explicitly provided.
       new  932c155   Bug: must be strictly less than.
       new  b071cb0   FIX: make the 20 newsgroups loader explicitly decode latin1 content
       new  23a330a   optimisations to Ridge Regression GCV
       new  56c8ebd   faster GCV for Ridge for n_samples > n_features
       new  1021a82   fixed tests to work with Ridge GCV
       new  037cef6   updated RidgeCV docstring and changelog
       new  acb20ca   fixed bug with _values (thanks @mblondel)
       new  646bf7b   fixed bug with > 1d y arrays
       new  f5ddde0   svd fails for sample_weights, use eig instead
       new  83ecb9b   coerce sparse matrices to dense before SVD
       new  8195db2   refactoring (thanks @GaelVaroquaux, @mblondel)
       new  cbb67aa   Pep8.
       new  36602c6   Don't raise warning in auto mode.
       new  560b0b8   Fixed missing image and GUI download link -Missing image and link for it fixed in basic tutorial -Seperate page for SVM GUI added instead of direct download link
       new  120e628   Image page fixed
       new  a82f7a2   added div.green to the theme for Exersizes in scikit-tutorial
       new  82f3c06   fixed link/updated some code
       new  a435978   DOC: add a link the Gohlke's 64bit windows binaries
       new  858c11e   shorten example a bit with train_test_split
       new  70fa9d5   manually rescale C in face recognition example
       new  0b3981c   renamed file-names, finished model-selection, changed cv plot to use C
       new  6783466   DOC doc fixes
       new  6e2223b   MISC added removal version for scikits.learn deprecation warning.
       new  f8b605d   MISC remove ball_tree and cross_val namespaces
       new  f758445   MISC scikits.learn removal at .12. I'm not so good at counting, sorry.
       new  f534013   Merge pull request #660 from amueller/remove_namespaces
       new  a02d47e   cross_validation.py: fixed bug in text of error message
       new  509bb5d   Merge pull request #664 from conradlee/663-kfold-init-bug
       new  43b962a   Section 4 done - images/links/htmls for images -_plot_pca_3d renamed to plot_pca_3d -plot_pca_3d still needs quiver, but plot works now -plot_lena_ward added to plot ward clustering of lena
       new  c6f619c   All scikit tutorial images and links redone -plot_face_recognition's code was updated to latest face_recognition
       new  f3ed2d7   Fixes for doctests
       new  e9c3570   modified makefile for doctesting - not permanent
       new  25ad1d7   ENH: use shift-invert in spectral clustering
       new  b55d62c   add detailed comment on ARPACK usage
       new  3a5e181   COSMIT renaming scikits.learn to sklearn in some places
       new  540b438   DOC/TEST: HMM fix doc layout and doctest
       new  7417591   Merge remote-tracking branch 'origin/master' into WIP_tut
       new  4479b57   ENH: Add controled random_state in HMMs
       new  7c958d7   DOC: prettify HMM sampling example
       new  bd2405f   Cosmit
       new  bc942e8   COSMIT pep8
       new  1cddb61   MISC Update all the other deprecation warnings that I forgot.
       new  73320c4   Merge branch 'master' into gradient_boosting
       new  51b6faa   renamed and updated covertype benchmark.
       new  05b5e38   uncomment RandomForest
       new  6bf9501   FIX: class_weight only in classifier Ridge classes
       new  52263a6   DOC Documentation for RidgeClassifierCV
       new  5886f7c   DOC add removed docstring.
       new  c597c6b   COSMIT pep8
       new  4508438   ENH Added tests and fixes
       new  a7c1805   DOC remove "for dense data" heading for SVM classes
       new  9061131   Merge branch 'master' into linear_model_class_weights
       new  a510d17   Flatten the feature extraction API
       new  e179df0   Merge branch 'master' of github.com:scikit-learn/scikit-learn into text-feature-extraction-simplification
       new  882dd38   missing C re-scaling in example
       new  11d6116   missing C re-scaling in example
       new  0187838   DOC document classification plot
       new  61b0e28   MISC removed deprecated api from examples
       new  e7c740b   WEBSITE: make example gallery look even better!
       new  2c496e7   MiniBatchSparsePCA and MiniBatchDictionaryLearning still use chunk_size as argument
       new  631705c   merge master
       new  38d4332   factorize feature names array
       new  16fc907   DOC added reference to r2 score
       new  9f6cdc0   make CountVectorizer able to output binary occurrence info
       new  fdb7844   add a test for custom dtype
       new  139283e   ENH rename parameter "multi_class" of LinearSVC to "crammer_singer", add docs, add tests
       new  882a7d5   FIX forgot doctest
       new  3fd3571   DOC minor addition to SVM kernel parameters
       new  e546e17   DOC more readable make_friedman docs....
       new  da9528f   Merge pull request #649 from daien/GridSearchCV_precomputed_kernel
       new  137b4dd   COSMIT don't use deprecated names
       new  40c45ca   ENH new samples generators for classification and clustering. Refactored label propagation example a bit
       new  db1e186   ENH cluster comparison example (starting)
       new  0e8c229   DOC: improve docstring for Vectorizer
       new  28471b6   Flatten the combined vectorizer as well
       new  85950bf   Merge remote-tracking branch 'upstream/master' into text-feature-extraction-simplification
       new  53b9dea   Merge pull request #669 from amueller/example_gallery_css
       new  689d672   ENH added "shadow" parameter class_weight_ as @ogrisel suggested.
       new  06e726c   MISC changed parameter name back but changed semantics, as @mbondel suggested.
       new  b620aa1   Fix grid search example
       new  155ead7   Fix charse in mlcomp example
       new  1c78860   COSMIT pep8
       new  bd58331   DOC added one more sentence about crammer-singer
       new  1a46df1   COSMIT typo. thanks @ogrisel.
       new  88085f0   DOC crammer_singer docstring by @ogrisel
       new  11ba3f7   ENH clustering example with spectral clustering and ward with connectivity. looking better now, still not perfect.
       new  c93c1a2   FIX broke label_probagation example, now fixed it again.
       new  3f2d0dc   Merge branch 'master' into sample_datasets
       new  5a88e0c   Merge pull request #673 from amueller/crammer_singer_rename
       new  98cf508   DOC add new dataset generators to class reference
       new  429cfd7   WEBSITE: another css enhancement to give figures a max width.
       new  667ef37   DOC move references from Notes to References section in docstrings
       new  3a030bb   MISC simplified kpca example with new dataset generator, another minor fix in generator
       new  977f7d5   DOC lasso/enet regression example with coefficient plots, corrected r2 score
       new  ada1520   DOC Basic docstrings for LDA and QDA classes
       new  b5260de   DOC lda/qda examples: remove redundant example, prettyfied other.
       new  70347cc   DOC Added QDA to references, narrative docs, improved docstrings
       new  c7a19aa   COSMIT newline in LDA doc
       new  846e4b5   DOC explanation for plot in lda/qda narrative
       new  41e864e   COSMIT: underscores are better than unseparated words
       new  9bcd00a   MISC use Gaels pretty plot, add dbscan, normalize data...
       new  1dc7dc4   closes #677 - improved affinity propagation docstrings
       new  124b19f   COSMIT cleanup, pep8
       new  a8359d6   ENH issue #661, plus some renaming and minor cleanup
       new  d41a1ea   MISC forbid mle initialization of PCA for n_samples < n_features
       new  30f7b5a   TST: fix trivial bug and control the rng
       new  3b5f328   DOC added clustering example to the docs
       new  adaef32   COSMIT make plot look more like other coef plots
       new  d00c601   MISC: fix the random number generators
       new  4f31269   DOC: started section on text feature extraction
       new  0dd256b   Merge remote-tracking branch 'upstream/master' into text-feature-extraction-simplification
       new  84edda6   COSMIT removed debugging print
       new  842f366   MISC added xlim and ylim for @ogrisel's weird matplotlib ;)
       new  24661d3   ENH fixed seed, added center positions
       new  267015f   switch back to the old vocabulary constructor argument
       new  346506f   Merge pull request #674 from amueller/sample_datasets
       new  02d238f   Merge remote-tracking branch 'upstream/master' into text-feature-extraction-simplification
       new  de25118   better blob seed so that both DBSCAN and meanshift are working well
       new  800e8c0   FIX minor doc fixes
       new  372f03f   DOC add link to narrative in lda and qda references
       new  a202926   Merge branch 'master' into text-feature-extraction-simplification
       new  e1ac7c3   DOC add ``estimate_bandwidth`` utility for MeanShift to the references and narrative
       new  0b9d572   MISC make Ward check if input is sparse.
       new  73c3082   MISC make Ward test if connectivity is a valid connectivity matrix.
       new  863e1b5   COSMIT changed error message for Ward
       new  fb21662   DOC another coefficient plot
       new  1f542c2   finally the right API with plenty of efficient overrides
       new  3f1570e   COSMIT Adjust title for example gallery
       new  10da35d   Merge pull request #679 from amueller/logistic_l1_l2_sample
       new  94da2b7   cosmit
       new  13ea995   ENH 2d plot for l1l2 digits example
       new  4968fef   COSMIT last try to make my plot pretty....
       new  e361207   Merge pull request #2 from GaelVaroquaux/hmmc
       new  a307e45   Filter stop words before ngrams
       new  b7a77e8   demonstrate stop words in example (+ slighly faster convergence)
       new  95946d0   missing sklearn.semi_supervised package in setup.py
       new  c6f3824   remove redundant file
       new  75e2d10   removed redundant file
       new  bd271ab   Better doctest time(wip),removed duplicate examples, update plot_ols.py - I wish to replace the old plot_ols.py example with a new one which   uses the diabetes_dataset instead of synthetic data. - references to changed example paths updated
       new  864e4fb   ENH: remove useless array wrap for feature names + more TF-IDF tests
       new  1216bc5   Make Vectorizer not inherit from TfidfTransformer while preserving direct gridsearchability
       new  24a1b3a   expose 'ls' loss function for classification added new example (GB regularization) fix: min_split -> min_samples_split in examples
       new  7f0048b   cosmit: pep8
       new  0d1daad   FIX: division by zero errors and negative IDF
       new  0ce35bb   DOC: TF-IDF and customizing
       new  5fe1cc4   DOC: updated parameters
       new  0572197   Merge branch 'master' into text-feature-extraction-simplification
       new  1556399   updated whats new
       new  9166a2d   s/Bags/Bag/ and Vector Space Model
       new  8999140   better explanation for bigram features
       new  3872d44   No accent stripping by default + various doc fixes
       new  25a9c4e   update strip_accents in Vectorizer as well
       new  56c186a   typo
       new  7e58310   typo
       new  1fa191f   Merge branch 'master' into sgd-weight-vector
       new  0272892   typos
       new  c686232   remove lambda + better comment position
       new  57e32f4   enable stop words in clustering example
       new  3ea19f0   renamed largescale -> large_scale
       new  0d498c6   Merge remote-tracking branch 'origin/master' into WIP_tut
       new  fc86904   3 files moved into main example pool - links to them updated
       new  aa29428   Merge remote-tracking branch 'upstream/master'
       new  130d87a   fixed error in graphviz export code causing graph error raised with certain valid node configurations
       new  c9d076c   Merge pull request #684 from clayw/graphviz-fix
       new  2a52d97   Test added for multiple-outcome: - linear regression - sparse linear regression
       new  5d0a63d   bugfix: lstsq coefficients output needed to be transposed added: sparse linear regression supports now multiple-outcome
       new  c55214c   fixed spelling error data set generators used instead of custom data sets
       new  3735220   docstring updated and list append replaced with preallocated array
       new  5a2f016   consistency
       new  c987520   spelling
       new  30150cf   Merged some examples into examples folder. - updated links accordingly - renamed miss-spelt excercice to exercise
       new  bca803e   Merged a few examples into the example pool
       new  754788c   typo
       new  db38edb   delete redundant file, merged some examples and updated links
       new  49c657c   examples merged to example pool
       new  97291bd   deleted unused file, tutorial examples folder removed
       new  c920fd8   API: use C=None by default in libsvm/liblinear bindings so (C=1, scale_C=False) which is libsvm default  == (C=None, scale_C=True) which is the scikit default
       new  5507f08   FIX : remove useless C definition in non-fit methods
       new  1caad7d   ENH : adding scaled_C_ attribute
       new  c0541fa   Renamed Vectorizer to TfidfVectorizer + deprecation warning
       new  f59495f   Merge branch 'master' into gradient_boosting
       new  5d0ea6c   updated what's new + backward compat for vocabulary attribute
       new  d9fd053   fixed and inheritance bug in TfidfVectorizer.fit_transform + removed vocabulary backward compat that breaks grid_search
       new  4cca090   replaced silence paramenter in makefile, links removed in stat_learn tutorial, big_toc_css copy deleted, heading changed in tutorial index, tutorial index info added
       new  be89040   useless import
       new  4c58199   Merge branch 'master' into sgd-weight-vector
       new  6c0110a   add support for multi-class
       new  18d6ce1   add todo
       new  7fcb2db   moved WeightVector und SequentialDataset into seperate modules.
       new  ef6e825   re-cythonized
       new  61c3fdb   fix: min_samples_split
       new  c4d7edb   Merge branch 'master' into sgd-weight-vector
       new  f78e1b5   don't need self here.
       new  d4a6ebd   pep8 errors fixed
       new  9b46046   factored out norm updates and moved them to a dedicated subclass
       new  aef85ba   cythonized
       new  436c257   Merge branch 'master' into gradient_boosting
       new  10d641c   Merge branch 'gradient_boosting' of https://github.com/scottblanc/scikit-learn into scottblanc-gradient_boosting
       new  cd0133e   Merge branch 'gradient_boosting' into scottblanc-gradient_boosting
       new  e6fa596   pip8 errors fixed
       new  86f621b   ENH add DictVectorizer
       new  e7a08ad   added ELLIPSIS to 4 examples
       new  17fa297   added ... to ellipsis
       new  4ee5b5e   Merge remote-tracking branch 'origin/master' into WIP_tut
       new  e818a84   cosmit: pep8
       new  d143947   cosmit
       new  6737f3b   added serialization test case
       new  b49c2d0   merged ols and ridge variance + some neating
       new  1b06362   fixed links & neatening
       new  1714f79   Merge pull request #668 from ogrisel/text-feature-extraction-simplification
       new  7a218ff   parallelized
       new  3f5dbe2   ENH test feature_extraction.DictVectorizer
       new  775588e   use `deviance` instead of `medviance` and `bdeviance`
       new  db71fb4   DOC syntax error in DictVectorizer docstring
       new  5f592df   COMPAT turns out collections.Mapping has an iteritems member
       new  7f155e9   moved exercises into seperate folder, neating up
       new  43f1819   ENH add test for DictVectorizer.restrict
       new  0337bde   Merge branch 'hmmc'
       new  2930bfe   TEST: fix doctest on non 64bit boxes
       new  e236ce3   DOC + ENH DictVectorizer: complete docs, add dict_type param
       new  9f9fd2d   COSMIT: readability
       new  03f0887   wip: refactor ``fit_stage``; fix feature importances regression; tests still not green (performance regression on Example 12.7).
       new  7cf907b   trailing whitespace
       new  cef8de7   parameter n_jobs added
       new  f209950   BugFix, matrix was not flagged as sparse.
       new  1dfc768   fix: make binary classification a special case.
       new  db7d0d2   refactoring for multi-class
       new  9aec586   test case for multi-class
       new  e6a5713   comment out - yahoo learning to rank dataset
       new  c72dbd5   BUG fixed error that I introduces earlier: connectivity can also be `None`
       new  1b8b088   DOC fixed reference to an example (that I also broke before)
       new  8920d48   some profiling
       new  493c7cc   path fix of moved figure
       new  5ecc161   FIX: broken doctest under OSX
       new  251fa60   impl. deviance for MultinomialDeviance.
       new  1890d6e   fixed typo,changed 2.2s numbering, fixed 4 examples in exercises
       new  a161f3d   fixed numbering in main User Guide
       new  fdfcf9c   fast tree prediction methods.
       new  6058567   faster ``_predict`` by using low-level tree predict functions.
       new  603eadd   cosmit
       new  6de5944   forgot to remove debug function
       new  b9f62bf   changed self.classes to self.classes_
       new  d340e0e   fix: forgot to rename classes
       new  be82adc   added a StratifiedShuffleSplit in the cross validation schemes
       new  9419451   added test for stratified shuffle split
       new  17ecf4e   updated stratified shuffle split test
       new  68d77d4   fixed sss test
       new  0f21fce   cleanup of arg check and doc update
       new  1c75bd8   put sss validation in external function
       new  1caa9e7   updated doc/whats_new.rst, doc/modules/classes.rst and doc/modules/cross_validation.rst for the sss
       new  50a42f0   sss raises error if a class has only one sample, added associated test
       new  88acd3c   pep8
       new  0f0a5aa   changed train_fraction to train_size
       new  5a79a4c   Fixed random state, changed _validate_sss name, fixed _validate_stratified_shuffle_split bug
       new  d14f7b7   TEST: Fix cross_validation tests
       new  44df9a2   BUG: fix cross_validation on numpy 1.3
       new  de3991a   updated documentation: plots for gradient_boosting, new sample generator
       new  249ea7b   new predict utils for early stopping; updated examples
       new  835af05   Merge remote branch 'upstream/master' into gradient_boosting
       new  5b3cb5c   updated benchmark script
       new  f548b99   delete benchmark scripts - include them in dedicated branch or ml-benchmarks
       new  bc5deea   NearestCentroid classifier, with test suite. Unimplemented shrink parameter, will be done in future commits
       new  3ea18da   Shrink threshold working, along with a test
       new  b064bb9   added collapsable sidebar - still WIP
       new  42b1510   Cosmit typo
       new  eba34de   Collapsable sidebar adding complete - appears to work well
       new  e3da7a4   FIX plot example fix for old matplotlib, so that it shows on the website.
       new  ebc7a6c   Merge branch 'master' of github.com:amueller/scikit-learn
       new  76cb9e9   DOC: Various minor fixes to "Contributing" docs
       new  570b4a3   COSMIT disable liblinear I/O code
       new  c467d54   Deleted redundant files
       new  84ea07d   Skip k-means parallel test on Mac OS X Lion (10.7)
       new  404f24c   Merge pull request #694 from njwilson/skip-kmeans-2-jobs-mac
       new  f9c5458   Merge pull request #692 from njwilson/minor-doc-fixes
       new  72174a7   Had a link to autopep8
       new  4000124   color change for button
       new  9c3e66e   FIX: Delete temporary cache directory
       new  6306c18   Merge pull request #695 from njwilson/tmp-dir-for-cache
       new  f3e4357   BUG: Fix metrics.aux() w/ duplicate values
       new  489b738   Merge pull request #696 from njwilson/issue-691
       new  6eb84f8   Sparse tests, but they are currently failing. Committing for comment
       new  4be3be5   FIX: Add NORMALIZE_WHITESPACE to broken doctest
       new  64c674e   Merge pull request #698 from njwilson/master
       new  dbea9d8   OPTIM: skip buffer unpacking in kmeans
       new  b864eb0   Typo for "neighbours", and converted to en-US
       new  0321ded   Stop passing keyword arguments for positional args
       new  c90a329   Add verbose parameter to SVMs (fixes #250)
       new  40df9b4   Merge pull request #699 from njwilson/issue-250
       new  f53aaa1   comment added to gen_rst. Arrow added to button
       new  dd55e12   Merge pull request #693 from jaquesgrobler/Collapse_Sidebar
       new  3d9c9e6   Test for sparse matrices. Tests fails, my guess is that centroids are the same. Committing to save, will work on later
       new  e932baf   Next button added:position correct,but does nothin
       new  0b15b04   ENH implement one-of-K/one-hot coding in DictVectorizer
       new  c128b14   COSMIT rename DictVectorizer source files
       new  49e2953   ENH optimize DictVectorizer (sparse case)
       new  f326366   TEST more strict test for one-of-K coding in DictVectorizer
       new  edd7748   DOC narrative documentation for DictVectorizer
       new  d5126a7   DOC + pyflakes in DictVectorizer
       new  e05e1a4   ENH reduce memory usage of DictVectorizer.transform in sparse case
       new  9fa061c   COSMIT make cross_validation nosetest slightly more readable and more pep8 respecting
       new  7af5f6a   BUG fix doctests for DictVectorizer (nose 0.X compat)
       new  922dc91   FIX make class weight nosetests work
       new  58148a2   FIX get rid of some doctest errors (with the stricter nosetester)
       new  6c82f70   Fixed bug in nearest_centroid, and removed boston test.
       new  06cda29   Narrative documentation
       new  6981505   Merge remote branch 'upstream/master' into gradient_boosting
       new  86bb509   removed ``store_terminal_region`` from ``build_tree``. removed redundant ``predict_tree`` from ext module.
       new  15e46bf   mention multi-class
       new  9d49982   use ``apply_tree`` to compute terminal region. This is faster and reduces code complexity.
       new  0aa244e   added __all__
       new  4fe5520   enhanced documentation
       new  2359ac0   type (differentiable)
       new  67ac31d   boston -> Boston
       new  dbaabfa   ENH refactoring of dot-file export
       new  c3fb05f   COSMIT comments
       new  4083bbd   COSMIT minor visual enhancement
       new  af3b08a   ENH: don't fail on "yeast" dataset
       new  20b0468   Merge branch 'dictvectorizer'
       new  0d1777c   Merge remote branch 'upstream/master' into sgd-weight-vector
       new  3ac4be8   cleaned some examples
       new  63c7478   un-done NormedWeightVector factorization; performance decrease on RCV1 is neglectable.
       new  19bdd9e   fixed SparsePCA.transform returning NaN for 0 in all samples. (fixes #615)
       new  b4aa308   cythonized sgd files
       new  c831764   Merge branch 'master' into gradient_boosting
       new  3c428f4   TST : add test on ridge shapes for different y shapes
       new  ba7ab75   TST : add test failing test to reproduce #708
       new  eefa816   FIX : fix test for #708
       new  50b516f   FIX : fix test failing with OMP
       new  03443da   ENH: y_mean with consistent shape in _center_data
       new  f5b56fa   Added test for SparsePCA.transform (checks #615)
       new  d770c8f   Merge pull request #711 from davidmarek/sparse_pca
       new  f6a36a3   PEP8
       new  ff09f48   ENH: move _compute_feature_importance into Tree
       new  3509e16   ENH: Use DTYPE instead of float64
       new  ec9ed89   Cosmit
       new  80438a2   DOC: Better doc string for l1_min_C
       new  934d373   ENH: Moved _build_tree into Tree
       new  47a335b   Cosmits + Fix to a test
       new  d101d74   Merge pull request #712 from agramfort/fix_y_center
       new  9011901   button is mostly working
       new  11408e7   spelling fixes
       new  cc2bab9   Revert "ENH: Use DTYPE instead of float64"
       new  a12a8fc   Merge branch 'pprett/gradient_boosting' of https://github.com/glouppe/scikit-learn into glouppe-pprett/gradient_boosting
       new  34f4c14   Sparse tests pass when using shrinkage Code cleaning and optimisation still to go
       new  5628074   cythonized
       new  c5b5a05   added Gilles to authors
       new  76ab911   whats new? Gradient Boosting!
       new  20ae0ed   Merge remote branch 'upstream/master' into gradient_boosting
       new  1a04623   cleaned up
       new  4bf8262   more cleaning-finished off
       new  8f9e2a9   added util func to create random sample_masks
       new  a52a526   use random_sample_mask (issue pointed out by @glouppe); rename train_deviance and oob_deviance to train_score_ and oob_score_; doc attributes
       new  0dc117a   update examples
       new  6c09415   update tests
       new  6079a8d   remove np.seterr
       new  98310f5   closes #703 - KFold has now an option to shuffle the data
       new  e4c8911   Added unit test for shuffle option in KFold
       new  8ce9396   Now tests the randomness of the KFolds when shuffle is True, and that all indices are returned in the different test folds
       new  e660ab3   FIX : prevent ICA with defined n_camponents and whiten=False (fix for #697)
       new  67adfaa   TST: capture warning in test
       new  66d8f4e   spelling errors,edit curse of dimensionality, explain top-down
       new  94088b8   bug fix - layout
       new  baaeafa   changed hover colours for button
       new  ed6f1fb   previous button added with hovering-effect
       new  0fe20c3   DOC Added clustering functions to references.
       new  2119eb1   Merge pull request #685 from ibayer/master
       new  9fe4408   FIX : use joblib from externals
       new  6cb7d84   cosmit: comments + rm unnecessary variables
       new  9ef2a0f   cosmit: add comment to replace ``random_sample_mask`` if numpy requirement allows to do so
       new  ceb8583   Merge pull request #709 from ibayer/cleanExamples
       new  22880a4   cosmit: fix ClassPriorPredictor docstring; rm comment
       new  b96bb4b   typos
       new  eb0b27c   typo
       new  fb4c81b   ENH local variable in ``fit`` instead of modifying the estimator parameters. thanks @GaelVaroquaux
       new  c8d3617   Merge pull request #705 from agramfort/fix_ica
       new  4cc8641   Merge pull request #714 from jaquesgrobler/Next_button
       new  c4fb366   mv *Predictor to *Estimator
       new  85d333f   mv classification init estimators; use np.bincount for PriorProbabilityEstimator.
       new  8babc3d   is_multi_class now is a class attribute.
       new  057caef   update docs
       new  923a759   don't need to store n_classes.
       new  0d63d78   Merge branch 'master' into WIP_tut
       new  6c7a166   fixed new doc-test error
       new  9f413c8   cosmit: no need for float literals
       new  d452cde   Merge branch 'master' into gradient_boosting
       new  5964664   combat for sp_linalg.lsqr
       new  0b08145   Merge branch 'shuffle_kfold' of https://github.com/NelleV/scikit-learn into kfold-shuffle
       new  24b80e3   Test indices=False case.
       new  1bb5a3c   Factor tests.
       new  af3a547   Merge branch 'combat' of https://github.com/ibayer/scikit-learn into lsqr_fix
       new  3b81dd9   Fix lsqr for scipy 0.7.
       new  ef19861   DOC: Added ElllipticEnvelop to the References
       new  a7b66ba   FIX: return; instead of return NULL;
       new  741b047   MISC: better verbosity in lars
       new  ff1ca34   DOC added reference for EllipticEnvelop and fixed some sphinx errors.
       new  0a4f976   Merge pull request #717 from jaquesgrobler/Issue714
       new  9672670   typo + cosmetics
       new  d07f946   FIXed nosetests. Thanks @pprett
       new  5a2681c   Merge pull request #707 from amueller/graphviz_dot_refactoring
       new  57e8147   Merge pull request #648 from amueller/linear_model_class_weights
       new  da20633   DOC: more visible version remark
       new  118c00d   ENH Ward: better behavior for non-fully-connected graphs
       new  e2f9c2a   ENH: Don't modify connectivity unless specified
       new  f0d7eb6   DOC: affinity-propagation in clustering comparison
       new  cbc9fa8   DOC: add clustering example on front page
       new  e9d6673   ENH: a few typos in docstrings
       new  a7520fb   Merge pull request #726 from emmanuelle/doc_correction
       new  e64b7a6   ENH: summary table on clustering
       new  c1eeb51   DOC: better clustering comparison table
       new  c2b356b   DOC clustering comparison: link table and figure
       new  df558e8   MISC: tweak example layout
       new  14f37d4   DOC: finish table to compare clustering
       new  e38fa17   Merge branch 'WIP_tut'
       new  43ca4b7   DOC: Better narrative for DBSCAN
       new  7278dfe   DOC: finish misc in tutorial
       new  6b1bc17   BUG: no plotting in doctests
       new  cc1765c   COSMIT: layout tweak
       new  73588be   Redo CSS layout killed by commut 94088b81
       new  2edfa60   BUG: fix doctests
       new  a247a3e   point out scalability problem with large numbers of classes; use ``n_classes`` instead of ``K``
       new  1bd5e83   cosmit; mention scalability issues w.r.t. large number of classes
       new  f62966d   COSMIT Typo
       new  91b42dc   Turned on final test (it works!)
       new  42e691b   Broadcasting used to remove a loop
       new  91d3797   Removed asserts in code
       new  72916f6   Test use assert_array_equal where appropriate
       new  707da6a   pyflakes on test
       new  6c1333b   Update to documentation
       new  f5179b1   Moved to the `neighbors` namespace
       new  fd30ddc   Example of nearest neighbor, getting an improvement when using a shrink threshold of 0.1
       new  4168ab5   Explain example in docs
       new  bc01881   fix: convert input arrays to float
       new  efe14a7   fix: force copy to True in case underlying default behavior changes.
       new  239039c   tst: added test for feature selection. this test would have failed in the previous case. closes #727
       new  f1c6f5e   Merge pull request #728 from satra/fix/f_regression
       new  f56ffb5   Made old EllipticEnvelop deprecated class -renamed to **EllipticEnvelope** -Examples updated
       new  69350f1   changed message to *Use EllipticEnvelope instead*
       new  87fda34   Merge pull request #730 from jaquesgrobler/rename_EllipticEnvelope
       new  46c2c3f   corrected Gaussian naive-bayes to correctly computer the class priors
       new  58e15bc   Update examples/neighbors/plot_nearest_centroid.py
       new  399fcff   Update doc/whats_new.rst
       new  2556fa1   Update doc/whats_new.rst
       new  f5de79a   Removed unneeded numpy.array call in test Added NearestCentroid to 20 newsgroups example
       new  d213168   metric fixed in tests
       new  8b76679   Merge remote-tracking branch 'origin/nearest_centroids' into nearest_centroids
       new  77c844a   DOC: timings in cluster comparison example
       new  2784be9   COSMIT: prettier plot
       new  efd467c   Fixed broken image link
       new  446a225   Merge pull request #733 from jaquesgrobler/master
       new  0ae685f   COSMIT simplify input validation in KMeans
       new  91a7b65   DOC small fixes to NearestCentroid classifier
       new  9e21adf   Merge branch 'master' of https://github.com/udi/scikit-learn into udi-master
       new  ea442a3   added prior test
       new  32c32db   more test cases for naive bayes
       new  f7f6849   GaussianNB: use epsilon to overcome zero sigma problem. GuassianNB: check input size
       new  3511c90   BUG disallow shrinking with sparse data in NearestCentroid
       new  7d00051   rm print stmt
       new  82cc2d0   BENCHMARK covertype: select classifier via cmd line opt
       new  46fcc12   Merge pull request #5 from larsmans/nearest_centroids
       new  e6c2be9   Merge pull request #736 from paolo-losi/bench_covtype
       new  88cf3c5   DOC: misc wording
       new  befad3c   TEST GNB: test that class_prior sum to 1
       new  98aef42   Add test for grid search with only one grid point.
       new  bbc2499   Check param grid.
       new  3f6e67a   Return early if there's only one grid point.
       new  ca3283b   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  bea9264   added gbrt extension module (faster prediction methods)
       new  de7ed14   rm custom regression tree prediction method
       new  5467dac   faster prediction methods
       new  f349e42   wip
       new  00e17f2   add prediction method for specific stage
       new  11aa107   add staged predict
       new  4b5aada   use staged predict in gbrt examples
       new  eb9ad47   ENH : speed up f_regression
       new  2d5c2a7   FIX : array copy for compat pb
       new  215eca7   DOC typos, line-width and minor stylistic fixes in pipeline module
       new  8a6fe20   COSMIT shallow copy of steps in Pipeline + code style
       new  11dd55b   fast tree prediction based on mystic cython kung-fu
       new  92b8e86   cosmit
       new  54cf09f   staged_predict for regression
       new  c5dc1f6   test for staged predict and cosmit
       new  5ace41b   more test cases
       new  faf035d   more test cases (input check at prediction time, degenerate inputs)
       new  d1a122e   use approriate data types (Py_ssize_t)
       new  b82c76e   better input checks at prediciton time
       new  8efc6d2   ENH: Added p to classes in sklearn.neighbors
       new  e761d69   TEST: tested different p values in nearest neighbors
       new  db30c2b   DOC: Documented p value in nearest neighbors
       new  a26cd5c   DOC: Added mention of Minkowski metrics to nearest neighbors.
       new  a8b5110   Fix doctest failure.
       new  4ac7a41   Merge branch 'nearest_centroids' of https://github.com/robertlayton/scikit-learn into nearest_centroids
       new  6901d7e   Fix doc mistakes.
       new  00eff77   Precomputed distance matrices can be rectangular.
       new  cf811d2   Add test for precomputed distance.
       new  4685bab   Doc cosmits.
       new  3e391c8   COSMIT pep8
       new  8474e94   DOC sphinx/rst errors
       new  705a368   DOC Believe it or not - this fixes the annoying sphinx error. And don't dare to abbreviate singular value decomposition with svd. Sphinx does not approve of that.
       new  d0fcb18   COSMIT minor fixes to docs
       new  f1ffee6   COSMIT fixed references to covariance.EllipticEnvelop in docs
       new  abc6a02   FIX+TEST: Special case nearest neighbors for p = np.inf
       new  ded3a8d   FIX: pep8
       new  3101cb4   test for positive constrained lasso added
       new  a78715a   positive constrained option for lasso added
       new  9aaf711   lasso docstring update
       new  76f34bb   remove outcommented lines
       new  2509de2   wording
       new  7df65a5   example for lasso with positive constraint
       new  8fbdde0   renaming
       new  ce0f7cb   reset wrongly committed file
       new  be11f8f   use scikit function to make train test split
       new  2e4d8db   set w[ii] = 0 if tmp > 0
       new  a7215af   - changed parameter from positive_constraint to positive - moved parameter from fit to constructor
       new  e4ce5cf   ENH: sort features in dict vectorizer + new doc
       new  9711ef7   indent
       new  83e1ead   add examples for positive constraint lasso and enet
       new  0df07e3   merged into plot_lasso_coordinate_descent_path
       new  759bcf4   fix doctest
       new  81c222f   ENH: refactored the HMM tests to ease PY3K transition
       new  1b0b500   Fix bad reference to LFW in example
       new  b7b7b62   fixed doctest
       new  40f2132   rm old tree prediction methods; cosmit
       new  b35aa6f   cosmit
       new  8588a1e   Merge branch 'gradient-boosting-enh2'
       new  588bb7d   Merge pull request #741 from ogrisel/sorted-dictvectorizer
       new  28e012c   COSMIT use sorted instead of list.sort in DictVectorizer
       new  18556c1   ENH: Use squared euclidean distance for p = 2
       new  eadb1f5   DOC small fixes to DictVectorizer documentation
       new  ca50d31   Removed `_plot` from the face recognition example updated tutorial page to use static image.
       new  987f0fb   Merge pull request #751 from jaquesgrobler/master
       new  8e4ece7   useless import
       new  0c04e8f   FIX #752: raise explict ValueError if k is too large
       new  9748565   FIX: missing string formating argument in MBKMeans error message
       new  cb76a77   BUG fix issue #753, "Sparse OneClassSVM missing argument to super()"
       new  2c293eb   BUG re-allow zero-based indexes in SVMlight files
       new  3a0414e   removed useless assert
       new  bdcb017   Merge pull request #748 from ogrisel/hmm-test-hierarchy-simplification
       new  7a87d7f   Merge pull request #742 from davidmarek/pdistance
       new  32f9a7c   COSMIT replace utils.testing.assert_in with Nose-compatible functions
       new  4861dd8   DOC: Manhattan distance == l1 norm
       new  ef6be04   COSMIT pep8
       new  fb7b5b7   DOC correct links to face recognition example, take care of trailing underscores.
       new  ef62a60   DOC + FIX DictVectorizer: actually support single Mapping arg in transform
       new  2bbc531   FIX : missing self.copy = copy in PLS GH Issue #758
       new  5b93ccd   cosmit : consistent linestyle in plot_lasso_coordinate_descent_path
       new  1926152   ENH : add duality gap check with Lasso(positive=True)
       new  3613180   Merge pull request #1 from agramfort/posCoeff
       new  6d6c9a8   ENH zero_based="auto" support + better n_features=None in load_svmlight_files
       new  370a0b5   FIX: avoid dividing by zero in Tree.compute_importances
       new  fb4bedf   Merge pull request #747 from ibayer/posCoeff
       new  7804a83   ENH: parallel computation of X_argsort
       new  7f7113e   ENH: better argsort
       new  7806bb0   ENH: cosmit and doc
       new  d8ce359   COSMIT vanity + license for ArrayBuilder
       new  2ad9115   Merge pull request #761 from glouppe/master
       new  e003185   ENH: add verbose option to LinearSVC
       new  4cdf513   BUG: fix LibLinear verbosity for L2R_L2_SVC
       new  f172c7e   MISC: verbose should be int, not bool
       new  2d80ea6   TST: add smoke test for LinearSVC's verbose option
       new  7e4ea9e   BUG fix LinearSVM doctest
       new  4956c00   MISC: verbosity in SVMs
       new  097d85e   add test for multiple fits w/ different input shapes
       new  6ef572c   fix issue 762: SGDRegressor does not clear coef_ from previous fit
       new  b68e6be   asarray not needed because of check_arrays stmt above
       new  5d2d7d1   rm unused vars
       new  87bda9c   Merge branch 'fix-issue-762'
       new  a5a02d4   COSMIT pep8
       new  38cdfb2   ENH grid_search forgets estimators
       new  0295809   DOC slightly better docs for ``refit``, document ``best_params``.
       new  db57c39   FIX clone base_clf before setting params.
       new  f33403c   FIX messed up something in the short cut method.
       new  a633b28   ENH pre_dispatch for foresters
       new  a6f0e2a   FIX redundant code is redundant
       new  6f1c384   Merge pull request #773 from amueller/forest_pre_dispatch
       new  10928c3   typo: Viola-Jones
       new  d6c29c0   Gradient Boosting also provided OOB estimates
       new  c2eeca7   COSMIT refactor SVMlight loader
       new  806bb97   ENH fit_predict convenience method on KMeans and MiniBatchKMeans
       new  e7a33b2   FIX: #774 Add documentation for lprun config in qtconsole and notebook
       new  29772fd   Merge pull request #729 from larsmans/fit-predict
       new  2ddf1a7   COSMIT pep8 SVMlight loader
       new  7b4842d   BUG close files in time in SVMlight loader (with statement)
       new  96af5ae   TEST + FIX zero_based="auto" behavior in SVMlight loader
       new  2584c25   DOCFIX: make math block render
       new  d38f4d0   DOC + PEP8 SVMlight loader
       new  31546a1   COSMIT add todo comment to grep
       new  d7f86f5   Merge pull request #770 from amueller/oblivious_grid_search
       new  7bc2ca7   Merge pull request #756 from larsmans/svmlight_fix
       new  1894252   ENH normalized_mutual_information
       new  eb09cec   Fix bug when refit=False.
       new  9f3369b   Revert "Merge pull request #773 from amueller/forest_pre_dispatch"
       new  4ae7551   Added the name change for the recent change EllipticEnvelope
       new  5942665   Merge pull request #782 from jaquesgrobler/Update_Changelog
       new  5f6d012   Fix math notation for exp and tanh.
       new  f2a7a7d   Add pointer to kernel equations from SVC docstring.
       new  40745ce   Rephrased narrative doc reference in docstring.
       new  f402556   Added RST comment about where to find narrative docs.
       new  fb2453d   Merge pull request #783 from dwf/svm_docs_minor
       new  3970de3   DOC typo
       new  bf301f8   change web site for agramfort
       new  43571bc   FIX : fix SVC pickle with callable kernel
       new  a13138a   cosmit
       new  e00044c   FIX : callable kernel for prediction
       new  89d73c7   FIX : sparse SVC clone with callable kernel
       new  b3ca4ee   ENH: train_size and test_size in ShuffleSplit (#721)
       new  3c64e1b   TEST: Added more tests for ShuffleSplit
       new  324abdd   TEST: Tested ShuffleSplit with different types of test_size
       new  1503bd7   Changed deprecation warning.
       new  58a8742   DOC: Added changes in ShuffleSplit and sklearn.neighbors
       new  95aba96   Error checking now works for more types than just int and float.
       new  f987efb   Use numpy dtype.kind instead of isinstance
       new  f18a5f4   TEST: assert_equal instead of assert
       new  0040a41   ENH: use warning.catch_warnings
       new  94c2c3a   ENH: neighbor warning always raised
       new  0adaacc   API: n_test -> test_size in Bootstrap
       new  d19abab   Changed GMM's API to suite rest of sklearn - parameters not to be set via 'fit' anymore - parameters are to be set in __init__ - previous way of doing it set as deprecated with warning
       new  c9ccf24   1.Fixed typo 2.Removed has_key entries
       new  436589f   restored last changes
       new  691d1e1   Fixed syntax error
       new  8ba6995   mixture/plot_gmm* examples updated  * plot_gmm_sin will get another update when    DPGMM and VBGMM get their API's updated too
       new  d2ed2db   restored last changes
       new  7752db8   DPGMM API updated, along with plot_gmm_sin example
       new  da56baa   DPGMM and VBGMM API change, example updated
       new  a3eb84d   modified test_gmm to match API changes in gmm.py
       new  51ddf4a   updated documentation for gmm,dpgmm and vbgmm
       new  3fb9333   Changed variable name `x` to `covar_type`
       new  2f3c9ea   Updated `whats_new.rst` with API change
       new  36418a9   COSMITs on GGM
       new  55959ed   COSMIT don't use deprecated attributes in tutorial.
       new  43d1aa4   COSMIT pep8
       new  1ee61ea   FIX don't use parameters to fit in GMMHMM.
       new  baab5f1   FIX don't use Python 2.5 method of checking for warnings
       new  29826ba   MISC Don't warn on equidistant on iris. iris has duplicate datapoints.
       new  61fb19e   FIX don't use fit parameters in grid_search test
       new  56b3167   COSMIT pep8 document classification example
       new  7613c48   ENH convert X to float in k_means predict.
       new  c7bec7a   DOC typo in example
       new  b8e40d3   DOC clarify zero_one_score
       new  42e33f3   Merge pull request #796 from amueller/kmeans_dtype
       new  80e6ec2   MISC don't use private ``set_params`` method as that raises a warning.
       new  a0498b6   MISC don't use iris in testing as it has duplicate data entries. Add some noise to simple examples.
       new  da32a0a   Check that scikit-learn implementation of PLS provides exactly the same outcomes than 3 R-packages: mixOmics, plspm and pls. By the way ensure that PLSRegression implements *exactly* the PLS2 algo. and  that PLSCanonical implements *exactly* the Wold algo.
       new  ae5fc4c   TEST: Fix doctest
       new  e7c4acc   MISC added note that we need better tests
       new  68e6c12   DOC typo
       new  0e8e38e   DOC typo
       new  072cfbb   ENH check if backport of sparse scipy ARPACK is needed. The backport breaks with scipy 0.11
       new  65a70f7   Added mutual_info_score to the references
       new  baea563   DOC narrative docs for normalized_mutual_info_score
       new  94fee95   DOC make formulars for clustering metrics more pleasing to the eye
       new  5cbd197   ENH fix if entropy is zero in normalized_mutual_info_score
       new  958cfbe   COSMIT cleanup + pep8 in examples
       new  b22aa4f   MISC extended example, fixed doc build warning
       new  b73d925   DOC made it more explicit that AMI is better than NMI
       new  692e07d   revert PLS param rename + move input validation out of loop
       new  dcc1c74   COSMIT + MISC pep8, pyflakes, typos and some other cleanup of examples.
       new  2453600   DOC typos (thanks @ogrisel) and some elaboration in docstring.
       new  079a5d4   Merge pull request #800 from amueller/less_neighbors_warnings
       new  0747f47   FIXed pca example that I broke when "cleaning up"
       new  b22eb2a   BUG chi² feature selection didn't work for COO matrices
       new  b17a820   ENH export f_oneway from feature_selection module
       new  258f193   BUG ensure that SelectKBest actually selects k features
       new  5cf3b4c   ENH checked for scipy version
       new  be59bf9   Fix kernel pca example.
       new  9b4be76   FIX #807: non regression test for KPCA on make_circles dataset
       new  5f4b320   Fixed typo in documentation
       new  71aa65e   Merge pull request #809 from zaxtax/master
       new  ff4d216   Fix doctest in PLS.
       new  7b75227   Rename "p" to "espilon".
       new  6204829   Allow regression losses for classification.
       new  3e762a0   Add epsilon-insensitive loss.
       new  288a247   ENH add ``decision_function`` to ``Pipeline``
       new  1c6797b   Fixed some typos
       new  d43e8e7   Update examples/exercises/plot_iris_exercise.py
       new  09c9dbe   Merge pull request #814 from invisibleroads/master
       new  cf5ac69   Merge pull request #813 from invisibleroads/patch-1
       new  8fc6a5c   ENH joined tests for less duplication, checked shapes as @ogrisel suggested.
       new  62f4c8c   predict_proba with loss="modified_huber".
       new  751da55   Update doc.
       new  04cecd2   Doc: predict_proba.
       new  9189efa   What's new.
       new  292f68d   Document API change.
       new  e1fbe35   Easier to understand formula.
       new  504956a   FIX : make plot_ica_vs_pca.py deterministic (fix for #815)
       new  b6f8fe7   Merge pull request #802 from amueller/arpack_backports
       new  cf6de05   Merge pull request #812 from amueller/pipeline_decision_function
       new  87389f5   FIX we need to do "LooseVersion" to support dev/git versions of scipy
       new  4e19dc3   typo
       new  c551504   Add note for port install py27-scikits-learn
       new  b11c2bd   COSMIT pep8
       new  978ba44   COSMIT make test more explicit
       new  33f758d   COSMIT removed unused "verbose" option in dbscan
       new  b75833f   COSMIT removed unused import in test
       new  eba29c2   FIX copy/paste error
       new  67dc16d   Cosmit: comment on 'clever' code
       new  83d9d61   FIX removed verbose also from main DBSCAN class
       new  cb132cd   Warn: Passing params to fit is depreciated
       new  ac63d30   DOC added reference to Hila's thesis, added comment about equivalence.
       new  51c49c9   ENH replaced v_measure_score computation with nmi computation.
       new  0ac3c65   DOC removed NMI from example plot as it is the same as V-measure
       new  44ff58a   COSMIT dbscan test doesn't use fit params
       new  c7d5293   DOC comment on normalized mutual information
       new  c8f8b9d   DOC: testing without sklearn.test()
       new  60e882f   COSMIT: macports package name
       new  a97e6c8   ENH simplified entropy calculation
       new  91fa192   Revert "DOC removed NMI from example plot as it is the same as V-measure"
       new  cfafb83   Revert "Revert "DOC removed NMI from example plot as it is the same as V-measure""
       new  a494b30   Revert "ENH replaced v_measure_score computation with nmi computation."
       new  c0e46e1   COSMIT typos by `git grep independant`
       new  0c7a06b   DOC corrected relation of V-measure to normalized mutual information.
       new  36fadb2   COSMIT: better warnings
       new  f150b0c   ENH MiniBatchKMEans: increase init_size for large k
       new  33006bc   MISC removed unused lines, see #666.
       new  2a593ce   DOC: better description of init_size
       new  d829f7d   DOC create example section for datasets
       new  f47d910   DOC title for the tutorial examples
       new  6f3cc74   EXMPL: fix legend in sgd sample weights
       new  f921e8e   COSMIT rst in example
       new  3698317   Add warnings and clean up tests
       new  54ebafc   COSMIT we no longer support Py 2.5
       new  645db95   ENH adjusted examples to new matplotlib 1.1.1
       new  e749771   MISC don't use ``set_cmap``
       new  3b3dc2e   MISC use logsumexp in DPGMM for less warnings
       new  e924f2e   FIX typos in examples
       new  05e4105   typo
       new  ca28fde   fix for #824
       new  3ec174e   FIX one more example
       new  66d00f4   MISC trying to remove scale_C
       new  4286188   MISC forgot two
       new  5edfa06   DOC docs and examples have scale_C removed
       new  77954f2   FIXed many tests
       new  0a4db5c   DOC : update SVM examples with scale_C
       new  1545526   API : change back default C to 1. explicitely and epsilon 0.1
       new  4fbe9c6   FIX : svm decision function test
       new  3885dd8   FIX: doctests for scale_C, took some liberties
       new  833fc05   COSMIT simplify a bit examples
       new  a721593   DOC: restructure what new
       new  060669e   BUG: explicit adding of libm at build
       new  3c2e01f   DOC some doc corrections
       new  123d70d   BUG test_oneclass_decision_function: fix RNG
       new  a707c34   ENH remove duplicate definition of "assert_lower" in tests
       new  0f57bc0   COSMIT: no capitals outside of class names
       new  da022b8   FIX ditto (numbers are to random)
       new  8a354dd   COSMIT: remove print
       new  690aaf1   - fix missing links to the C math libray
       new  b32180d   BUILD: add libm onlyon posix systems
       new  6ee1274   Radius-based classifier now raises exception, if no neighbors found
       new  f063837   Corrected some mistakes, added optional outlier_label parameter, which can be given to outliers
       new  5013dd7   Fixed weight calculation from distances (1. / dist), and weight function in tests (lamda d : d ** -2)
       new  fab8344   MISC: simpler faster code with vectorization
       new  2ce92a0   SPD: Minor speedups
       new  c919b7f   SPD: minor speedups
       new  3e5f2e5   ENH backport "assert_less" and "assert_greater", rename "assert_lower" and use it everywhere :)
       new  5ad3bc6   ENH rename out_dim to  n_components in manifold module
       new  e629d08   FIX: handle deprecation with estimator API
       new  104fe38   BUG: fix assert_greater/assert_lower
       new  1cf2faa   BUG: fix assert_greater
       new  acac958   FIX assert_greater message
       new  816bb28   DOC Added pipeline user guide
       new  5c464b8   BUG: fix doctests
       new  6c42e9d   DOC: cosmits in docs
       new  eeb6dcc   ENH use random states everywhere, never call np.random.
       new  31c6745   FIX don't do anything in the __init__
       new  fead9ac   COSMIT: only classes should have capitals
       new  5bf5266   ENH: make LinearSVC copyiable
       new  1eae414   TST: do not raise warnings in sklearn.test()
       new  df3dac7   BUG: fix testing on older numpy
       new  b5f1a5e   WEB Added page with links to various tutorials/presentations on scikit-learn
       new  aa2c839   DOC added some explanation to video page
       new  bae4e41   DOC: cosmits on tutorials and videos
       new  e75dbc9   DOC: wording of whats_new
       new  23c042e   ENH added random_state to Gaussian Process
       new  caea207   BUG: use permutation rather than shuffle
       new  34c2904   FIX testing: random state problem in forest testing.
       new  d1e34dd   CLEAN sparse_encode: remove unused arguments
       new  65abbb6   ENH: avoid an underflow
       new  5b29e82   Revert "ENH: avoid an underflow"
       new  4e3f008   DOC minor fixes to rst and image paths
       new  92ab122   DOC banner 14 duplication?
       new  87f06b7   DOC more minor fixes
       new  c2bc420   DOC: instructions on testing
       new  e6ecf57   DOC fix last docstring error. Don't remove redundant docstring. I dare you, I double dare you mother******!
       new  b93105e   DOC: faster and more meaningful example
       new  ed58083   ENH: prevent multiprocessing in tests under Windows
       new  22f2da2   DOC: avoid 2 rows of images
       new  25112d0   DOC: more readable title
       new  7b34fba   DOC: Feature extraction vs feature selection
       new  ae1cfd5   DOC: image to graph utilities
       new  2b287c3   FIX: bug in test_setup. Actually avoid multiprocessing now.
       new  9d6a0b0   ENH: update joblib
       new  6a1f2b2   FIX: wrong cover-package, misleading coverage as 100%
       new  73fdf6a   RELEASE 0.11
       new  7806572   COSMIT typo in whatsnew
       new  67cad73   RELEASE HEAD is now 0.12-git
       new  7a6af66   COSMIT typo in whatsnew
       new  13028c5   BUG: remove n_jobs=-1 from examples
       new  42c2ce4   BUG: remove n_jobs=-1 from examples
       new  5fed64f   DOC: updated testing instructions
       new  2652a74   Merge branch 'install-windows' of https://github.com/vene/scikit-learn
       new  4ae44b0   DOC: updated testing instructions
       new  5b136ab   Merge tag '0.11' (theirs) into releases
       new  a16fc5d   Merge branch 'releases' into dfsg
       new  a749af9   Merge branch 'dfsg' into debian
       new  39efa66   Initial changelog for 0.11.0-1
       new  c48e0ec   ENH: adjusted Format in debian/copyright
       new  7d72558   Adjusted patches/deb_use_system_joblib to avoid submodule import
       new  726b026   Made running unittests verbose
       new  f6d7f45   FIX: control RNG seeds in ICA tests
       new  2114142   DOC: fix rst layout
       new  cd74047   MISC: clean up top-level namespace
       new  b2d1517   added patch up_ICA_test_seeding to "cherry-pick" f6d7f45a45d21a779d1a2d59a6f7ff30de83b76e (FIX: control RNG seeds in ICA tests)
       new  7eb39fa   exclude test_sparse_svc_clone_with_callable_kernel from tests
       new  91293a3   DOC LabelBinarizer
       new  207cd79   BUG: now build works.
       new  9d1b59c   DOC clarify __check_build messages
       new  1b41ccb   DOC: TODO items
       new  c9250d4   Some more non regression test on PLS
       new  94a604b   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b842bb7   BUG: price accidentally used instead of volume
       new  bd23d74   Add LabelNormalizer.
       new  37ba420   Documentation for LabelBinarizer and LabelNormalizer.
       new  699d832   Pep8.
       new  54310a9   Cosmit: LabelBinarizer and LabelNormalizer are not classifiers.
       new  530a0d7   More useful error message.
       new  270bd30   Doc cosmit.
       new  769c481   Add test for non-numerical labels.
       new  da2bf54   LabelNormalizer -> LabelEncoder.
       new  86f543c   Add documentation for non-numerical label case.
       new  87e9e5a   What's new.
       new  3498e1c   Cosmit: be more explicit why LabelEncoder is useful.
       new  b49c7e6   Address @larsmans' comments.
       new  06d3fd1   P3K: 'type' has been renamed 'class' in python3
       new  6a3413f   P3K: Fixed dtype doctests for Python3
       new  68062a2   P3K: Fixed print related Python3 errors
       new  47f459e   P3K : Fixed range iterator to be list
       new  ef7b166   PK3: __len__ returned float instead of int. Typecasted.
       new  cf46424   P3K : Convert int type checking to np.integer
       new  9fea659   P3K : Typecasted float to int
       new  704b861   P3K : Changed / to // to typecast float to int
       new  ad66a25   P3K: Modified RuntimeError message args
       new  66ceb05   P3K : Replaced / by //
       new  598dbff   P3K : Refactored test cases to use setUp
       new  66b9628   P3K: print back compatible with python2.6-7 with  __future__ import
       new  46ce069   P3K: Fixed None < Float Python 3 error
       new  23e9050   P3K: Fixed unicode pickling error by changing to BytesIO
       new  6592fdb   P3K: Fixing prints and dtypes
       new  2727820   P3K: Fixed RuntimeError.message
       new  05886db   P3K: more Py3k compat changes
       new  45733b3   BUG: multiple jobs in dict_learning
       new  4302abc   Merge branch 'sgd_losses' of github.com:mblondel/scikit-learn into sgd_losses
       new  85eccb9   Address @ogrisel and @pprett's comments.
       new  0186ffe   Fix remaining merge conflict.
       new  5543303   Fix doctest.
       new  611d1b5   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6030c8a   What's new.
       new  508abed   BUG: fix install bug for _check_build
       new  065b905   BUG: casting error with recent numpys
       new  e27b310   Merged svm parameter selection visualization
       new  bec5c65   Merge pull request #847 from kwgoodman/master
       new  09d4241   fix: gradient boosting regressor does not check if X is c-continous
       new  bc1dc1b   Merge remote branch 'upstream/master'
       new  5fab809   split plot_rbf_parameters.py's plot into two
       new  43c39f4   Added plot_rbf_parameters example to SVM doc
       new  069bf4a   Fixed bug in plot_rbf_parameters.py causing only one figure to show
       new  06622a9   Fixed location of ".. _svm_mathematical_formulation:" in svm.rst
       new  dc86f0d   Merge pull request #851 from duckworthd/master
       new  9a566bb   Convert input dtype to float in pairwise_distances
       new  eef72db   DOC instruct new devs to *always* work in branches
       new  d0b20f0   FIX inconsistent coef_.shape in LinearRegression
       new  e906d48   COSMIT pyflakes + pep8 linear_model/base.py
       new  1ebcc4a   COSMIT pep8
       new  a2c3112   MISC don't use fit parameters in example
       new  d92be8d   Convert input dtype to float in pairwise_distances
       new  02d1d85   ENH rename unmixing_matrix_ to components_ in FastICA
       new  f6c62bc   Added positive constraints for the elastic net
       new  3fd759f   Made the code pep friendly
       new  5b73af1   Added fit_intercept for sparse ElasticNet as well as corresponding test
       new  9a1d501   Corrected bad comment and the use of a typedef
       new  963cd8a   Made code pep compliant
       new  c6acacd   DOUBLE does'nt stand for a dtype
       new  a8ed0da   Added utility functions for csc sparse matrices
       new  4e7bcca   Modified: uses utility function for sparse csc matrices
       new  103ff6a   Modified data generation so it can generate data adapted to positiveness constraints
       new  beef9a8   Removed most python function calls
       new  fb9d18d   Removed duplicate definition of csc_mean_variance_axis0
       new  56d9bc1   Made the code pep8 compliant
       new  520e48e   Corrected doctring: CSR -> CSC
       new  cc01c12   Regenerated with Cython
       new  95ba957   Corrected missing import of csc_mean_variance_axis0
       new  c2c35da   Made code pep compliant
       new  892b603   Modified: in 'center_data' makes a copy only when needed
       new  e01dca7   Made code pep8 compliant
       new  4842ebb   Unified access to 'mean_variance_axis0' for CSC and CSR matrices
       new  48ba97f   Removed undeed functions
       new  9a9ad41   TST : tesitng intercept_ between dense and sparse
       new  81723cc   adding alexis to authors
       new  47351db   typo
       new  0e2fbc3   Fix typo.
       new  991da84   Note regarding multilabel example.
       new  5647d0e   WEB: update the designer's URL
       new  7b06026   ENH generalize LabelBinarizer to arbitrary Sequence types
       new  1e0e330   BUG remove debugging statements from multiclass
       new  6d04162   BUG in LabelBinarizer (forgot to run the full testsuite)
       new  66bdd96   Note on one-vs-all classification in SGD module.
       new  f672984   Unused import.
       new  e47039d   Fix warning.
       new  799a2ff   FIX: latex underscore
       new  d8c80ab   Explitit cmap for background.
       new  c1a4b69   DOC fixed sentence that was missing a verb
       new  3afedca   DOC document 'labels' argument of confusion_matrix
       new  b1cef07   DOC fix see also in gmm
       new  7e40dc7   update tip on svm C param
       new  3a7fbd2   DOC: note on heat kernel for spectral clustering
       new  df22296   Typo
       new  d69c5d3   Typo
       new  bc9f799   started work on Huber loss function for robust regression
       new  3da2c77   rm deprecated euclidian_distances synonym
       new  cfc2fd3   add dense attribute and dummy for sparse fit
       new  01b29e5   Some doc for the example "Lasso path using LARS"
       new  443929d   Some documentation for example plot_ridge_path
       new  855257f   ENH fix and test LabelBinarizer's handling of string labels
       new  e04fe92   BUG: reassigning cluster centers with X sparse
       new  ea2bacd   ENH import liblinear 1.91
       new  209c046   COSMIT make a liblinear C private helper function static
       new  684e9ec   ensure that std is not zero
       new  f26d3c6   add test case for scale div through zero
       new  efb80b2   BUG set new p parameter in liblinear helper
       new  f20296a   Merge branch 'master' into gbrt-huber
       new  104bb82   add huber loss to test
       new  a170d84   implemented huber loss for robust regression
       new  583ce70   fix errors in huber loss
       new  2f89cd1   Added `note` to tutorial index for `doctest_mode` in `ipython`
       new  7f436ab   add alpha parameter for huber robust regression loss
       new  8df4b2c   FIX made "unmixing_matrix_" a property as @larsmans suggested.
       new  8b24cbd   COSMIT pep8
       new  388c1af   ENH rename 'k' in KMeans and MiniBatchKMeans
       new  805c0f3   ENH renamed 'k' to n_clusters in SpectralClustering
       new  d9a6243   ENH rename k in clustering examples and doctests to n_clusters
       new  8199bfd   ENH fixed ``n_cluster`` to ``n_clusters`` in examples. Thanks @agramfort
       new  89f8618   ENH check whether "k" was used in fit, not init, as GaelVaroquaux suggested.
       new  06dcc21   BUG: k_means k -> n_clusters
       new  522942a   COSMIT: k -> n_clusters
       new  7054c7a   COSMIT: avoid deprecation warnings
       new  038a489   fix: ensure X is c-continuous
       new  7d89ab0   fix: make sure X is c-continuous
       new  196fd69   Merge branch 'master' into gbrt-huber
       new  1d63e20   MISC: os.name -> platform.system()
       new  486add4   Merge pull request #869 from pprett/pls-scale-by-zero
       new  4865777   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  dda0341   Merge pull request #872 from jaquesgrobler/master
       new  6e2472a   Remove a warning from kmeans tests
       new  9c53c8c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2b2a893   Remove deprecation warning in sparse_encode
       new  ab39014   Merge pull request #873 from vene/remove_sc_warning
       new  5d58089   MISC privatize/deprecate internal function of gaussian process
       new  5a37086   typo
       new  3cacf51   Updated mailmap
       new  43278f3   Updated mailmap (bis)
       new  3a25fb0   Fix a bug in the ward clustering.
       new  2db03c0   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into merge_cd
       new  43a363e   Add a non-regression test for the bug of connectivity fixing.
       new  7e7d9e8   Put conversion after component computation
       new  55c7e5e   Merge pull request #874 from temporaer/master
       new  abbec42   Added warm restart option and completed docstring
       new  8e8d476   Merge pull request #858 from amueller/fastica_components_rename
       new  faa53bd   Merge remote-tracking branch 'upstream/master'
       new  9abb2c7   Python 2.6 bugfix for plot_rbf_parameters.py
       new  b070cff   Merge pull request #877 from duckworthd/master
       new  4e5f811   Completed docstrings, factorized some tests and added checks on dimensions
       new  33c7633   Added test case for warm_start
       new  419c6b2   Added size check on coef_init
       new  4045b41   Fix test function name.
       new  d329227   changes to `fit` and `__init__`
       new  4a12b24   decision logic removed from __init__
       new  a12531b   API update for HMM types with docstrings
       new  14ecaa1   tests updated to match API
       new  cb16f80   fixed example`s fit(..) to new API
       new  59942ef   made `diag` explicit in example
       new  9c5f9af   FIX: unique in old numpy
       new  25bdb53   Added Pool Adjancent Violator
       new  4dbe1a0   SMACOF algorithm for MDS
       new  081cb87   Added tests and documentation to the smacof algorithm
       new  6a218d8   PAV now uses Kruskal's first approach to ties
       new  f568e36   Added a new dataset: traveling distances between 17 cities in france
       new  6366a8e   MDS now computes the SMACOF algorithm several times, and returns the results with the lowest stress
       new  ec72b72   Added documentation on MDS
       new  fe5ac32   MDS can now run several jobs in parallel thanks to joblib - when initial array passed, MDS will also only run once. If n_init is not set to 1, it will raise a warning
       new  09a002d   FIX mds tests where failing because of an interface change
       new  d09ed68   Added docstrings to MDS
       new  d3f0792   Cleaned up MDS's documentation
       new  ebaee5a   Added more documentation on the cities dataset
       new  81b75a8   Fix errors due to previous refactoring on MDS
       new  62a3484   Changed dataset from france's mileage to knuth's USA mileage dataset
       new  5d80149   Replaced MDS US mileage distance example by a generated, more representative one
       new  096e792   Added paragraphs on metric and nonmetric MDS, explaining the difference
       new  5b53e8d   MDS: out_dim → n_components
       new  c7b4e25   MDS: added documentation for n_jobs parameter
       new  518fe78   MDS - fixed some latex error in the documentation
       new  3a55480   Added a fit_transform method to the MDS class
       new  0ee486f   Pool Adjacent Violators now does a max_iter number of iteration
       new  e2e545f   DOC: added references to papers and licence - fixed the MDS example
       new  06fc359   a += a.T is different from a = a + a.T
       new  73b98d4   Small explanation on the plot_mds example
       new  d3cb559   np.diag raised a red flag - used broadcasting instead
       new  4707152   Set the seed of the random_state generators to have nicely aligned results
       new  0619753   Knuth load_cities dataset isn't used anymore
       new  6b6fb39   COSMIT in plot_mds.py example
       new  797605a   DOC: misc improvements in MDS docs
       new  c4d2fe0   MDS: renamed positions_ to embedding_
       new  7a0511c   Added MDS to manifold comparison methods
       new  d26c7cf   MDS: documentation fixes
       new  ea6bac3   FIX: load_cities doesn't exist anymore
       new  25bffd9   DOC: minor MDS doc/example changes
       new  468dcaa   MISC: update whats_new with MDS
       new  5d5c07d   Made code pep8 friendly. Used random state with fixed seed.
       new  4f8bb77   Made code pep8 friendly.
       new  fab9857   added feature subsampling to GBRT (via max_features)
       new  48a08d7   fix: forgot comma
       new  c696ce9   added test for max_features
       new  3e45173   COSMIT pep8
       new  2e08435   FIX typo in example. My bad.
       new  a682c9a   FIX renamed what was `components_` to `sources_`
       new  18c7cf1   COSMIT rst error
       new  b6b564d   fix: alpha needs to be scaled by 100
       new  c039766   COSMIT fixing doc building errors.
       new  6a4f214   COSMIT typo
       new  4bcf9f4   Merge pull request #776 from amueller/normalized_mutual_information
       new  5a24da5   Merge pull request #868 from larsmans/liblinear-1.91
       new  86baa8a   ENH "fit_pairwise" for spectral clustering.
       new  28fd13f   ENH Starting on affinity propagation
       new  faea50d   DOC typo
       new  7054701   DOC Improving docstring for SpectralClustering
       new  1d88edc   ENH fixed affinity propagation test. Need more tests.
       new  c39ad04   ENH fit_pairwise, transform_pairwise for KernelPCA
       new  37dbc5b   ENH base svm has fit_pairwise and predict_pairwise.
       new  4f11178   ENH fit_transform_pairwise for KernelPCA
       new  d91cf3c   ENH isomap uses new interface.
       new  8903668   COSMIT get rid of debugging output
       new  9f20a9c   ENH GridSearchCV uses the new API
       new  a26edc4   COSMIT forgot one print...
       new  5072bb0   DOC Deprecation warning with removal version 0.13.
       new  ef82d75   ENH going for a universal property ``_pairwise`` instead of many functions.
       new  ac1fb00   ENH Cleanup
       new  6c21e25   FIX Fixing rebasing problems...
       new  f93d94a   COSMIT avoid errors in tests.
       new  3e5f482   FIX : use RandomState rather than global seed
       new  e3369c1   ENH slight improvement to mds speed, modified examples to not run mds that long.
       new  dc01ed3   Faster confusion_matrix implementation
       new  b6e38ce   ENH added old confusion_matrix implementation as alternative for few labels.
       new  4eedded   ENH support opening compressed files in SVMlight reader
       new  9a2534f   Merge pull request #881 from amueller/fix_ica_components_rename
       new  6d9a864   FIX: fix buildbot ICA pb
       new  1deddaa   Fix a few spelling/grammar errors in the docs
       new  ef8ec5c   Merge pull request #887 from danohuiginn/master
       new  a06a14a   Fixed typos, spacing errors & updated `Whats New`
       new  d29211d   fixed broken GaussianHMM documentation generation
       new  93eddef   correct some wrong fixes
       new  8cf5896   BUG fixing bug in entropy that I introduced, adding regression test.
       new  de7eb72   FIX faces_decomposition example. That this broke only now is a sign of deep magic, better left unexplored.
       new  eeac6fb   Merge pull request #888 from jaquesgrobler/master
       new  9f33581   DOC removed irrelephant/confusion reference, added pointer to source (as there is no other possible reference).
       new  4366766   DOC user guide pdf building. Kicked out a formular that rendered neither in html nor latex. Please don't hit me.
       new  dd6d6b1   Merge pull request #876 from alexis-mignon/master
       new  f26e237   ENH: make_regression supports multiple targets
       new  3c02e6c   Update make_regression return shapes in docstring
       new  f59f72a   Merge pull request #889 from vene/generate-multitarget
       new  7078bce   add dense attribute and dummy for sparse fit
       new  9462086   Merge branch 'merge_cd' of https://github.com/ibayer/scikit-learn into merge_cd
       new  8f8f1c0   Merge pull request #875 from AlexandreAbraham/ward_coo_bug
       new  aa8354a   wip: added quantile regression loss; this allows for prediction intervals; adopted the GP regression example to show-case prediction intervals
       new  14b5c05   added title to example
       new  3d32de1   FIX : fix a division by zero in LARS #63
       new  a087e29   performance improvement for random split (ctyped two variables).
       new  505abb4   import random split
       new  84ff4dd   support of sparse input data added
       new  3689c44   tests of sparse coordinate_descent applied to the modified dense version
       new  f09c98a   -remove sparse option - decision function does not change state anymore
       new  8f397b6   remove sparse_coef_
       new  84ddd60   Test is redundant since _set_coef function as been removed.
       new  a7f7859   add property for sparse_coef_
       new  f1663e4   add test for sparse_coef_ property
       new  cce7889   docstrings updated
       new  91a5b42   merge cd_fast and cd_fast_sparse
       new  313bf63   remove redundant tests
       new  934f954   remove redundant files, functionality has been moved to cd_fast.pyx
       new  b281561   code removed and deprecated message added
       new  0586bfe   fix docstring example
       new  4f9e956   Merge pull request #892 from ibayer/fix_mldata_docstring
       new  069213c   COSMIT pep8
       new  2cc0af8   MISC raise more helpful error message in GaussianProcess if optimization fails.
       new  294d4b6   MISC added bigger "tiny" in lars_path. least_squares is float32.
       new  82d795f   MISC reduce code duplication, fix "self.gamma" modification
       new  028a33e   MISC A bit more cleaning up in BaseLibSVM
       new  f2603f7   add test to check normalize option in sparse enet
       new  3e98b44   Revert "remove redundant files, functionality has been moved to cd_fast.pyx"
       new  b8ec5c4   Revert "remove redundant tests"
       new  096e3e0   add sparse_std that has been wrongly removed in commit 48ba97f1 from the sparse version
       new  2e4d442   update sparse_std call
       new  f3f3892   some tests didn't use the numpy sparse matrix as input data and therefore tested the dense and not the sparse lasso and enet
       new  4587d60   make sure X is of dtype float64 in _sparse_fit
       new  bfc1671   change input to inplace_csc_column_scale
       new  f513dbd   modify test_normalize_option
       new  91ea012   test data changed for test_normalize_option
       new  533729b   remove redundant folders in linear_model/sparse
       new  c5a7eff   trailing space
       new  420144a   add missing attribute estimators_ to the docstring of forest models
       new  78595aa   FIX #898: narrative documentation for feature importances in forest models
       new  eb63974   FIX: C range in plot_cv_digits
       new  3b32420   remove unused imports
       new  91f1013   fix pip8
       new  fae2f13   move sparse_center_data to linear_model.base
       new  88929a6   avoid copy if X has proper type, modify docstring
       new  55862c1   Merge pull request #891 from ibayer/merge_cd
       new  b15401f   KNeighborsClassifier now has a predict_proba() method
       new  7bb3c6d   reversed changes to KNeighborsClassifier.predict()
       new  3e0810d   FIX : cleanup classes.rst + pep8 after merge of coordinate descent
       new  8b5142a   an simple test case for KNeighborsClassifier.predict_proba()
       new  c2e5a45   Merge pull request #900 from kernc/neighbors_predict_proba
       new  35dbb6e   DOC added "fetch_mldata" to references.
       new  ddb7b4d   ENH always support file descriptors in SVMlight loader
       new  1af8520   FIX : fix predict_proba in KNeighborsClassifier for old numpy
       new  3f1b189   CLEANUP remove linear_model.sparse.setup.py
       new  fb714df   Fix #904.
       new  5709d53   Removed needless method redefinition.
       new  e2ce08e   DOC typo in docstring
       new  669bbf2   BUG do not close fd passed by user in SVMlight loader
       new  55df2a7   COSMIT pep8
       new  127415b   DOC rename lambda to alpha in plot_lasso_model_selection. Closes #903.
       new  6bbdcfa   TESTING check that SVC checks the shape of precomputed kernels.
       new  7feff8c   Fix: RidgeClassifier must not inherit from RegressorMixin.
       new  878ddaf   test for quantile loss function
       new  688b612   Use BaseEstimator for constant predictors
       new  5b67d11   cosmit
       new  7e2ffbf   huber and quantile loss for gbrt
       new  09674ee   Clean up unused code.
       new  5f1f60d   ENH Check that X is non_zero for MultinomialNB.
       new  1832641   better docs for quantile reg
       new  4e7d996   ENH fixed doctests, addressed comments.
       new  727429a   FIX NearestCentroid.fit could not handle sparse formats other than CSR
       new  5314954   DOC typo
       new  802ed82   Merge branch 'master' into gbrt-huber
       new  cf20230   DOC improve kmeans init doc.
       new  8ea9262   Merge remote branch 'upstream/master' into gbrt-huber
       new  9f9fe4f   Modified chi2 kernel approximation such that it deals with zero elements
       new  b12076a   kernel approximation: simplified mangement of non zero elements
       new  adcf1cf   Merge pull request #894 from amueller/svm_sparse_dense
       new  2d8ffc7   DOC fix dead link
       new  0984cf6   For the sake of clarity, creates new temporary arrays instead of copying the same one several times.\n Modified error message for negative valued arrays.
       new  34cb85a   pep8 compatibility
       new  aa33e0e   FIX more doctests that I broke.
       new  2e89eeb   DOC comment in whats_new on changed behavior of ``gamma`` in SVM
       new  db0475b   Merge pull request #914 from alexis-mignon/master
       new  37cadf0   Merge branch 'master' into fit_pairwise
       new  537911e   MISC callable kernel gridsearch fix...
       new  67dee7b   DOC + COSMIT additive chi² sampler
       new  11e9842   ENH scipy.sparse support in additive chi² sampler
       new  b2e3858   DOC output from additive chi² sampler
       new  4f1e871   ctyped variables in ``find_random_split`` and use for loop over index range instead of array elements
       new  3ca2c71   Merge branch 'master' into gbrt-huber
       new  a25522e   ENH: MultiOutputTree (wip)
       new  4918692   fix: np.arange dtype issue; fix dtype to be np.int32
       new  863d53c   ENH factorize common tests.
       new  1e9c337   ENH don't list abstract base classes
       new  26c31bf   ENH make base classes abstract meta classes
       new  86cedaf   ENH make all Estimators default constructible (except SparseCoder)
       new  04d8160   ENH Add MetaEstimatorMixin, make RFE default constructible
       new  520da07   ENH make GMMs and LLE cloneable.
       new  7cdeecf   COSMIT get rid of warnings (can't get rid of deprecation warnings only :-/)
       new  21c0d1a   ENH make BaseLabelPropagation abstract base class, make OutlierDetectionMixin not inherit from ClassifierMixin
       new  282d267   BUG fix testing for abstract classes
       new  f71185d   ENH default score func for univariate feature selection: f_classif
       new  ee7d78c   Make sparse svm base class ABC
       new  30d6246   FIX better class selection, more strict testing.
       new  7a585e4   ENH more tests
       new  310a8ce   MISC raise NotImplementedError instead of value error in decision_function of sparse SVM
       new  c02938f   ENH do zero mean, unit variance on iris, don't test naive Bayes (for the moment)
       new  cd0b531   ENH change defaults on SGD (works on digits and iris and I just guessed them). Turned up alpha and n_iter. This corresponds to more regularization and more carefull SGD. On which kind of problems do the old defaults work?
       new  f0026be   ENH avoid division by zero in LDA, also avoid reusing variable names.
       new  4264acd   MISC don't test SVM for the moment, rest works :)
       new  bab0539   ENH make LinearModel and LinearModelCV abstract base classes
       new  6aa167c   ENH test regressors
       new  5338bda   MISC shuffle iris for SGD based methods
       new  a44318a   Revert "ENH change defaults on SGD (works on digits and iris and I just guessed them)."
       new  ddf5693   ENH Fix seed that makes SGDClassifier work.
       new  971d131   ENH create BaseRidge base class
       new  a93e0af   ENH test more shapes, test non-consecutive classes, test accuracy on test set
       new  a3e86dc   use np.int32_t instead of Py_ssize_t
       new  773ea6c   Merge branch 'master' into gbrt-huber
       new  7ce5b22   FIX: sparse ElasticNet tests that were not testing much at all
       new  94708a3   FIX minor rebasing and other problems
       new  bd94b49   MISC cleanup common testing
       new  549d82f   Merge pull request #893 from amueller/common_test
       new  4728c79   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-mo
       new  c210c0b   FIX for filtering of meta estimators in python2.6
       new  163d777   ENH better input validation for prediction in SVC, LinearSVC.
       new  e3f62d2   DOC Also added some notes on my recent merge with tests and stuff to the whatsnew.
       new  de8a370   added test that fails because Scaler.fit changes a sparse input vector when Scaler is initialized with copy=False
       new  7bd5fb5   removed bug in Scaler.fit
       new  7cb9468   improved test_scaler_without_copy
       new  f09a2f6   Merge pull request #921 from fhoeni/scaler_bugfix
       new  eac35cc   ENH: Multi-output decision trees
       new  064a48c   ENH: Regenerate .c file
       new  74bf03c   FIX: graphviz test
       new  55dbb49   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-mo
       new  be8ea69   FIX: test_classification_toy
       new  afacf44   TEST: test_multioutput (1)
       new  7e41cf2   MISC fixed random seeds in LLE tests.
       new  8792cbd   Merge remote branch 'upstream/master' into gbrt-huber
       new  6cf4d26   TEST: test_multioutput
       new  cf09064   use dtype float32
       new  b22b1f6   ENH: make forests support multi-output
       new  7b6ef37   TEST: test_multioutput
       new  5ee718c   ENH: Patch GradientBoosting
       new  41cd38f   ENH: Patch GradientBoosting (2)
       new  b4131f9   FIX: log_proba + DOC
       new  d560372   DOC: What's new
       new  9f7a0dd   PEP8
       new  0cae649   ENH: graphviz
       new  36fd9c3   proper pylab import
       new  e00d789   DOC: narrative documentation
       new  358884a   DOC: typo
       new  c549cb6   DOC: Scikit-Learn -> scikit-learn
       new  0d4719e   ENH: Cython improved code
       new  18a2e23   ENH: Cython improved code (2)
       new  5333afa   DOC: narrative documentation
       new  f178fe6   FIX: use and modify own y
       new  b14c23a   COSMIT
       new  071cdb0   FIX: heisentest for robust covariance: seed MinCovDet
       new  6e9c271   FIX: fix grid search when X is list #925
       new  5ec4a0a   Merge pull request #926 from agramfort/fix_X_list_grid_search
       new  309173b   Merge branch 'master' into gbrt-huber
       new  37b440e   ENH do not fail the test reslying on numpy div 0 warnings if those are not spit out by numpy in general
       new  f0961c2   Merge pull request #928 from yarikoptic/master
       new  f1bdd99   FIX: segfault
       new  f11ff94   DOC: Example
       new  264737e   DOC: typo
       new  386631e   DOC: example
       new  91963b8   DOC: typo
       new  a08a910   DOC: narrative documentation
       new  5d9f2ff   Patch up_inconsistent_numpy_warnings
       new  4918c9c   Patch up_MinCovDet_test_seeding
       new  114822b   Changelog entry for 0.11.0-2
       new  5a27534   reversed the order of the thresholds array
       new  a2170a5   BUG: address ill-conditionned designs in Lars
       new  46bbddd   Cosmit: PEP8 :P
       new  9f7a0ec   Merge remote branch 'upstream/master' into gbrt-huber
       new  4689d69   added test case for symbol labels
       new  545f324   y must be one dimensional
       new  549c0d0   more tests
       new  4dd43fc   DOC: docstrings for criteria
       new  89f7276   metrics.py
       new  a4974cb   DOC: docstrings
       new  41c5df9   test added for this
       new  89cb705   removed quantile regression example
       new  6fee3da   added max_features to gbrt regularization example
       new  abdfcfc   fix: section label for gbrt was wrong
       new  51661e7   FIX #937: preserve double precision values in svmlight serializer
       new  3edc52b   add a what's new entry
       new  886b6b2   fixed typos,updated `whats new`
       new  637ab82   added multi-ouput tree example
       new  f87f518   typo fixed in what`s new
       new  75c4415   Merge pull request #932 from jaquesgrobler/master
       new  81a1f90   updated documentation to reflect multi-output DT regression
       new  e5a61dc   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-mo
       new  94a5f3f   added link
       new  dc8e65a   Merge pull request #3 from bdholt1/glouppe-tree-mo
       new  9cf0a7c   add quantile example again
       new  e9590b6   added new features to whatsnew
       new  1d4c087   Merge branch 'gbrt-huber'
       new  532c54c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-mo
       new  f14601a   DOC: format
       new  aad531f   Merge pull request #923 from glouppe/tree-mo
       new  886945b   Fix broken bot (sorry for that!)
       new  e462bb4   Fix broken bot (again ;))
       new  8f55a18   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-speedup
       new  6b94adc   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  1864bca   COSMIT pep8
       new  0240e3d   work on smmlight serualizaer to preserve double precision values
       new  6b34e14   track master
       new  285d9cc   Merge pull request #938 from ogrisel/svmlight-double-precision
       new  d2625bb   DOC: What's new > Missing links
       new  60464de   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-speedup
       new  e122dc0   Tree refactoring (1)
       new  3786458   Tree refactoring (2)
       new  3054660   Tree refactoring (3)
       new  f3b203c   fix typo
       new  b07dfaf   Typo fix
       new  57162d9   Merge pull request #945 from cpa/master
       new  cd3df54   ENH: faster design in FastICA
       new  6c3f16d   tree: check length of sample_mask and X_argsorted.
       new  0ee5568   Cosmit: PEP8
       new  b7e3b6e   COSMIT: intermediate variable
       new  e976713   Tree refactoring (4)
       new  0015350   Tree refactoring (5)
       new  db9cb78   Tree refactoring (6)
       new  a868024   Tree refactoring (7)
       new  c9ac2ff   Tree refactoring (8)
       new  30f62f2   Tree refactoring (9)
       new  1e5aac8   Tree refactoring (10)
       new  a29897d   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-speedup
       new  516d292   DOC: fix typos in tree docstrings.
       new  89f8e12   Merge pull request #948 from mrjbq7/trees
       new  1bb8526   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-speedup
       new  a4d4bb6   DOC: fix value error text in Tree.compute_feature_importances.
       new  0e3cbbd   Merge pull request #950 from mrjbq7/trees
       new  df04b4c   Merge branch 'master' of github.com:scikit-learn/scikit-learn into tree-speedup
       new  2347423   ENH: Tree properties
       new  b6e68a3   Tree refactoring (11)
       new  c9da1f4   ENH: make Tree picklable
       new  cd7b5af   added alternating columns for tables in documentation and a tighter layout in pre
       new  13cad8c   Tree refactoring (12)
       new  6bc9b82   Tree refactoring (13)
       new  f1410e5   FIX: avoid useless data conversion
       new  d656721   FIX: avoid useless data conversion (2)
       new  16165c0   Tree refactoring (14)
       new  f482ebc   Tree refactoring (15)
       new  ec38852   Tree refactoring (16)
       new  2b72a1a   FIX: @mrjbq7 comments
       new  c08f40b   Tree refactoring (17)
       new  30f0a03   Tree refactoring (18)
       new  b6d9492   FIX: sample_mask
       new  4d43fcc   Merge branch 'tree-speedup' of github.com:glouppe/scikit-learn into tree-speedup
       new  e91c341   Merge pull request #953 from jaquesgrobler/nature_css_addons
       new  001a681   COSMIT pep8
       new  b463f8f   ENH in OvR, use constant predictor if one class always present or never. Closes #559.
       new  839c4b0   MISC address Gael's and Lars's comments, make ECOC tests deterministic.
       new  ed2721b   Begin updating the developers performance documentation
       new  0613f83   ENH: backport gen_rst changes from NISL
       new  310ada3   FIX: init/del => cinit/dealloc
       new  349a1e4   Added _tree.pxd
       new  e88f560   Update and fix errors in memory profiling documentation
       new  1f282c6   FIX: gradient boosting (1)
       new  2436bf2   COSMIT
       new  2fe48dc   Tree refactoring (19)
       new  923e471   FIX: PyArray_ZEROS -> np.zeros?
       new  2eb9e2f   FIX: gradient boosting (2)
       new  c59ee33   Tree refactoring (20)
       new  38ddb3d   What's new
       new  3015731   PEP8
       new  5c3aa7f   [Docstring Typo] making there -> making their
       new  918b9e1   Merge pull request #956 from Carreau/patch-1
       new  669c980   COSMIT
       new  d4dc61b   FIX trying to fix long-standing linker issue
       new  463ea61   Turn off warnings
       new  734cf7d   FIX: test_feature_importances
       new  1e0be6d   ENH: minor speedup in Ward
       new  999a749   ENH: factor 2 speedup in Ward
       new  2562842   FIX: test_feature_importances?
       new  316424d   ENH: minor speed up in ward
       new  5d8a941   ENH: minor speed up in Ward
       new  f865bc7   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  6027252   ENH: add store_loo_values attribute to _RidgeGCV see Issue #957
       new  29d5dc2   FIX: expose loo_values_ in RidgeCV instead of the private _RidgeGCV
       new  cfb6510   COSMIT: rename M matrix to loo_values
       new  129fd7b   COSMIT: -loo_values +cv_values
       new  c70afcb   FIX: use rng with fixed seed
       new  78517a3   COSMIT: Use np.array.fill for scalar values.
       new  c6aa568   TEST: disable test_feature_importances for now
       new  a2bb8f7   Merge pull request #946 from glouppe/tree-speedup
       new  44c503f   FIX: dtype conversion of y
       new  e25fa50   EXAMPLE: plot importances with bars
       new  9ad5d32   docstring fixes
       new  57d49e7   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  f397adb   Merge pull request #969 from jaquesgrobler/master
       new  38c3c30   COSMIT refactor input validation code and tests
       new  1426087   COSMIT + DOC input handling and docstrings in RandomizedPCA
       new  a484535   ENH classes_ on OvR classifier
       new  a30da06   Issue #339: minimizing number of calls in tests.test_hmm.
       new  c50fa29   COSMIT pep8
       new  bbd1350   Merge pull request #971 from acompa/master
       new  4fff8ed   trying out some testing stuff
       new  c1c734a   ENH put atlas checking in one place and load from there.
       new  69bef87   DOC typo / wrong parameter in lle docs
       new  bb92f9a   Improve test-coverage ;)
       new  be8319f   DOC typos
       new  fb9b7ce   COSMIT remove some dead code
       new  4a579f8   BUG remove predict{_log,}_proba from SVR
       new  f6bd86e   COSMIT cleanup tests with pyflakes
       new  bf990e2   COSMIT some RST fixes for the docs
       new  911a3c4   Remove empty statement
       new  2530791   ENH better input validation for dump_svmlight_file
       new  e1ba110   ENH make generated SVMlight files self-describing in a comment
       new  3472967   change dtype of y to float64 (aka DOUBLE_t)
       new  2cce14c   cosmit: better docstrings
       new  ea8c899   forest uses DOUBLE for y
       new  7896f72   COSMIT don't call magic methods directly
       new  b8ad24e   Update doc/support.rst
       new  180312b   DOC doctest failed on my box because I had higher precision...
       new  78ecda6   COSMIT typos in covertype benchmark
       new  b9098ef   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  398ed37   Merge branch 'master' into tree-y-float64
       new  35b24ca   ENH allow user-specified comment in SVMlight dumper
       new  494dd74   Fixed broken links on Support page
       new  5000ef0   Fixed broken links on Support page
       new  b6c8776   Merge pull request #974 from jaquesgrobler/master
       new  5a43dfa   This test repeats issues 960, with the silhouette coefficient returning nan
       new  2522de0   nan values are converted to zeros
       new  bbc6b80   k-means now no longer needed in test.
       new  91c6c3b   Distance matrix doesn't matter, and was therefore removed
       new  c72600d   MISC: avoid unprotected np.random
       new  8f71eb7   DOC: better phrasing about memory profiling
       new  ae82bc7   rm the long-deprecated scikits.learn package
       new  d9226d0   Merge pull request #886 from amueller/multiclass_always_present
       new  c24702e   rename mahalanobis_values to raw_values in covariance decision method.
       new  53b4e93   ENH: make LedoitWolf estimation scale (memory usage) with n_features.
       new  f3fc229   The LedoitWolf object has to return a covariance estimate or breaks.
       new  2e5f874   Put Ledoit-Wolf shrinkage coefficient estimation in a separate function.
       new  7b34256   Avoid extra computations + clean `assume_centered` argument use.
       new  b3d3c2a   Remove forgotten line related to previous commit.
       new  c9d7c05   Catch non-invertibility errors within MinCovDet computation.
       new  c6d92aa   TST: testing without hard-coding the values
       new  3fdd1e3   TST: test on diabetes rather than iris
       new  cc7937e   Cosmit
       new  d41b02a   BUG: example now needs 'assume_centered'
       new  6aa7450   ENH: using slices rather than indice masks
       new  c76f4f5   ENH: avoid unecessary steps (covariance)
       new  6aaf15f   Cosmit: more explicit names
       new  9150354   FIX: remove leftover print
       new  4915d4d   ENH: sklearn.setup_module to preseed RNGs to reproduce failures
       new  f96e352   Note on control of the RNG seed during testing
       new  8db2065   FIX: added logsumexp and nan_to_num to avoid underflows and NaNs
       new  09a7d78   Begin updating the developers performance documentation
       new  4c53380   Update and fix errors in memory profiling documentation
       new  0a8dbbe   DOC: better phrasing about memory profiling
       new  cc7f96c   DOC: cosmit performance instructions
       new  fbcde2d   Merge pull request #955 from vene/mem_prof
       new  f293630   TST: test check_build
       new  408d3ef   COSMIT, removed scikits.learn things, removed orphan file.
       new  9300d96   ENH trying to catch that damn thing.
       new  b90a30d   ENH better error messaged in multiclass as @mbondel suggested.
       new  bebe5de   ENH: remove setuptools
       new  73b7663   ENH: restore 'develop' mode install
       new  0df89f0   FIX: remove executable bit on joblib files
       new  4701319   TST: improve coverage of feature_selection.SelectorMixin
       new  af2602e   BUG: fix setup.py for develop
       new  f8c19f9   COSMIT suppress warning from qr_economic + docstring on Counter
       new  8074524   Test default input.
       new  3fe5875   Credits and license.
       new  73a274e   Added test to sklearn.utils.bench's total_seconds method
       new  e80efe6   Improve covariance module test coverage.
       new  4ffe0a8   More tests for the covariance module.
       new  68693b8   BF: adapt a svm test to recent numpy versions.
       new  e8af7c6   Test for "amg" mode for spectral clustering added.
       new  bddf256   docfix: spectral_cluster doesn't return n_centers
       new  bbc8038   pep8
       new  26dd857   Spectral will raise an error if the mode is set to amg and pyamg is not available
       new  993b6b4   Test that an unknown mode raises the appropritate error
       new  3b15f44   TST: test the setup.py using the configure step
       new  2fd7813   MISC cleanup old coverage info in Makefile
       new  d8e30f1   REF: hack to be able to share distutils utilities.
       new  4e1912a   TST absolute imports in spectral clustering tests
       new  1a5e4c3   ENH more specific warning filter for qr_economic
       new  bb73ff8   Merge pull request #1 from cournape/linking_arrayfuncs
       new  d0d9588   ENH corrected errormessages for always present labels. ugh
       new  48055de   FIX doctests for changed dtype
       new  c88a667   ENH fixed warning for output code
       new  fd2283f   FIXed another doctest.
       new  cc834d8   ENH add verbose warning about too little trees for oob. Should we catch the divison by zero warning for classification?
       new  066108e   DOC made the pls example plots so much prettier
       new  7a86168   ENH: Faster ward for large n_clusters
       new  7bc92af   BUG: fix ward tests
       new  df1e711   DOC: ward docstring and testing
       new  99f5813   TEST: improve test coverage in hierarchical
       new  8022008   FIX: make ward_tree work on 1D data
       new  d30eea0   MISC: very minor speedup
       new  9dbc576   COSMIT: remove left over profiling
       new  bdb32d4   TST: More testing in hierarchical
       new  5ad4d0e   TST: test TypeError in Ward
       new  6d0c727   TST: more tests for hierarchical
       new  01cf172   DOC: notes on improving code coverage
       new  ecda4de   COSMIT: explainations of the partial import
       new  446611d   MISC: build_utils: module rather than a subpackages
       new  9457f0c   ENH: use sklearn.__version__ in setup.py
       new  b5519f0   Merge branch 'linking_arrayfuncs'
       new  6de8466   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  9cfdff5   missing pl.show() in plot_digits_agglomeration.py
       new  f01907d   TST upgrade trivial (single-class) k-NN problems to binary ones
       new  c389d7d   Cosmit: comment
       new  01572bf   TST: fix doctest
       new  6a37b8f   Update whats_new
       new  d4a020b   Clean: remove debug print
       new  3c73706   PEP8
       new  6c4ca28   Typos
       new  1601b27   Merge pull request #983 from GaelVaroquaux/faster_ward
       new  f4882b5   DOC + TST vocabulary arg in CountVect docstring
       new  7c8f435   COSMIT move BaseSGD to its only place of usage
       new  b3bedb5   DOC: add tutorial links
       new  d6d9ec0   Update doc/whats_new.rst
       new  ee55901   Update doc/whats_new.rst
       new  0695e83   MISC : update my web site URL in what's new
       new  913e222   COSMIT minor refactoring of SGD
       new  44d57bb   Typo.
       new  7db59b1   feature_extraction.text.CountVectorizer analyzer 'char_nospace' creates character n-grams but only inside word boundaries, e.g. 'this is it' to 3-grams: ['thi', 'his', 'is', 'it']
       new  8b2893e   Oneliner docstring
       new  e83fe8b   words for n-grams padded with one space on each side
       new  9f0ec0b   missing unicode modifier
       new  ed89ddf   replaced str.format() with string concatenation as it's 3 times faster
       new  08338cf   char_nspace -> char_nospace, thanks Lars
       new  da96817   DOC tutorial: explain what an estimator is
       new  6f54691   Check that feature indices are sorted.
       new  ef8aa40   Add missing test file.
       new  5c3853c   Merge branch 'master' into fit_pairwise
       new  242e553   Fixed merge problem
       new  440cafb   ENH Removed stupid ``_pairwise`` property in BaseEstimator.
       new  d0e21ac   MISC minor cleanup in spectral clustering
       new  9a52077   FIX/TST test anc fix grid search with kernel pca and precomputed kernel in pipeline.
       new  9ddb180   COSMIT comments not docstrings in tests
       new  194b0f1   Merge branch 'master' into fit_pairwise
       new  5a7a1a1   TST precision issue on my windows box :-/
       new  dcbf5a6   ENH : MultiTaskLasso works (still draft)
       new  371f49a   FIX : fix docstring in MultiTaskLasso
       new  287898d   ENH : add multi task lasso example
       new  21706e3   ENH + DOC : add MultiTaskElasticNet + doc + 1 example
       new  243233c   update what's new
       new  8999e4f   FIX : support 1d y in MultiTaskFoobar
       new  02446b8   rename ylabel in MultiTaskLasso example
       new  955c8ef   moving MultiTaskLasso doc after E-net
       new  86e3616   FIX : remove unnecessary dgemm in cd_fast.pyx
       new  0819b10   BUILD: add gemv cblas routine
       new  4bcf28d   BUILD: add dger cblas function
       new  0d50313   BUG: keep same shape for y in MultiTaskLasso
       new  dab7dcf   DOC: explicit MultiTaskLasso.coef_ dimensions
       new  a7aed50   DOC: formatting and rephrasing in MultiTaskLasso
       new  6df5d76   ENH slight cleanup in LDA, QDA, support for arbitrary class labels.
       new  8dae4c6   ENH use LabelEncoder
       new  bb6b45d   COSMIT typo in pairwise docs
       new  7ff81e4   DOC added LabelEncoder to the References.
       new  fcced58   Optim in LabelEncoder.
       new  ba78c11   Fixed small typo, even though the message is kind of the same and the one with the typo is waaay funnier.
       new  97eacbe   Merge pull request #1001 from serch/master
       new  a746332   FIX - the eps option of the MDS was overwritten
       new  eedb105   Merge pull request #1005 from NelleV/MDS
       new  9bf746c   COSMIT: doc fixes to sklearn.feature_selection.univariate_selection.
       new  a733f3b   COSMIT: fix typo of homoscedasticity.
       new  8b29486   COSMIT: fix reference to scipy.stats.kruskal.
       new  ce856db   COSMIT: fix more typos.
       new  0a2d8b6   Merge pull request #1008 from mrjbq7/doc-fixes
       new  82a762f   COSMIT pep8
       new  a12bd7b   ENH: understandable error message for X sparse
       new  cd3745a   FIX : catching pb with sparse input in MultiTaskElasticNet
       new  45b88b7   ENH just a little more input validation testing
       new  b0c08ee   DOC added default value of shrink_theshold to NearestCentroid docstring.
       new  fdf6bfa   DOC added ``lowercase`` to CountVectorizer docstring.
       new  c41b03b   FIX feature selection dies on non-csr sparse matrices (that are unsubscribable). Regression test should go in common testing.
       new  a93945a   DOC added class_weight to LogisticRegression docstring
       new  8d0b71c   ENH auc_score and average_precision_score. Closes issue #158.
       new  320035e   ENH added to ``__init__.py`` and references.
       new  fd171dc   We already have the inverse at that step
       new  0b8b4e0   Replase pinv calls with dgetri
       new  ba4f3ed   DOC explained RFE default behavior in docstring.
       new  22531a3   More lapack inverting
       new  23e1c06   MISC Added unconfigured windows box to mailmap. Sorry about that.
       new  1f0daf9   DOC add parameters to TfidfTransformer docstring
       new  b346184   changed 'char_nospace' keyword to shorter and meaningful 'char_wb'
       new  27164f5   some narrative documentation...
       new  08f00a0   BF: Make MinCovDet work with n_samples >> n_features.
       new  44173ac   Merge branch 'cov-speedup' of https://github.com/vene/scikit-learn into cov-speedup
       new  3d6b4aa   Add comments on optimized precision computations.
       new  0b7c93c   ENH slight cleanup in LDA, QDA, support for arbitrary class labels.
       new  53c1126   ENH use LabelEncoder
       new  9ecca83   FIX Removed code-duplication introduced in rebase.
       new  d4ff4f0   FIX Fixed variable names. Thanks @mblondel
       new  3d8c5df   DOC Added wikipedia references to docstrings
       new  4c3532a   mentioned 'char' vs 'char_wb' in the narrative
       new  05dd59c   Merge pull request #1013 from amueller/auc_score
       new  0d0d3e7   Refactored fast_pinv without lapack calls
       new  e5af8e4   DOC Updated whatsnew
       new  46ebdd4   Compute pseudoinverse using eigendecomposition
       new  bdc7069   Merge branch 'cov-speedup' of git://github.com/vene/scikit-learn into vene-cov-speedup
       new  d2722a4   speed up symmetric_pinv
       new  be68d3a   additional speedup: all eigenvalues are real for symmetric matrix
       new  79ced0e   Vectorize singular value inversion
       new  fb167da   Remove unused import
       new  55bc245   changed shape of predict_proba
       new  bd9cece   adopted tests because of changed shape of predict_proba
       new  f97629c   adopted tests because of changed shape of predict_proba
       new  2bec0d7   cosmit in sgd docs
       new  5d38957   added change to ``whats_new``
       new  53ec265   Merge branch 'master' into cov-speedup
       new  d16c9d4   Merge remote-tracking branch 'VirgileFritsch/cov-speedup' into cov-speedup
       new  d917149   Merge remote-tracking branch 'jakevdp/vene-cov-speedup' into cov-speedup
       new  c520275   ENH sparse matrix support in univariate feature selection
       new  fc3cdc7   TST Simplified tests, test that sparse and dense versions give the same result, always return arrays, not matrices.
       new  00aa089   DOC Polished some docstrings
       new  e95d461   ENH Added copy keyword to safe_sqr, added to dev docs.
       new  1aec782   COSMIT Fixed commata
       new  52b1bef   ENH Addressed @mblondel's comments.
       new  c789e3c   ENH simplify as @mblondel suggests
       new  bef749b   ENH sparse matrix support for RFE and RFECV. Closes issue #1018.
       new  e8a0167   DOC updated whats_new
       new  e4aefd6   add quantile regression example to gbm doc
       new  c8dde78   Remove needless loop in inverse_transform.
       new  0017782   Update and rename pinvh (by @jakevdp)
       new  755a5dd   ENH going back to not using LabelEncoder.
       new  3a1aaf9   Merge branch 'qda_lda_1000' of github.com:amueller/scikit-learn into qda_lda_1000
       new  8e0528d   Merge pull request #995 from kernc/CountVectorizer_analyzer_char_nospace
       new  f779301   Simplify LabelEncoder.fit_transform.
       new  6998a0e   fix broken doctests for the new char_wb text analyzer
       new  17a09c5   Fix warnings in multiclass module tests.
       new  59bc583   Cloned @jakevdp's pinvh tests
       new  8918f37   Merge pull request #1000 from amueller/qda_lda_1000
       new  6ba3de1   DOC: better narrative for char_wb text analyzer + add a whats_new entry
       new  bb90d67   Remove duplicated line.
       new  2eeb3a7   Remove odd-looking period in tests
       new  641d38f   TST: change LLE test to stable seed
       new  ba07ae2   DOC: fix documentation of arpack
       new  dc0734e   typo in linear_model doc
       new  2fe9b06   ENH add verbosity parameter to cross_validation_score
       new  c583e39   MISC catch warnings in covariance tests
       new  70caf15   Use pinvh in plot_sparse_recovery example
       new  ad3587e   grammar
       new  c2a73f4   Use pinvh in bayes.py
       new  d8be1c3   Typo in last commit :-/ sry
       new  30c30f6   Use pinvh in GMM and DPGMM
       new  3990e13   ENH catch expected warning in ward clustering
       new  47c1562   ENH renamed ``min_n`` and ``max_n`` parameters in CountVectorizer to enable gridsearch over them together.
       new  bf787ac   ENH renamed parameter bounds_n to ngram_range, fixed doctests and tests.
       new  cd18098   ENH addresses @ogrisel's comments
       new  4156481   ENH fix merge with char_wb_ngram
       new  0d69c6c   ENH check that classifier decision_function and predict_proba validate shape of input.
       new  e00d706   Merge pull request #991 from jakevdp/doc-update
       new  52e39db   fixed long-name-references madness + removed some whitespace
       new  e116d2c   trainling whitespace removed
       new  8f91ec7   blank line removed
       new  e1bf857   Merge pull request #1043 from jaquesgrobler/master
       new  ba55424   Merge pull request #1039 from jakevdp/lle-test-fix
       new  8043efb   Add all_categories option.
       new  e55e2c5   Normalize training and test times.
       new  c677f0b   Typo.
       new  f3c603d   Enable graphvis export function to export trees as well as regressors
       new  b59e100   FIX : make as_float_array keep fortran order on dense array when copy
       new  bbe17a2   Merge pull request #1045 from agramfort/fix/as_float_array
       new  dbcf8bb   Merge pull request #1046 from TimSC/master
       new  7fb4fa3   Fixed docstring for C param in BaseLibLinear/SVM subclasses.
       new  ceb0d8b   Added version info to deprecation warning
       new  ea6d903   Merge pull request #1049 from fsav/c-docstring-patch
       new  ccd1384   add support for non-ndarray lists
       new  fbd0082   Merge branch 'master' into peter-dev
       new  a89604e   Merge pull request #1063 from welinder/peter-dev
       new  2311a4c   COSMIT pep8
       new  da80997   ENH rename paramter ``p`` of AffinitPropagation to ``preference``, slightly change the meaning of scalar parameter. Scaling the medium seems more intuitive that giving absolute values.
       new  56c6bfb   Simplify LabelEncoder.transform.
       new  7107544   Test LabelEncoder.fit_transform with arbitrary labels.
       new  5762f74   Ignore joblib folder.
       new  f408d76   DOC fixed renaming of ngram_range in feature_extraction narrative
       new  6829a2f   DOC rewrote logistic regression docs
       new  1aaa7ac   TST check that transformers fail gracefully on sparse input
       new  f6aaa4d   ENH affinity propagation now has an ``affinity`` parameter, instead of a ``precomputed`` parameter, to support other affinities in the future.
       new  7c4f112   ENH renamed ``gaussian`` affinity to ``rbf`` in spectral clustering for consistency.
       new  2b328ca   COSMIT renamed n_points to n_samples everywhere, fixed shape docstring that @mblondel pointed out.
       new  eb89d5a   FIX Worst feature in RFECV missing. closes issue #681.
       new  949ddd1   ENH renamed ``neq_sqr_euclidean`` to ``euclidean`` so we it is easier to parse
       new  e42c435   ENH Convert input into float in GMM
       new  0df98bb   ENH add test, revert affinity propagation to previous parametrization (was a bit over-eager there)
       new  dcbe01f   TST added tests for different spectral clustering affinities
       new  b797265   Merge branch 'fit_pairwise'
       new  546f7df   MISC add verbose keyword to AffinityPropagation
       new  faef6e7   slight adjustment to header size
       new  7883d7a   FIX fixed horrible bug in spectral clustering!!!!
       new  c59a7dd   ENH updated whatsnew for bugfix, removed warning box, tightened test.
       new  74a99dc   Merge pull request #1075 from jaquesgrobler/master
       new  42e9029   TST classifier behavior with only one class present
       new  0b707d3   ENH also test MultinomialNB
       new  e4c71ea   FIX removed ancient templates from manifest to make sklearn pip-installable.
       new  fed8259   correcting typos in the doc
       new  ff01105   Merge pull request #1077 from ludwigschwardt/minor-fixes
       new  1fb13d4   FIX: forest / check_random_state in fit
       new  4e3576f   FIX: tree / check_random_state in fit
       new  93beee2   Update README.rst
       new  97c290c   Merge pull request #1078 from buguen/docs
       new  ca36d73   BUG: casting rule with recent numpy
       new  4a15c20   BUG: do not use diag_indices
       new  1378012   BUG: choose seed to get affinity test working
       new  ae377f6   BUG: fix my fix for affinity :(
       new  e18a32d   Fix #1080.
       new  5d73ecd   Merge pull request #1009 from amueller/one_class_check
       new  053d1a5   Merge branch 'master' into sgd-predict-proba
       new  59cf938   Merge branch 'sgd-predict-proba'
       new  8b3e4d0   ENH some cleanup in grid_search. Closes #777.
       new  9d0080a   DOC: link to Randomized sparsity in Lasso section
       new  1ad6926   Merge pull request #1068 from amueller/grid_search_cleanup
       new  7a5e118   ENH add test for consistend predict_proba shape also in the two-class case.
       new  a11c38d   tst add check for isotropic data in spectral clustering
       new  fef96c7   FIX try to be a bit nicer to arpack - any one with a different setting care to try to make a more stable test?
       new  694e575   New stratified shuffle split version that only return indices arrays
       new  5d8b384   stratified shuffle split can return masks
       new  6445dee   Fixed StratifiedShuffleSplit issue for unbalanced classes
       new  4d25843   Fixed n_test issue in StratifiedShuffleSplit
       new  781326e   pep8 fix
       new  5b5b3b3   Added new tests for StratifiedShuffleSplit
       new  d259e1f   Fixed SSS test
       new  d5db0fa   Removed redefinition of variable i in SSS
       new  e453af9   Permute the train and test sets in SSS to avoid class-sorted folds
       new  0f16b8a   Added validation for some corner cases in SSS
       new  1d21109   Updated tests for SSS
       new  55907d5   Added tests for the StratifiedShuffleSplit to check the sizes of the training and testing sets, and that they don't overlap
       new  b2e6e19   Minor cleanup of StratifiedShuffleSplit
       new  de100d9   FIX doctest corrected (hopefully this is deterministic) + cosmit
       new  689cd96   FIX removed isotropic spectral clustering test because of arpack problems.
       new  bccfefa   FIX use backport of np.unique
       new  c9ee91b   FIX forgot some uniques
       new  fd2a0a2   added failing test for 2d y
       new  c7e65d2   rm redundant input check (we check in _partial_fit)
       new  0902ed0   ravel y; use atleast2d_or_csr for input validation
       new  e0434de   _tocsr not needed because of atleast2d_or_csr
       new  a1dfd31   inline comment
       new  d9ff3f5   cosmit: constants for penalty types and learning rate types; inline comments;
       new  920bf84   Decision threshold is now 0 in RidgeClassifier.
       new  7dd7bad   Optim + cosmit in StratifiedShuffleSplit.
       new  7ae3356   DOC fix minor sphinx errors and stuff
       new  072b3c6   enh: try to get decision function to work in two class case
       new  2019260   ENH make QDA and LDA decision functions adhere to standard shape [n_samples,] in two class case.
       new  54076d5   Fixed tests for RidgeClassifier
       new  7ca5577   DOC updated whatsnew, moved @pprett's api fix into the api section.
       new  f540875   ENH addressed @agramfort's comment, also removed the special case from testing as @mblondel fixed it :)
       new  05630b0   ENH added min_df keyword to CountVectorizer, default=2
       new  f9d6b2f   ENH more robust testing for int
       new  4c8c3be   ENH more robust testing if parameter is int or float, as suggested by @larsmans in #1066.
       new  b27dde2   FIX typo
       new  80b4a5c   fix warning: add underscore to: grid_search.best_estimator_ and grid_search.best_score_
       new  53e7a87   Merge pull request #1094 from ibayer/warnings
       new  a30359c   COSMIT Typo. Englais svp. Closes #1090.
       new  0fccfbe   COSMIT trying to fix doc issues
       new  b57e01b   DOC added min_df change to whatsnew, made more estimator names clickable.
       new  2ec38b3   ENH rudimentary testing of tranformer objects
       new  cfe11f9   MISC added comment to explain SelectKBest k in common tests
       new  3cafdaa   COSMIT copy+paste error
       new  f3b76d2   ENH test that regressors can handle integer data.
       new  069936c   ENH add ClassifierMixin with ``fit_predict`` and some tests.
       new  d33c2ea   COSMIT remove commented out score
       new  d0efa6f   Remove deprecated _set_params and the call in grid_search
       new  fc3e680   Remove chunk_size from k_means
       new  f692faf   Removed load_filenames and load_20newsgroups
       new  4575b32   Remove sparse_encode_parallel
       new  89813e9   Removed deprecated parameters in GridSearchCV
       new  02a831a   Remove LARS and LassoLARS
       new  2516b62   Remove fast_svd.
       new  2769dbd   Remove _get_params
       new  bc90358   Corrected deprecation schedule in cross_validation
       new  b515066   Remove deprecated properties in naive_bayes
       new  9e7c03f   Add or fix deprecation schedule in warnings.
       new  0a382ac   Fix example using deprecated API, output was misleading.
       new  b0cb959   Remove deprecated load_20newsgroups from classes.rst
       new  3f8b454   FIX: randomly failing CountVectorizer test
       new  1cdc9e1   DOC CountVectorizerDocstring readability
       new  e0c6c11   MDS is not a transformer, fix the test to skip PLS
       new  ad339a4   Merge branch 'master' into mixins
       new  9829179   Improve the common tests, make fast_ica pipelinable
       new  f058593   Support y-dependent transform as in PLS
       new  30e1597   fit_transform in PLS to support y
       new  2b7c336   Make PLS degrade gracefully on sparse data
       new  4452154   FIX in the makefile - we should delete pyc and so only from the source code, and not from everything in the root folder
       new  a10928c   Rename Y to y in PLS
       new  a5541d9   Check for sparse input in isomap and lle
       new  a36a190   Check for sparse data in MDS despite not being tested
       new  3a7125f   Skip CCA in test_regressors_int
       new  c4b6b39   DOC Added section on issue tracker tags to development docs
       new  4d13bd6   Merge pull request #1100 from NelleV/makefile
       new  5954727   Deprecated sparse classes from the SVM module - refs #1093
       new  4631a21   FIX sparse OneClassSVM was using the wrong parameter
       new  163dbe5   ENH raise ValueError in r2_score when given only a single sample.
       new  d8f576c   Merge branch 'master' into mixins
       new  58ed136   Revert "Rename Y to y in PLS"
       new  9fbf66c   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  e737aa4   FIX the AP was using a deprecated parameter
       new  9f5fd65   Decrease the number of convit in the AP
       new  d70ee4e   Renamed parameter convit to convergence_iteration and deprecated the old API
       new  45a5bd2   FIX typo in deprecation warning in the AP module
       new  ce6cfbf   DOC better documentation on the AP
       new  4789910   FIX The new parameter of the AP is called convergence_iter and not convergence_iteration anymore
       new  57b184b   ENH: simplify dict learning with gram and reg_param handling
       new  2184c46   ENH : add copy arg to array2d and new atleast2d_or_csr usual for sparse coordinate descent
       new  9dfdcc7   First effort in multitarget lassolars
       new  f74f5f0   ENH: move Gram precomputation outside of the loop
       new  abc9ccc   TEST: precomputed lasso and lars
       new  fee95e1   Unnecessary copying
       new  519e204   FIX: add test, fix memory initialization bug
       new  2c9ede5   ENH: multidimensional y in ElasticNet (WIP)
       new  e483883   ENH : add copy param to array2d_or_csx
       new  48687a7   ENH : add support for multitarget in sparse enet + simplify input checking
       new  dffd56e   ENH : use multitarget in dict learning
       new  ab7c8cc   FIX : fix tests
       new  73cb0a3   DOC : getting over docstrings
       new  3109c99   ENH : avoid a copy in MultiTaskElasticNet
       new  83d5f18   add note on what's new
       new  36cdf1c   ENH : add support for sparse data in ElasticNetCV/LassoCV (not optimal)
       new  9c6422c   ENH : use multitarget Lars and LassoLars in dict_learning
       new  9cbdad8   ENH : simplify handle of copy of Gram and X with array2d in OMP
       new  01d032c   style + typo
       new  4e3b66a   DOC : better reg_param docstring in dict learning
       new  51d02d3   ENH : use build_dataset in multi target test
       new  8bf1acb   ENH : update warn for multitarget
       new  eac812b   return_path option in lars_path
       new  c1ef133   Add possibility to ignore the path in Lars objects
       new  d40f7a4   Fix doctests
       new  105a678   update coef_path_ docstrings
       new  9e3354b   use assert_true
       new  466479f   API : consistent use alpha_/alphas_ for alpha/alphas estimated by CV in linear models (issue #1041)
       new  5af5b7f   DOC : add useful comment in code
       new  28cecfe   addressing for round of reviews
       new  7a04ca0   BUG: sparse matrices in ElasticNetCV
       new  b763ee2   DOC : better docstring for fit_path
       new  273468a   MISC rest
       new  0441dc5   Merge branch 'master' into sgd-yshape-fix
       new  81bca36   fix typo
       new  50868f2   make smoke tests explicit; check ValueError on 2d inputs
       new  31af752   ENH support custom kernels on sparse matrices
       new  79ebf6c   ENH added low-level bail out in sparse svm
       new  a6a8771   MISC use assert instead of value error.
       new  50a2d05   FIX add exception, check exception, if sparse.SVC is called with kernel='precomputed'
       new  254e15f   ENH fix error by removing unnecessary test.
       new  1c1ea99   DOC added some comments to the sparse precomputed kernel tests.
       new  1575e50   FIX: this fixes issues #746 ProbabilisticPCA minor things
       new  ea14dc3   FIX: this further fixes issues #746 with API compatibility warning and integer division fix
       new  a190eb4   DOC updated whatsnew with ProbabilisticPCA fix by @kuantkid
       new  bcbcc5c   DOC Added SGDCLassifier support only binary prediction probabilites.
       new  2a9e095   Merge pull request #1109 from buma/predict_proba_doc
       new  4d46856   DOC Fixed a return in predict_proba in SGDClassifier
       new  e691007   DOC add support for sparse arrays to SGDCLassifer
       new  d3139db   DOC forgot dot in SGDCLassifier documentation
       new  a682cfb   FIX affinity propagation typo
       new  63e9902   DOC fixed some sphinx errors, issues in docs....
       new  5236f8c   DOC Fixed a return in predict_proba in SGDClassifier
       new  ab7e942   DOC add support for sparse arrays to SGDCLassifer
       new  f524edc   DOC forgot dot in SGDCLassifier documentation
       new  bd4292a   work on BaseGradientBoostingCV
       new  425d1f9   refactored prediction and decision_function (rm duplicate code)
       new  554799f   ENH: use gini for feature importance
       new  d7fc786   GradientBoosting classes with built in cross-validation; implemented via Decorator pattern
       new  e93a6a5   wip: aggregate fold via groupby
       new  970326f   wip: fixing some set attr errors but still buggy if params not lists
       new  e6f0b1d   remove *CV classes - only pick decision_function and staged predict refactoring
       new  2c97793   rm CV class tests
       new  edee5ca   rm CV class legacy
       new  a8a04c6   remove CV class legacy
       new  d5658b6   add API changes and feature_importance fix to whatsnew
       new  5b9cd36   Merge pull request #1110 from buma/predict_proba_doc
       new  73f991c   DOC yet another AKA
       new  713ba60   COSMIT pep8
       new  e35b3e5   DOC fixed reference in whatsnew
       new  0f4208b   DOC added some more API changes to whatsnew
       new  7f76aa1   Add __all__ for half of the scikit
       new  006e7ef   Add __all__ for the second half of the scikit
       new  ef2f4ac   Expose ENGLISH_STOP_WORDS
       new  317a928   FIX removed sparse_encode_parallel
       new  3d02176   Added scale_c fiasco example
       new  6b04635   DOC: improve scale_c_example
       new  254b109   gael`s suggestions/tweaks
       new  e50fad6   docstring change
       new  8367109   docstring fixes
       new  d117cb7   changed includes back - change broke JENKINS build
       new  75ba083   not the problem afterall - switch back
       new  656d701   docstring changes
       new  868bcd2   typos and alex`s review changes
       new  e686e38   small tweaks
       new  4bc8822   COSMIT pep8
       new  e53274a   COSMIT typo, thanks @ogrisel
       new  a4a3d35   MISC changed version number for release, change maintainer to myself
       new  8e8e9a1   DOC add people and commits do whatsnew
       new  fc5fe58   COSMIT typo, thanks @ogrisel
       new  c38c547   DOC add people and commits do whatsnew
       new  f1e4427   MISC starting 0.13 cycle
       new  3bfb4e7   DOC added link to 0.11 docs to support page.
       new  ec61542   We already have the inverse at that step
       new  4a5a1d6   Compute pseudoinverse using eigendecomposition
       new  bfac9f2   Vectorize singular value inversion
       new  57c1abe   Add comments on optimized precision computations.
       new  776b060   @jakevdp's version of pinvh speed up symmetric_pinv
       new  f4a4016   Cloned @jakevdp's pinvh tests
       new  e6ca49e   Use pinvh wherever it helps in the codebase.
       new  d11e6d1   ENH more robust transformer testing.... don't ask why that came up
       new  8d103b5   We already have the inverse at that step
       new  0946f59   Compute pseudoinverse using eigendecomposition
       new  4829341   Vectorize singular value inversion
       new  602bac4   Add comments on optimized precision computations.
       new  12f4eb9   @jakevdp's version of pinvh speed up symmetric_pinv
       new  6d66d0f   Cloned @jakevdp's pinvh tests
       new  97717ba   Use pinvh wherever it helps in the codebase.
       new  0fede44   ENH more robust transformer testing.... don't ask why that came up
       new  9e3e9ad   DOC copyediting
       new  4f46545   FIX: bug in multi-output forest.predict_proba
       new  6eb4bb3   TST (near-)empty lines and explicit zeros in SVMlight loader
       new  0e5b876   DOC: add a reference on multi-output trees
       new  3f45f67   MISC: docstring work
       new  9b642e5   Merge tag '0.12' into releases
       new  46e9cfb   Merge branch 'releases' into dfsg
       new  551af8d   Merge branch 'dfsg' into debian
       new  a6d581f   changelog for fresh upstrem release: all debian/up_ patches removed, deb_ patched updated
       new  c805989   Boosted policy to 3.9.3 (should be ok without changes)
       new  b4041c3   added failing test for clone
       new  70099dd   rm instance variables learing_rate_type, loss_function, and penalty_type; create them before plain_fit
       new  dfd2140   move get_loss_function to _partial_fit
       new  185d326   add test for proper loss instantiation
       new  0c59872   n_iter must not be 0
       new  63608c6   refactored input validation; special loss function factory for huber and epsilon insensitive loss
       new  6a23fc8   use DEFAULT_EPSILON consistently
       new  01d3486   patches/deb_disable_test_spectral_old_scipy - disable unittest on older scipy's due to failure
       new  702ab6e   updated copyright years and added owner for borrowed ATL_* code
       new  e797a9b   fixed typo
       new  b9ed384   removed obsolete (and not used any longer) python-psyco from Recommends
       new  60c730d   rename get_loss_function to _get_loss_function
       new  f54c0bc   COSMIT use property.setter in sklearn.svm
       new  2933920   ENH performance of TfidfTransformer
       new  5a320d7   COSMIT replace useless safe_sparse_dot in chi2 with np.dot
       new  7dd9dd9   Merge remote-tracking branch 'upstream/master' into sgd-clone-fix
       new  16a5719   BUG fix broken top-10 features printing in text clf example
       new  cde7d2a   BUG: fix setuptools feature
       new  2425381   Print running time as a floating-point number with two decimals.
       new  2caa484   DOC copyedit HMM documentation
       new  cb98e77   Merge pull request #1138 from fabianp/doc_float
       new  420a006   COSMIT const and void* correctness in liblinear wrapper
       new  9987b61   DOC : fix rho=1 is L1 penalty #1139
       new  39a7b26   ENH: pass verbose consistently in forest module
       new  ac5e028   added test to expose the predict_proba w/ sparse matrix regression
       new  ad9cc06   fix the predict_proba w/ sparse matrix regression by using shape instead of len
       new  523ba8f   cosmit
       new  e14dd67   ENH address issue 1028: clone estimator in RFE
       new  57574dc   ENH issue 1115: grid search support for rfe via ``estimator_params``
       new  6fff265   ENH fixed bug in sparse RFECV, fixed bug in RFECV init (step was always 1), added decision_function and predict_proba to RFE and RFECV
       new  aa20467   MISC rfe outputs loss, not score
       new  ea3f32b   FIX typo
       new  9de51ea   add y to tfidf vectorizer
       new  0eca37d   cosmit
       new  7063196   ENH refactor liblinear prediction code and add classes_ member
       new  6fd64d4   COSMIT liblinear C code cleanup
       new  5d357f7   followed @larsmans tip to get rid of _decision_function
       new  4390eff   Check for memory errors (1)
       new  8a3c77b   Check for memory errors (2)
       new  14c63aa   Check for memory errors (3)
       new  71f9388   Mechanism to propagate optional estimator.fit arguments when using CV
       new  4de6d3f   changed **fit_kwargs to explicit fit_params dict
       new  ba857ef   make sure that param has len attr + a test
       new  cf9444b   replaced assert with assert_true + error msg
       new  c599668   fix failing test
       new  13be59d   TST : use nose assert_true and not python assert
       new  ae3a1fb   Merge pull request #1015 from vene/cov-speedup
       new  9feb095   fix docstring of predict_proba
       new  7bc9b71   Avoid useless if-statements
       new  c763dea   BF: Address issue #1059 in GMM by adding a supplementary check.
       new  3ed794e   COSMIT comment out more unneeded liblinear code
       new  3bbed29   Added a comment to clarify initial capacity
       new  3a73c2b   BF: Fix broken tests: change a check for compatibility with HMM.
       new  8727405   DOC + COSMIT LogisticRegression: docstring + rewrite predict_proba
       new  5ecb9ac   add predict_log_proba and test; better docstrings
       new  8aac72f   Merge pull request #1141 from pprett/sgd-predict-proba
       new  69176b1   DOC small fixes to SGD docstrings
       new  5528334   Robustify LARS. Fixes issue #487
       new  1e9f7ef   COSMIT rm svm.sparse tests to prevent deprecation warnings
       new  91aaabc   ENH micro-optimizations in SVMlight loader
       new  55a2070   WEBSITE updated logo, changed scikits-learn to scikit-learn.
       new  179eabc   BUG rm RidgeClassifier from 20newsgroups
       new  f28d088   ENH: Isotonic regression
       new  be9261b   MDS is now using the new isotonic_regression submodule
       new  2734835   Added tests to isotonic_regression
       new  c1abbf2   DOC - added paragraph in user documentation on the isotonic regression + an example plot.
       new  ccd1ad3   More documentation
       new  4bc740c   ENH : proper IsotonicRegression model + example + test
       new  af22f29   remove support for extrapolation
       new  f7a5fc1   FIX : for test_common sparse support
       new  01b5a97   New (faster) implementation of isotonic regression
       new  462f304   pep8
       new  b0bdd3f   adding my name in IR example
       new  1f59e07   FIX IsotonicRegression only takes vector input, hence don't test it in the common estimators
       new  aac3b2c   ENH IsotonicRegression now uses variable names that have more than 3 letters
       new  bc8fd0b   ENH better error messages on the IsotonicRegression
       new  2543d93   Added a predict method to the IsotonicRegression
       new  d8d4af0   ENH : finish addressing @GaelVaroquaux comments + improve coverage + add linear regression to example
       new  74d6f80   typo
       new  8511d96   MISC: small docstring work
       new  1acd247   FIX : fix LLE test (don't ask me why...)
       new  65cf665   ENH remove some deprecated parameters / functions /estimators
       new  52a9f4d   FIX remove test for deprecated parameter.
       new  b48b7f5   Use fixed random state in isotonic regression example.
       new  1be816d   Example: added a pretty PCA 3D plot of iris, as this dataset is used in so many examples. Unfortunately the mlp bug about 3d rotations and colors makes it not as nice as it could be.
       new  c3ea87d   ENH minore example beautification
       new  81c8b6e   Note on the use of X in isotonic regression.
       new  85693d6   Fix confusing notation in isotonic regression.
       new  6fbb0ea   DOC fixed default value of ``compute_importance`` in DecisionTreeClassifier docstring.
       new  6f13021   Merge pull request #1143 from larsmans/refactor-liblinear
       new  7d31400   metrics.py: modified precision_recall_curve to lower computational complexity
       new  d6d2dcf   metrics.py: pep8 and other cosmetic changes
       new  c979924   Fix latex formula in isotonic regression doc.
       new  a02e58c   Release manager change + fix Satra's URL.
       new  e6a9f3a   metrics.py: Added more comments to precision_recall_curve.
       new  8a96ca1   ENH no more distinction between "sparse" and "dense" LinearSVC
       new  f15e580   COSMIT rm deprecated SGDClassifier.classes property
       new  9d33497   COSMIT clarify L1/L2 LR sparsity demo
       new  e9d227d   DOC fix link for IsotonicRegression
       new  2247a31   misc
       new  d03b605   DOC fix IsotonicRegression docstrings
       new  ae24236   BUG allow array-like y in RFE
       new  b4aa0cb   DOC RFE docstring + link RFECV in narrative docs
       new  c2aabf0   BUG rm LARS from linear_model.__init__
       new  dc94372   COSMIT refactor linear classifiers
       new  0696777   DOC : avoid mentioning ElasticNet in Lasso.fit docstring
       new  8d318b2   DOC typo in ElasticNet docstring
       new  acda3f6   DOC add isotonic regression to References (even if we move it soon), also OCD.
       new  c4d9e87   FIX error in error message ^^ closes #1155.
       new  66c8a70   ENH fix percentile tiebreaking, add warning
       new  1373bcd   DOC document attributes scores_ and pvalues_ in feature selection docstrings, some superficial cleanup.
       new  9fb1499   DOC somewhat improved feature selection example
       new  88a0e5d   TST improve Ridge test
       new  4183cb3   COSMIT use LinearClassifierMixin in RidgeClassifier
       new  109b07a   ENH in NMF only use svd initialization by default if n_components < n_features.
       new  912f2bc   FIX fixed typo in code, added smoke test.
       new  47fd441   Merge pull request #1144 from glouppe/tree-malloc
       new  cf778c0   BF: fix issue #1127 about MinCovDet breaking with X.shape = (3, 1)
       new  2e48248   COSMIT + DOC univariate feature selection
       new  f5bcc3d   COSMIT re-indent docstring for safe_mask
       new  f11454d   BUG make GridSearchCV work with non-CSR sparse matrix
       new  ea51101   COSMIT remove unused imports
       new  35704c8   COSMIT rm deprecated class_weight from fit in Ridge
       new  33efa17   ENH Improve Ridge's conjugate gradient descent
       new  9d5936e   Revert "BUG rm RidgeClassifier from 20newsgroups"
       new  d93b33e   ENH add max_iter argument to Ridge estimators
       new  6300abd   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  684f17f   Adding guide on how to contribute to project
       new  f0c816d   Fix indentation
       new  3143bbe   Removed tabs from indentation
       new  aa03dc5   COSMIT: noting that PRs don't send mail to mailing list
       new  c3a510c   Moved link for further info to be more prominent
       new  c57402f   Minor changes to contributing
       new  53009c1   DOC added Conrad Lee's PR to whatsnew
       new  1313f25   COSMIT pep8
       new  c877742   FIX unicode support in count vectorizer. Closes #1098.
       new  986ad8c   FIX docstring for count vectorizer. Sorry about that.
       new  0e00956   DOC Ridge improvements in whats_new
       new  46f4c55   Merge pull request #1169 from larsmans/ridge-cg
       new  77e69ed   COSMIT rm deprecated stuff -- lots of it
       new  3ca3e95   DOC rm references to deprecated stuff
       new  a164d57   TST writable coef_ and intercept_ on LogisticRegression
       new  4c11ccd   COSMIT remove unused import
       new  b5339a9   Added the paper I used to implement isotonic_regression.
       new  77dc060   ENH add MinMaxScaler, #1111
       new  cf4940a   ENH do normalization in single pass over data
       new  485854a   DOC added missing docstrings
       new  c0a0578   ENH rename Scaler to StandardScaler everywhere
       new  3941a4b   COSMIT pep8
       new  8c0bb9e   BUG: parallel computing in MDS
       new  d82ff7b   DOC remove sparse support from docstring as there is none. Also cosmit on docstrings.
       new  c742e2a   BUG: deprecated k parameter in MiniBatchKMeans
       new  6fd39fb   Add support for preference contraints in svmlight format.
       new  7842748   BUG: copy and keep ordering
       new  e0d45a9   BUG: remove leftout debug prints
       new  398fa9d   FIX: query_id parameter and other cosmetic changes
       new  fa93e20   FIX: wrong probabilities for OvR LogisticRegression
       new  2cae34b   Move solver option to constructor.
       new  107abb6   Add lsqr solver.
       new  0471092   BUG: transmit parameters correctly from Ridge to ridge_regression.
       new  8c9cc85   Can afford better precision in news20 example.
       new  abee624   Fix docstrings and doctests.
       new  ec07896   Add minimalistic test for each solver.
       new  c5892aa   Fix damp parameter.
       new  1e11730   Fall back to dense_cholesky if sample_weight is given.
       new  15ce055   lsqr is not available in old scipy versions...
       new  71fd662   Better documentation on the choice of solver.
       new  7575410   PEP8!
       new  2a7ebd9   Cosmit: not a fan of defining a function in a loop :)
       new  ef336b6   Update what's new.
       new  73c21e2   More accurate API change description.
       new  0c2b07f   Fix warning message.
       new  58c1b5d   Add test for load_svmlight_files
       new  c0d6c6c   Merge pull request #1182 from fabianp/svmlight
       new  6c1e0ed   FIX: typo in ValueError message.
       new  04b5e98   Add support for query_id in dump_svmlight_file
       new  8aeb9ea   DOC: added svmlight qid support to whats_new.rst
       new  531dcb0   ENH: make test_common check normalized probabilities
       new  db37872   DOC: Enet alpha=0 => advice to use LinearRegression
       new  749ea70   ENH let DictVectorizer build a CSR matrix directly and use array.array
       new  d7e118b   MISC: add ltsa in docstring
       new  baf39ad   DOC DictVectorizer returning CSR in ChangeLog
       new  339c854   Merge pull request #1193 from larsmans/dictvectorizer-csr
       new  4b4d8fd   ENH add FeatureStacker estimator
       new  7b78abb   ENH add feature stacker example
       new  67c5a07   COSMIT + DOC more dosctrings, minor improvements
       new  26d17c6   ENH implement get_feature_names
       new  d27e8fc   TST added tests, fix feature names.
       new  3d5e6cb   ENH add parallel fit and transform with joblib.
       new  ec08e8c   ENH add transformer weights
       new  dac5e59   TST add test for feature weights in feature stacker
       new  098cfbc   DOC move example (there is nothing to plot) and add some text
       new  78789a1   MISC renaming FeatureStacker to FeatureUnion, adding docs
       new  d087830   DOC added FeatureUnion to whatsnew.
       new  aa3641f   Merge pull request #1189 from fabianp/svmlight
       new  244e8a6   Merge pull request #1187 from ogrisel/bugfix-logistic-ovr-probabilities
       new  e0f9a97   FIX: broken doctest for DictVectorizer
       new  3f8f8ee   changed includes back - change broke JENKINS build
       new  41266a9   not the problem afterall - switch back
       new  92aaf2d   add first collapsible toctree test
       new  24e5fc6   moved buttons to themes
       new  3d4c900   working version
       new  305f2f1   Links now clickable
       new  5a96cf8   -collapse toc moved to front page- -firefox highlighting bug fixed -examples bug fixed
       new  75dc22a   button colour change + comments
       new  975d1f4   fixes - seemingly good version -sidebar removed from start page -sidebar size fixed on other long pages
       new  edf0d56   highlighting of + implemented
       new  271f06d   -line highlight bug fixed, buttons changed, full expansion added
       new  10b8b13   small bug fix and colour tweak
       new  5db0dd5   nitpick fix
       new  ed0a9dd   cleanups
       new  844c4b0   cleanups
       new  1a381d4   toggle bug fixed
       new  690e6d9   highlight fix
       new  6227904   what`s new updated
       new  4062b3e   Improve doc and error msg in MinCovDet in response to issue #1153.
       new  7a75fa3   ENH remove deprecated sparse SVM class from cross-validation test.
       new  9905618   ENH: MCD for large dataset
       new  24e7190   BUG in error message for k-means
       new  7ac8a3d   BUG in error msg for spectral clustering
       new  109b63d   BUG: propagate random-state in MCD
       new  bc0ff18   travis config file
       new  60f10ae   update travis config
       new  3f49b21   DOC: return values of make_moons and make_circles
       new  6f4e17a   Merge pull request #1197 from glouppe/master
       new  3c0fc66   put the requirements at the right place
       new  175dd82   added requirements to travis config file
       new  d50c0cb   COSMIT pep8
       new  40328ed   FIX bug in pipeline.inverse_transform, improve coverage.
       new  dc52457   ENH support for string labels in Neighbors classifiers
       new  ffc8656   ENH rename ``_classes`` to ``classes_``, fix outlier labeling, remove unnecessary mapping to indices.
       new  ee9b9ae   COSMIT reuse variable name
       new  adf3e2a   ENH added non-regression test
       new  0984a2a   cd_fast: use square norm directly
       new  aa69c56   COSMIT error messages in GenericUnivariateSelect
       new  19a1a41   ENH perform feature selection on scores, not p-values, when possible
       new  99f13ef   DOC some improvements to FeatureUnion docs
       new  e662539   ENH: using coo matrix construction to accelerate calculation of the contingency matrix
       new  a3dacb6   FIX: numerial issues in NMI
       new  9964264   COMIT pep8
       new  117177d   ENH add refs to issue #884
       new  8ce26ed   FIX: ADD test cases for exact 0 case, and nmi equal to v_measure case
       new  2d3f70a   FIX: accelerate v_measure calculation based on mutual information
       new  94147ed   COSMIT add doc to clearify how  nmi is normalized and pep8 fix
       new  9b38790   COSMIT pep8 fix for test_supervised
       new  7c8521c   FIX: fixes error caused by break line
       new  9d3fa6e   Using coo_matrix to accelerate confusion_matrix calculation
       new  f07378b   COSMIT
       new  39070ea   ENH add test for testing v_measure is a variant of nmi
       new  64b0832   COSMIT typos in doc strings
       new  310c558   FIX let test use random_state(seed)
       new  b23bbaa   PEP8..
       new  fa669d5   FIX typos and vague comments
       new  30e8a44   DOC add comments for log(a) - log(b) precision
       new  2834ed5   COSMIT fails to see the function name use mi rather than mutual information
       new  ffe44eb   FIX doctest to check up to 6 digits precision
       new  ccb93e5   FIX: eliminate \ for continuation from doctests
       new  094c636   COSMIT removed unused import
       new  89a6d1c   FIX np.unique doesn't have return_inverse keyword, use backport from utils.
       new  fb7e3ab   Python3 compat: print()
       new  40d1399   DOC LaTeX error in SVM narrative docs
       new  07a9d2e   ENH better error messages in CountVectorizer for empty vocabulary
       new  ee8c8ff   FIX: bug in multi-output forest.predict_proba
       new  3beae02   ENH address issue 1028: clone estimator in RFE
       new  a999ce1   ENH issue 1115: grid search support for rfe via ``estimator_params``
       new  2aacb3b   ENH fixed bug in sparse RFECV, fixed bug in RFECV init (step was always 1), added decision_function and predict_proba to RFE and RFECV
       new  4322a2f   MISC rfe outputs loss, not score
       new  b3722b5   FIX typo
       new  8f13917   BF: Address issue #1059 in GMM by adding a supplementary check.
       new  1890c2e   BF: Fix broken tests: change a check for compatibility with HMM.
       new  4a3b4a7   WEBSITE updated logo, changed scikits-learn to scikit-learn.
       new  b7f6f19   added test to expose the predict_proba w/ sparse matrix regression
       new  7ea873b   fix the predict_proba w/ sparse matrix regression by using shape instead of len
       new  94c7881   cosmit
       new  58d5a5e   followed @larsmans tip to get rid of _decision_function
       new  c6308ae   fix docstring of predict_proba
       new  326e4f6   add predict_log_proba and test; better docstrings
       new  03d28ee   added failing test for clone
       new  1791575   rm instance variables learing_rate_type, loss_function, and penalty_type; create them before plain_fit
       new  ee838a2   move get_loss_function to _partial_fit
       new  55e8810   add test for proper loss instantiation
       new  767ec5d   n_iter must not be 0
       new  8283126   refactored input validation; special loss function factory for huber and epsilon insensitive loss
       new  237867a   use DEFAULT_EPSILON consistently
       new  7d0347d   rename get_loss_function to _get_loss_function
       new  02e6fd0   BUG rm RidgeClassifier from 20newsgroups
       new  9a3910d   BUG allow array-like y in RFE
       new  33a73bc   Release manager change + fix Satra's URL.
       new  9d0cc93   COSMIT clarify L1/L2 LR sparsity demo
       new  890bb8e   DOC RFE docstring + link RFECV in narrative docs
       new  d65fae7   BF: fix issue #1127 about MinCovDet breaking with X.shape = (3, 1)
       new  8b0219a   FIX unicode support in count vectorizer. Closes #1098.
       new  5dcaaf7   FIX docstring for count vectorizer. Sorry about that.
       new  9fe86f6   BUG: parallel computing in MDS
       new  f7241ac   BUG: deprecated k parameter in MiniBatchKMeans
       new  133aaf0   BUG make GridSearchCV work with non-CSR sparse matrix
       new  08d170d   BUG: copy and keep ordering
       new  9e16ea7   BUG: remove leftout debug prints
       new  13fd25e   Mark Cython outputs as binary so their changes don't clutter diffs.
       new  669f31e   TST CountVectorizer with empty vocabulary
       new  46e3c84   Merge pull request #1208 from larsmans/check-empty-vocabulary
       new  536829a   Merge pull request #1211 from kcarnold/gitattributes
       new  121cb20   DOC : fix rho=1 is L1 penalty #1139
       new  6716a81   DOC typo in ElasticNet docstring
       new  5b55cee   DOC typos in README
       new  7efd8fc   ENH slightly better error message for robust covariance
       new  55ef1d6   DOC: protect `classes_` for valid rst
       new  4d73f2b   DOC: protect `classes_` for valid rst
       new  ed90a99   enh even better error message
       new  10d9023   ENH make multi-class more robust in discovering scoring functions
       new  e9fc59a   ENH in all_estimators, skip testing modules. They have dummies.
       new  7c70198   FIX: doctests under Windows 64bit
       new  f18295c   TST improve test-coverage in base, remove unreachable code-path
       new  70bf00c   DOC escape one more instance of "classes_" in SGD
       new  3f9d73d   FIX: doctests under Windows 64bit
       new  2bb7b91   ENH/FIX add a lobpcg solver to spectral embedding
       new  354b648   MISC: decrease verbosity by default
       new  6c519c8   FIX: numerical stability in spectral
       new  914048f   Update changelog
       new  91f7848   Update changelog
       new  98c666c   DOC feature selection by scores instead of p-values
       new  852363e   Merge pull request #1215 from amueller/pipeline_muliclass
       new  9701bb8   add dual_gap_ and eps_ to Enet and Lasso docstring
       new  ff6fe90   wip on fx interactions for GBRT
       new  03b8829   extend eps_ description
       new  892e1c5   Merge pull request #1223 from ibayer/master
       new  f2c52f1   DOC: use nosetests rather than sklearn.test()
       new  70cab6d   ENH: support arbitrary dtype in kNN classifiers
       new  06cd9a7   DOC: use nosetests rather than sklearn.test()
       new  f26f33c   ENH: support arbitrary dtype in kNN classifiers
       new  a00b3ce   TEST: predict_proba in knn classifier with y string
       new  0dd2e39   REL: 0.12.1 Bugfix release
       new  bb5d924   DOC: fix doc mistakes
       new  7ca9faa   DOC: another layout fix
       new  7cab9bd   DOC: add doc on making a release
       new  e8443db   Merge tag '0.12.1' into releases
       new  7e33533   Merge branch 'releases' into dfsg
       new  8c4304f   Merge branch 'dfsg' into debian
       new  1253750   changelog for 0.12.1-1
       new  5700100   COSMIT pep8
       new  fe1cd4b   DOC added whatsnew entry for mutual info fix and faster confusion matrix.
       new  ceb80b0   ENH rename k to n_folds and n_bootstraps to n_iterations
       new  ad4abdd   DOC cleanup some docstrings (not scipy standard yet)
       new  1b367e5   ENH set n_fold default to 3, rename k to n_fold in all doctests, docs, and examples
       new  0a81827   COSMIT rename n_iterations to n_iter in cross_validation
       new  9bd0aad   MISC renamed n_iterations to n_iter in all other places.
       new  c3e046a   DOC added changes / renames to whatsnew
       new  dfb31f5   ENH rewrite K-Means m-step as loop over samples, cythonize.
       new  9283b10   ENH separate sparse and dense case, cythonize further.
       new  9423e9c   DOC various typos and other minor stuff
       new  262cbdf   DOC clarify zero_based's implications in SVMlight loader
       new  f41c98c   ENH fix int type in kmeans
       new  27256a3   ENH fix kmeans for old numpy (bincount minlength) Should this go into the utils/fixes?
       new  6e1aec2   FIX also the other function in kmeans. whoops
       new  a3fabe1   Merge pull request #1204 from larsmans/mi-feature-selection
       new  3bfc42c   Merge branch 'master' into gbrt-interactions
       new  f9df471   FIX bincount mess I made in kmeans.
       new  16aabc7   ENH rename rho to l1_ratio in ElasticNet and friends
       new  b39609a   ENH rename rho in SGD
       new  72e71bb   ENH address @agramfort's comments, fix some doctests
       new  143147d   DOC add changes to whatsnew.
       new  04b31e2   ENH simplify as suggested by @larsmans.
       new  20f6540   implemented partial dependecy plot
       new  5edcbc1   fix: grid and model
       new  d6655fa   FIX for len(result) > minlength
       new  27baaf3   TST: cater for 0.9 not > 0.9
       new  ac7f399   cleaned tree traversal and sorted out weighting
       new  ab78baa   cythonized and cosmit
       new  6cd2f57   Fixed typos.
       new  d637d86   Merge pull request #1237 from kalaidin/typos
       new  a744f31   DOC tried to clarify meaning of l1_ratio in whatsnew
       new  fba0d35   BUG: obey numpy 1.7's stricter rules
       new  f6dd870   Merge remote-tracking branch 'origin/pr/1234'
       new  d7152a5   BUG: cater for dev versions of numpy
       new  1929892   Factor Analysis -- implemented with EM + SVD.
       new  bf96a14   ENH : cleanup FactorAnalysis object
       new  dd58d6e   API : rename psi to noise_variance + some cleanup in FA
       new  f85b658   TST : add test that FA log like increases over iterations
       new  8666896   add Bishop's book to refs in FA
       new  46d1631   TST: Make factor analysis test repeatable.
       new  f881485   Extended faces decomposition example with Factor Analysis.
       new  63ce79c   Factor Analysis learns variance of generative model for every dimension. Illustrated with faces.
       new  cbe1b3d   pep 257.
       new  3a46f75   Make sure that psi=0 does not break em.
       new  3732ccf   Some documentation for FA.
       new  d3ea9f6   More or less same code already available.
       new  eac061b   Plot noise variance for FA. Changed some things to make plot_gallery usable for this, too.
       new  3d769f5   Adding some plots for FA. Ordering of articles must be adopted.
       new  8b8a1a8   Extended test a bit.
       new  1458a75   Added score function.
       new  7f8c087   Two iterations are enough for the test.
       new  d2b1573   score works like ppca.score().
       new  229f813   adapted to new signature of score().
       new  345d6fb   Moved paragraph on FA before ICA.
       new  5129029   update what's new with FactorAnalysis
       new  93a2fcf   DOC : adding FactorAnalysis to classes.rst
       new  97ed661   BUG + DOC l1_ratio in SGD and CD
       new  53b2203   COSMIT correct error msgs in SGD and make them more consistent
       new  3bc0276   Merge branch 'pr/1214'
       new  2d4dd89   FIX: missing figures in FA narrative doc
       new  9e7d1a6   MISC: use toarray instead of todense
       new  7f20345   automatically create grid from training data cosmit docstrings
       new  4bd2e9b   add cartesian product
       new  1b6161a   partial dependency plot example from ESLII 10.14.1
       new  cfed074   ENH remove some unreachable code from gridsearch
       new  4b66eba   BUG: RandomizedPCA needs random_state set
       new  8141cb1   ENH: make RandomizedPCA.fit idempotent
       new  20ede3b   DOC let BibTeX handle its own capitalization, except for {P}ython
       new  b77804d   BUG NaN handling in SelectPercentile and SelectKBest
       new  342c5be   ENH: Consider order in X for IsotonicRegression.
       new  04251e1   Better tests + cosmetic changes.
       new  8d3c75f   Store X as an ordered array.
       new  a06c367   ENH sparse matrix support in randomized logistic regression
       new  60ae168   COSMIT rm unused import
       new  8123769   ENH: adding iter_limit to libsvm
       new  8fa79ee   FIX: committing updated Cython-generated libsvm bindings
       new  fb48f29   ENH: Solver iter_limit emits warning instead of raising exception
       new  5d6540d   ENH: renaming iter_limit -> max_iter
       new  3c803da   FIX: missing file hidden among the Cython output
       new  4b8399b   ENH: hint about data normalization when SVC stops early
       new  7fab72d   FIX: adding missing c files from cython
       new  159fb12   ENH: assert -> assert_equals
       new  51b06b8   FIX doctests for max_iter
       new  7be6c2c   FIX two more docstrings. Sorry.
       new  00be697   FIX seed liblinear using srand. Fixes issue #919.
       new  f711bf2   TST: fix doctest
       new  fc236b1   TST: fix test counting warnings
       new  6ac7fbf   Merge branch 'master' into pr/975
       new  bfbcc78   docstrings for init and loss_
       new  ff3097f   cosmit
       new  b321d4c   added Emanuele to authors
       new  e64cea0   ENH add random seed to logistic regression
       new  66fc977   ENH don't use deprecated interface in PPCA & cosmit
       new  e09db85   FIX issue #1239 when confusion matrix y_true/y_labels has unexpected labels
       new  a42a91a   PEP8
       new  43d083d   ENH docstring misleading
       new  f314380   Merge remote-tracking branch 'upstream/master' into pr/975
       new  35860aa   Merge branch 'master' into gbrt-interactions
       new  83bfb4a   Merge exthmath tests into the same file.
       new  e69acb1   Add common assertions to sklearn.utils.testing.
       new  a2cd5fa   Fix density utility when input is sparse.
       new  3f42c71   Typo.
       new  f358799   Fix test failure.
       new  b24e60d   Adding Passive Aggressive learning rates
       new  ba52957   Added documentation to stochastic_gradient
       new  3dbd135   Added to documentation
       new  e9ebfeb   Added documentation and removed PA
       new  6610659   Added tests
       new  74ef738   COSMIT: spelling correction
       new  6f80695   Adding example
       new  8ff7407   Added smoothing to example
       new  70a966b   COSMIT typo
       new  297498c   PEP8 fix
       new  c68e607   PEP8 COSMIT
       new  dfe33d0   PEP8 COSMIT
       new  77b585f   Merge branch 'master' into gbrt-interactions
       new  8246645   Merge branch 'master' into gbrt-interactions
       new  021e9f9   add learn rate to partial dependency function
       new  0000eb1   common ylim; comment out 3d plot
       new  b7e9d17   Enforcing non-negative step-size
       new  50627fc   FIX: prevent early stopping in tree construction
       new  34cdc4e   FIX: prevent early stopping in tree construction (2) + Test
       new  aa58f86   make fit_stage private
       new  56d285f   return axes instead of grid
       new  807bc0e   3d plot of 2-way interaction plot
       new  b8dbef6   Fix typo in README
       new  9deaa43   Merge pull request #1266 from cdeil/patch-1
       new  f4e2e1a   REL put myself as contact / maintainer, fixed url
       new  e134a27   Merge pull request #1263 from glouppe/fix-1254
       new  67890e4   add verbose output for gradient boosting algorithms
       new  5dfcff0   Changed verbose to int, added a low verbose option to just print '.'.
       new  513d2d9   remove '\r' and format numbers to be fixed width, 7 digits of precision
       new  2954079   fix GradientBoostingClassifier by passing verbose as a keyword argument
       new  676b97f   FIX rebase mishap
       new  bf39cd9   Merge branch 'master' into gbrt-interactions
       new  d952ed2   multi-class is supported
       new  2a3ca6b   cosmit
       new  465c8f7   doc: use n_iter instead of epochs; remove backslash
       new  0062ba1   Merge branch 'master' into gbrt-interactions
       new  b49768f   california housing dataset
       new  40e483e   cosmit
       new  62d7a90   use California housing dataset loader
       new  7613fa9   DOC small example / doctest for kernel approximation.
       new  38155aa   First go at speeding up Euclidean distances
       new  886270a   Make it less yellow
       new  78a9fd5   More reusable code, speed up symmetric case
       new  8c45f55   Better cython style.
       new  480f462   Add dense sparse support and precomputation
       new  73c2c34   FIX: buggy case when X=dense, Y=sparse
       new  a7ae308   Consistent argument naming and useful maintenance notes
       new  5666b8d   FIX using out with sparse matrices
       new  f560fe4   Relative imports, fix todense bug
       new  a3b0c4c   safe_sparse_dot into preallocated output
       new  cb4879d   Add test for dense_output, fix bug, cleaned up logic
       new  3b1a0c2   Avoid reallocation in manifold.mds
       new  bad7f14   add type prefix to blas funcs
       new  8066f9b   DOC Clarify the docstrings
       new  8ec5ecb   Added Cython-generated euclidean_fast.c
       new  230fff7   BUG: follow scipy API change
       new  ab3304f   Separate dense_output and out parameters, document better
       new  1b1829b   API change: mutually exclusive preallocation and precomputation
       new  1103240   FIX: csr_matrix induced unwanted copying
       new  7da9bf0   Rename euclidean_fast to _euclidean_fast
       new  bd7badb   Clean setup.py in metrics
       new  ca703ae   ENH: improve test coverage
       new  88f7df2   COSMIT website address + copyedit in __init__.py
       new  ed1d5e2   DOC move implementation details on mixins to comments
       new  711fb96   DOC: fix 'Controls' typo in sklearn.ensemble.forest.
       new  9f9f5ce   Merge pull request #1269 from mrjbq7/doc-fixes
       new  0dd841f   Revert (rebased) merge of euclidean_distances speedup
       new  8f22486   Split out PassiveAggressive Classifier into its own object
       new  5590ce5   Adding PassiveAggressiveRegressor estimator
       new  5a3fc24   COSMIT
       new  c8bf18b   COSMIT: fix typo in AUTHORS.rst.
       new  9d3bcf1   DOC typo in whatsnew
       new  77358b5   DOC more typos in whatsnew.
       new  1406cdf   ENH use the numbers module introduced in Python 2.6 to check for number types. Fixes 1255.
       new  82dc425   DOC: typo
       new  afb666c   Use sklearn.utils.testing in tests.
       new  6e8696b   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  f65efcf   More use sklearn.utils.testing.
       new  76972db   Added documentation for new classifier and changed seed to random_state
       new  6eec2e6   Even more sklearn.utils.testing.
       new  eca762d   DOC: improve the model selection exercise
       new  fd9e1c2   ENH added OneHotEncoder
       new  2660482   DOC minor fixes / typos. Thanks @larsmans.
       new  0860182   ENH user-specified dtype, defaults to np.float, nicer numpy stuff :)
       new  c305629   TST skip test in common_tests, reach 100% coverage on new code.
       new  deebe78   DOC more typos omg. comment about automatically inferring maximum values.
       new  89257b6   ENH better example.
       new  14bf28e   DOC: further improvements to the model selection exercise
       new  3a6bb73   Fixed typo
       new  fd595a7   Renamed learning_rate loss in PassiveAggressive
       new  5d1bda6   Correct documentation
       new  60d20c1   Corrected doctests
       new  f2e9d5b   enh masking out zero features
       new  e04614c   TST fixed doctests, added more tests. Still 100% line coverage :)
       new  c37cef2   FIX : fix application example due to API change
       new  425c6e9   DOC: further improvements to the model selection exercise
       new  315d421   ENH/FIX add a lobpcg solver to spectral embedding
       new  ba400b7   MISC: decrease verbosity by default
       new  d4a6958   FIX: numerical stability in spectral
       new  9a4aefd   MISC: addressing @satra's comments
       new  2c857f2   ENH: make sure that spectral EVD is solve once
       new  b0f3a01   MISC: @agramfort's comments
       new  083c507   PEP8
       new  d0f7da2   doc: added reference to lobpcg and note about small number of nodes
       new  402a304   BUG: fix test error
       new  e8fe8e6   Fix indentation
       new  c326b1f   ENH removed ``remove_zeros`` parameter.
       new  90c0079   BUG: make precision_recall invariant by scaling probs
       new  2b85071   BUG: fix setuptools feature
       new  7eda399   Discretization method for spectral clustering added along with tolerence setting to loosen eigendecomposition constraints
       new  c615759   Documentation and small bugs fixed and code cleaned up
       new  b2502fc   Small comments/constants added
       new  bbfe61a   Fix typos
       new  851a27c   Added more info in documentation
       new  8ba410c   Small aesthetic fixes to discretization
       new  3f00ff2   pep8 formatting
       new  40701a7   More description of the discretization algorithm.
       new  e500133   Even more description of the discretization algorithm.
       new  36efa06   Documentation changes, removed more camel case variables
       new  01217f3   Fixed some memory inefficiencies and clearned up documentation and code semantics
       new  64bfe41   Example for spectral clustering embedding handling
       new  112597d   Added newline to the end of file
       new  0b53cb5   removed a hardcoded value
       new  efb1344   Modified lena segmenation example to include different embedding solvers
       new  d3f1d77   Removed savefig
       new  9b0ee1d   MISC: split example in two plots
       new  388d123   API: change 'embed_solve' to 'assign_labels'
       new  a69557e   TST: increase coverage in spectral clustering
       new  d4f53ed   DOC: add docs of assign_label in spectral clustering
       new  0b3bc3f   COSMIT: long remarks go in 'notes' section
       new  8b6f280   ENH allow more than 1000 linear SVMs with custom random seeds
       new  0a01292   BUG: restore numpy 1.3 compatbibility
       new  79fcb69   fixed bug of initialization in hmm.py
       new  0ad0b3c   added test_fit_with_init to tests/test_hmm.py also fixed a bug by this test
       new  6755803   pep8, ignored E126-E128
       new  e9f4645   - avoid startprob, transmat, emissionprob containing a zero element by   adding small number and re-normalize - test case for above problem
       new  d5fcb73   - check input format of MultinomialHMM.fit   raise ValueError in MultinominalHMM if the input does not pass the check
       new  e3d369d   MISC: minor clean ups in hmm code
       new  6f3ccc1   DOC more extensive classifier comparision on synthetic datasets
       new  0ddc287   ENH more noise, cross-validated parameters.
       new  7353a5e   ENH train/test split, plot accuracy, make plot pretty.
       new  72c360c   ENH simplify circles dataset generator, make classes balanced.
       new  7c0d1b6   FIX typo in dataset generation
       new  15b2168   ENH I'm more happy with the last example now....
       new  301733c   Merge branch 'master' into gbrt-interactions
       new  906b942   remove legacy code
       new  ff31326   FIX adjust gamma in kernelPCA tests to fit slightly modified circles with balanced classes.
       new  0563fa9   BUG halve the number of LinearSVCs
       new  77b4670   COSMIT use np.clip in SGD
       new  e683e64   Merge remote-tracking branch 'upstream/master'
       new  87cf514   DOC: another improvement to the model selection exercise
       new  00d1055   FIX HMM test failures
       new  6670c17   Merge remote-tracking branch 'upstream/master' into gbrt-interactions
       new  87d3246   renamed dependency -> dependence; docstring and cosmit
       new  18b3b1d    typo
       new  23fb18d   fix: feature_importances_
       new  df909c7   rename dependency -> dependence
       new  c8145ce   rename dependency -> dependence
       new  cc66f14   add partial dependence plot example
       new  3adaaa7   FIX : pass random_state to kmeans in gmm.fit
       new  c5e7807   Merge pull request #1292 from aymas/pass_rng_kmeans_gmm
       new  189af52   Merge pull request #1290 from tjanez/master
       new  edcfcd3   COSMIT: pep8 in arrayfuncs.pyx
       new  427a0e6   ENH used asarray to avoid copy
       new  917f309   ENH fit_transform on KMeans
       new  3c854a2   COSMIT pep8
       new  e8bdac9   enh: add code analysis target to makefile
       new  d70088c   FIX small bug in feature selection error message.
       new  5cfb764   COSMIT do less deprecated things.
       new  81b0f36   for now excluded some labile tests from testing
       new  ad3bfc8   added patch changeset_5528334660d8901263b4157038f41a0f4028a5f3.diff to "cherry-pick" 5528334660d8901263b4157038f41a0f4028a5f3 (Robustify LARS. Fixes issue #487)
       new  6db567b   changelog for prev commit
       new  01d4db9   ENH input validation in chi2, error for negative input
       new  4104f96   FIX revert useless change.
       new  bd3a6c2   DOC warn about parallel k-means on OS X.
       new  9babaa7   ENH minor improvements in testing, new utility function for safely setting random states.
       new  b4f1c4f   Test qda with 'priors' parameter
       new  a51f911   Test QDA.covariances_ attribute
       new  e4355fb   Don't cover this deprecated method
       new  a361f0e   Test non-normalized GaussianProcess
       new  c1ffbb4   Missing random_state in LinearSVC.
       new  0d59309   Test _BaseHMM._decode_map
       new  1e826c7   Test _BaseHMM.{predict,predict_proba}
       new  635e7f3   FIX cross_val_score now honors ``_pairwise``
       new  342dddc   Make this bit of code more compact (and improve code coverage).
       new  bd87dee   Remove unused code branch.  (_hmmc must be always available nowadays.)
       new  f67e0d2   Fixed docstrings and seed tests
       new  189ac2f   renaming 'learn_rate' in 'learning_rate'
       new  0687b2f   DOC added my last PR (cross_val_score fix) to whatsnew
       new  bbc272c   Remove stale test code
       new  470c769   Remove obsolete comment
       new  796f6c6   Improve cross_validation test coverage: 94% -> 99%
       new  2605a6b   ENH hompage add links to headers in left panel
       new  19f5dae   ENH add link to Citing
       new  a1568d7   WEB color fix for link headlines
       new  7f45f2c   Improve metrics.metrics code coverage: 95% -> 100%
       new  f35ab4b   DOC document callable kernels in SVM docstring.
       new  913fe77   DOC add user guide for MinMaxScaler
       new  3d1c8fb   Improve svm.base test coverage: 92% -> 98%
       new  4a477e1   Add docs for `vocabulary_` and `stop_words_` attributes of Countvectorizer.
       new  4fe5a92   Merge pull request #1323 from dnouri/countvectorizer_doc_1154
       new  e1cbdee   FIX: vocabulary_ maps to feature indices.
       new  d8fa623   Merge pull request #1320 from dnouri/test_coverage
       new  b32fc9c   ENH renaming 'max_iters' to 'max_iter' for consistency
       new  6941d8d   DOC missing class mention
       new  ca8a8eb   ENH renaming 'n_atoms' to 'n_components' for consistency
       new  e2efecd   ENH fix pep8
       new  33a5911   COSMIT in mean shift docs
       new  40f1872   FIX : missing import warnings
       new  2064745   typo
       new  a164a98   typos
       new  25284b9   Merge branch 'sgd_learners' of https://github.com/zaxtax/scikit-learn into passive_aggressive
       new  d9bb6ac   Rename pa.py to passive_aggressive.py.
       new  00c380b   Cosmit: random_state is not necessary.
       new  65b035f   Fix many bugs and test PA-I.
       new  1589d54   Do not expose C in SGDClassifier / Regressor.
       new  8b73f40   Implement and test PA-II.
       new  5a313ba   Add SquaredHingeLoss.
       new  89e06ad   Test different losses.
       new  97f594a   Add squared epsilon insensitive loss.
       new  0dcb61f   Test PA-II (regression).
       new  d38c0d2   Fix random_state in SGD.
       new  221d153   Update narrative documentation.
       new  d184d62   Fix example.
       new  6b6abcd   Credit myself.
       new  7fadf53   Fix see also.
       new  e9dde49   Fix a few test failures.
       new  da98830   Add one more test for PassiveAggressiveRegressor.
       new  4661b64   Fix underflow detected by test_common :)
       new  8788713   Update document classification example.
       new  462bbb2   DOC: typos in ensemble.rst
       new  9385763   Fix doctests.
       new  fdb4923   DOC: typos in ensemble.rst
       new  3407214   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  663a6ae   Better documentation for C.
       new  56df9dc   Add PassiveAggressive* to class reference.
       new  0f92da4   Remove sample_weight and class_weight from PassiveAggressive*.
       new  27be154   Add tests for partial fit.
       new  2255944   remove `steps` from Attributes of docstring
       new  06879d0   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  19042cb   Merge pull request #1331 from jaquesgrobler/master
       new  3d7ff31   FIX : add condition pos_label!=None for multiclass purpose in metrics.precision_recall_fscore_support
       new  e152473   TEST : add a test, test_precision_recall_f1_score_multiclass_break(), that breaks with current master and now works
       new  922c7ae   Change metrics.py as before and shorten test (test_precision_recall_f1_score_multiclass_break() in test_metrics.py) to show where it breaks
       new  9cb0af2   FIX : clean test + pep8 + reply fix to the code
       new  2d8bdbf   Update to the clustering.rst module file for k-means. Added a plain language description and the objective function.
       new  5706109   Document epsilon.
       new  3264268   Better documentation for epsilon in SGD.
       new  370e3ca   BUG: dot on sparse matrices broken in recent numpy
       new  a3670c0   Remove predict_proba from Perceptron and PassiveAggressiveClassifier.
       new  bf87bf5   Merge branch 'master' into pr/1279
       new  47d8645   DOC: Improved the code that shows how to export a decision tree to Graphviz and generate a PDF file.
       new  94fa4d1   Skip doctest for the Python code involving pydot.
       new  6b6e1ab   Skip doctest for the remaining line involving pydot.
       new  9be9072   FIX hotfix for NMF sparsity problem
       new  f07a10a   BUG: fix doctest bug
       new  05ea3e4   BF: GridSearchCV + unsupervised covariance shrinkage selection.
       new  180cbfb   Change legend + complete docstrings.
       new  67fabab   Improve example narrative doc (rewritten intro).
       new  cf54268   DOC: improve wording in covariance docs
       new  4d3ac03   Fix typos in doc.
       new  109d102   Add y=None to covariance estimators for API consistence purpose.
       new  02cbb24   DOC: typo
       new  f8d7bd4   ENH OneHotEncoder docs + TypeError + test active_features_
       new  09714bc   cythonized expected_mutual_information
       new  5e97624   added authors
       new  e33adcb   ENH cut down on memory use of text vectorizers
       new  3403730   Remove transform from PassiveAggressive*.
       new  7a2681d   Changed example svc kernel to be linear, however the error curve ends up flat under the new kernel.
       new  5f40bdf   Used more extreme values of C to show a more pronounced error curve. See the new curve here: http://imgur.com/UF7I4
       new  e6fd534   Took out a save image line
       new  c188755   Edited docs to reflect change in kernel used.
       new  205d658   added yticks
       new  81ff85d   added yticks
       new  f5cf0ee   added yticks
       new  c14e9c6   limited range of C cross validation
       new  252f73c   COSMIT: pep8, wording, layout
       new  76b3d9c   DOC: fixed string formatting in example
       new  177e813   Switched dynamic default args in random forest
       new  668979d   Added test
       new  c65c35c   Switched default parameter to tuple from lists.
       new  e0fdf95   move tuple back into arguments
       new  2d0ab3c   MISC: remove unused import
       new  4f41de9   BUG: LassoLars path ending contained junk
       new  e115469   TEST: one addition test on the length of the path
       new  48dd2ac   TEST: test that alpha is decreasing in LassoLars
       new  65b79cf   BUG: lars corner case with path length == 1
       new  e18465d   ENH: multi-target Lars: lists rather than arrays
       new  156566e   ENH: early stopping LARS for degenerate active set
       new  a911ad1   MISC: address comments
       new  caf995c   MISC: more precise warning
       new  f739070   WIP: drop for good correlated regressors
       new  c3b5ea2   MISC lars_path: cleaner code in degenerate case
       new  e290951   ENH: early stopping for lars
       new  2d9435c   TST: add a test for lasso and lars
       new  428439f   MISC: comment
       new  16dbe52   ENH lars_path: early stopping after drop for good
       new  1f3f497   TST: difficult test for early stopping
       new  de4fd32   COSMIT: better comments
       new  25707c2   BUG: missing import introduced by rebase
       new  0eb8578   DOC: Update whats_new with lars improvements
       new  9e52350   BUG: compat with numpy 1.3
       new  8bb0b68   DOC: spelling
       new  9f18586   BUG: AUC should not assume curve is increasing
       new  7c877ca   Fixed an issue where CountVectorizer.decode leaves file pointers open after reading the contents of the file. This produces unpredictable behaviour as closing the file pointer is left to the implementation of the python interpreter.
       new  07a2b24   Changed the CountVectorizer charset default back to 'utf-8' instead of 'utf8'. This was due to debugging on my local machine.
       new  1f92046   document sample_mask and X_argsorted in BaseDecisionTree.fit and validate inputs using np.asarray (added two tests as well)
       new  d3dfbec   Merge branch 'master' into gbrt-interactions
       new  ef60b17   tidy up deprecated warnings for learn_rate
       new  c0d80d6   Merge pull request #1344 from mattilyra/CountVectorizer.decode
       new  bdb0f64   COSMIT
       new  76b3130   FIX: missing # for comment in pyx file and readded missing AMI docstring
       new  970f8b2   DOC: LinearRegression document the shape of coef_
       new  fde726c   DOC: n_responses -> n_targets
       new  9cf5f00   FIX dirty fix for expected mutual info in cython.
       new  ff655ad   ENH added OneHotEncoder
       new  1a3c19c   DOC minor fixes / typos. Thanks @larsmans.
       new  b744d92   ENH user-specified dtype, defaults to np.float, nicer numpy stuff :)
       new  c861ad7   TST skip test in common_tests, reach 100% coverage on new code.
       new  2e1b455   DOC more typos omg. comment about automatically inferring maximum values.
       new  d126074   ENH better example.
       new  94218a4   enh masking out zero features
       new  8f55895   TST fixed doctests, added more tests. Still 100% line coverage :)
       new  93e4390   ENH removed ``remove_zeros`` parameter.
       new  eaf8757   Merge branch 'larsmans_pr' into one_hot_encoder
       new  bd514aa   Merge branch 'master' into gbrt-interactions
       new  a2edf49   raise error if both grid and X are specified
       new  7c31618   initialize estimators_ with empty array not None
       new  ac2e0c7   DOC copyedit tutorials
       new  207793f   COSMIT rm outdated file of changes to liblinear
       new  53d9418   more input validation for partial dependence and doctest
       new  191f9c3   tests for partial_dependence
       new  bf5c124   rename learn_rate -> learning_rate
       new  21d1260   input validation for grid
       new  9509959   test cases for grid
       new  0530c10   pep8
       new  6f08ae6   added test for cartesian
       new  886ccfa   add partial dependence to whats new
       new  92138a4   COSMIT pep8
       new  3f32833   DOC corrected whatsnew.rst. Thanks @ogrisel.
       new  af101cc   ENH check in all classifiers in fit and predict that input is finite. inspired by #1027.
       new  a351355   ENH add checks for clustering, regressors, transformers
       new  d71ba5f   FIX revert old behavior, all tests work :)
       new  261f07e   MISC address Gael's comments
       new  2f0dd13   DOC added comment about default for n_nonzero_coefs.
       new  0c27619   documentation for partial dependence plots
       new  70a4bf1   add module imports
       new  d859eb0   typo
       new  7cec21b   TEST: decrease precision in test_lars_drop_for_good
       new  bc2948f   cosmit
       new  830b405   call pl.show
       new  4249edc   renamed datasets.cal_housing to datasets.california_housing
       new  c9cd2b7   FIX: lars drop for good platform specific test failure
       new  12e87d9   FIX + ENH: catch custom function argument errors and inform user
       new  c97e281   FIX #1354: machine precision assertion failure in test_liblinear_random_state
       new  7cf0cba   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  2ffbf28   add plot titles
       new  e9583f6   cosmit
       new  f764581   Merge branch 'master' into gbrt-interactions
       new  63bae39   cosmit: docstrings
       new  fc41a88   better narrative docs for partial dependence
       new  cf77c10   cosmit: footnote header
       new  52ba220   empty instead of zeros
       new  d7316bf   fixed the __repr__ method of cross_validation.Bootstrap, which failed if self.random_state is None
       new  8ace61f   Updated fixes from larsmans
       new  13a3176   BUG: imports should be locals
       new  69e3554   MISC: wording of doc/comments in example
       new  859bbce   Merge pull request #1335 from robertlayton/clustdocs
       new  824c07b   DOC typo in k-means docs
       new  a5c11b6   Merge branch 'master' into gbrt-interactions
       new  ef9161b   more explicit typing (int32, float64)
       new  a4e4523   COSMIT pep8
       new  863816d   ENH added check for non-negative input.
       new  c63b7a1   P3K: Fixed imports.
       new  6d0ffb1   P3K: Cannot compare list to tuple.
       new  ca60c93   Replaced use of deprecated method.
       new  9496a96   P3K: Changed StringIO to BytesIO to fix a failing test.
       new  67e1770   Merge pull request #1279 from amueller/one_hot_encoder
       new  253a2c4   FIX don't use pl.subplots.
       new  541f01d   ENH adding "apply" to random forests
       new  973edd5   ENH add RandomHashingForest estimator.
       new  6e3472c   ENH added docs, example and tests.
       new  de31f19   DOC Some narrative documentation for Random Forest Hashing.
       new  b669568   FIX for sparse matrix in RandomForestHasher
       new  131ad58   ENH refactor inheritance structure.
       new  b9a08bd   ENH use random regression task to avoid memory overhead of n_sample classes. Modifying cython for tree building would still be better... (via new criterion?)
       new  ae8a7a5   ENH Added Example
       new  d02d08c   DOC added references
       new  ef4d208   MISC renamed RandomForestHasher to RandomForestEmbedding
       new  e9e60db   MISC don't use pl.subplots, fix typo
       new  8e58f83   MISC rename plot_random_forest_hasher to plot_random_forest_embedding
       new  9a7de0b   ENH fix plot in docs. thanks @ppret.
       new  5ee93ec   DOC forgotten rename
       new  277e944   DOC fixed links in whatsnew.
       new  c201281   DOC added dump_svmlight_file to the references
       new  e3fc34a   DOC improve MinMaxScaler narrative docs.
       new  a345761   Merge pull request #1361 from astaric/py3k
       new  c4e978b   metrics.py: bugfix in precision_recall_curve and added tests
       new  d70e110   metrics.py: more detailed comment in precision_recall_curve
       new  090bed6   metrics.py: pep8
       new  e31376a   metrics.py: COSMIT more commets on precision_recall_curve
       new  5c61755   metrics.py: COSMIT, replaced cryptic np.r_ with np.hstack
       new  6b096cf   DOC added new precision_recall_curve to whatsnew
       new  593f4bd   ENH: RandomForestEmbedding in lle_digits example
       new  46d2461   DOC: cross-ref Random Forest embedding and manifold
       new  a9b4a38   Fix typos and wording in RandomForestEmbedding.
       new  e481596   Indicate dimensionality in RandomForestEmbedding example.
       new  f92d63a   DOC fix some layout on the "presentations" page, add Jake's resent PyData NYC tutorial.
       new  23ac37a   API : move isotonic regression out of linear_model
       new  dcbddc7   DOC : fix move of isotonic in doc + examples
       new  b1ea48d   DOC: add google analytics theme option
       new  7d707c4   Fix broken links
       new  ae86c22   BUG: Fix path in doc cleaning
       new  1f6d7e0   Merge branch 'master' of https://github.com/jaquesgrobler/scikit-learn into fix_doc_clean
       new  193b7dd   Merge pull request #1367 from fannix/master
       new  45bc0bc   Merge pull request #1369 from AlexandreAbraham/fix_doc_clean
       new  2468cbf   DOC Change URLs of NNDSVD papers to avoid paywall
       new  b0ac8f9   doc fix - trailing underscore and init param update
       new  2287edc   goodness of fit fix
       new  15db3c2   trailing whitespace sentence
       new  9432b4e   Merge pull request #1366 from agramfort/move_isotonic
       new  56afa38   DOC grammar in isotonic regression narrative docs
       new  6ab3b5b   MISC rename RandomForestEmbedding to RandomTreesEmbedding
       new  1644a7c   Merge branch 'master' into gbrt-interactions
       new  bb1f1ea   DOC: make MinMaxScaler example snippet readable outside of other sections context
       new  daeeb99   DOC: more improvements / fixes on the MinMaxScaler doc
       new  2b8ab35   ENH feature hashing transformer
       new  2ffe980   DOC narrative documentation for feature hashing
       new  f5a4ad2   ENH speed up hashing and reduce memory usage by 1/3
       new  cbb9b01   ENH allow (feature, value) pairs in FeatureHasher
       new  c8e644a   ENH 20newsgroups example for FeatureHasher
       new  205a947   ENH + DOC FeatureHasher
       new  3de0f43   ENH add dict support to FeatureHasher and make it the default input_type
       new  c370f53   Merge pull request #1374 from jakevdp/doc_GA_flag
       new  c89d208   Merge pull request #909 from larsmans/hashing-trick
       new  0f24720   P3K: Fix build for py3k + pip.
       new  99be49e   DOC: list of dicts in GridSearchCV
       new  87daeea   DOC: wording and layout on front page
       new  8c99284   Cosmit: use less memory in feature hasher tests.
       new  54b4ba2   Cosmit: make KernelCenterer a private attribute in KernelPCA.
       new  a88ee54   Improve KernelCenterer docstring.
       new  2fc7080   Add add_dummy_feature.
       new  f93ebb4   BUG enforce and document max. n_features for FeatureHasher
       new  94f1218   Add RandomClassifier and tests.
       new  76253a2   Fix tests.
       new  3fb6425   Add docstrings for RandomClassifier.
       new  7358fd4   PEP8.
       new  637a491   random_state=None by default.
       new  146da09   Remove label encoder.
       new  ac90b7b   Implement predict_proba.
       new  30aa21c   Add some narrative doc.
       new  d756c33   Address @amueller's comments.
       new  607c136   Rename to dummy.DummyClassifier.
       new  cb8989a   Add DummyRegressor.
       new  09cdd40   Add dummy estimators to references.
       new  86aad25   Add what's new entry.
       new  aab8162   Add comments.
       new  38b0f59   Check returned types.
       new  a7ebd1b   Merge branch 'master' into gbrt-interactions
       new  8a171a8   add plotting convenience function
       new  4f70ab3   uses plotting convenience function
       new  b2dcfdd   ENH: update joblib to 0.7.0a
       new  6ead032   BUG: fix properties on joblib files
       new  926858e   BUG: Add forgotten file
       new  324f69b   moved partial dependence into its own module.
       new  60a8316   doctest fix + cosmit
       new  7137713   fix imports
       new  af3345c   remove partial dependence (moved to own module)
       new  6cb2b83   updated example
       new  3b90033   fix imports (partial dependence)
       new  3f6f333   fix: california_housing not cal_housing
       new  fa0622c   cosmit
       new  e960643   switch axis for 2-way plot; better to compare with above plot
       new  e0b7a22   BUG: update joblib to 0.7.0b
       new  7cd4392   added partial dependence and fetch_california_housing to classes
       new  0b5fd93   better documentation
       new  0163d94   fix links
       new  9bbc433   add partial dependence module
       new  9adeaa1   Test expectations.
       new  e157836   Test string labels.
       new  55f8ab0   Test exceptions.
       new  99f2375   Cosmit: save one line.
       new  ef50f3e   Address @amueller doc comments.
       new  6530f44   Skip common tests for Dummy*.
       new  8430d66   Typo :/
       new  cfd8b79   Add example in docstring.
       new  b4022ac   Add to references.
       new  77cd045   Merge pull request #1382 from mblondel/add_intercept
       new  0e4a693   Merge pull request #1373 from mblondel/random_clf
       new  55bbbc1   Remove unused import.
       new  c01e65e   COSMIT don't do deprecated things in test (hmm)
       new  dd04394   BUG: fix murmurhash compilation with recent Cython
       new  41a03a5   Merge https://github.com/scikit-learn/scikit-learn
       new  c113c85   Travis CI cfg + status in README + sklearn requirements
       new  e3df73a   with Ubuntu's scipy instead of pip's
       new  591b7c2   with python-nose
       new  3076320   removed requirements.txt from travis cfg
       new  6c8887b   removed requirements.txt
       new  7f43561   DOC update Ubuntu installation instructions
       new  c40c0ff   changed the build image URL in README for after pull-merge
       new  e256540   Merge pull request #1397 from SnippyHolloW/travis
       new  97445e2   Improve error message when vocabulary is empty.
       new  116b54a   Fix bug in sqnorm (used by PassiveAggressive).
       new  f49773d   Link to travis.
       new  3c46eba   COSMIT pep8, removing unused imports and recommend ``toarray`` instead of ``todense``
       new  343f895   trying travis cfg with system-site-packages
       new  31de016   Improved bench_covtype.py to load data faster and support configurable n_jobs
       new  a8df0d1   Specify branch in status button.
       new  33899d0   ENH make sparse svm test more robust, catch warning on deprecated class
       new  4c2d1c9   ENH use blobs instead of iris in the common classifier tests. Iris has duplicat datapoints which raises annoying neighbors warnings.
       new  c1eb792   Fix staged_predict_proba() in gradient_boosting.
       new  fca4802   ENH slight cleanup in common tests, less warnings.
       new  295feec   add test for staged_predict_proba
       new  e9a893d   Merge branch 'master' into pr/1409
       new  66df362   Merge branch 'master' into gbrt-interactions
       new  963c064   better formatting of xticks (prevent overlap) pass arguments to contour and line plots
       new  f47ff85   show how to use ``partial_dependence`` to generate custom plots.
       new  857c642   doctest skip for plot function
       new  2dc982f   fix doctests skip
       new  780ab02   ENH Check what ``__init__`` does in test_common
       new  021c999   FIX messed up memorizing gmms parameter in GMMHMM before.
       new  c3ea29e   DOC added comment to test.
       new  3fa378a   DOC explain what the test is doing.
       new  bf0dcf3   BernoulliNB: Fix the denominator of P(feature)
       new  e0da673   FIX smoothing in Naive Bayes and refactor the discrete estimators
       new  4a34f7f   Merge pull request #1372 from jaquesgrobler/doc-fix-dev_guide
       new  98b188f   ENH add chi2 and exponentiated chi2 kernel.
       new  c5935b3   FIX add generated c file
       new  cf6c8b2   DOC add chi2_kernel and exponential_chi2_kernel references.
       new  270dc05   TST added a test for chi2 and exponential chi2 kernel.
       new  e02a668   FIX input validation, test chi2 in pairwise function, add reference.
       new  b9f999c   ENH fused types for chi2_kernel
       new  54b1228   ENH renamed chi2 to additive_chi2 and exponential_chi2 to chi2, as usually the exponential version is meant with "chi2"
       new  3d4fac8   DOC updated whatsnew
       new  b4bdd57   DOC cleared up difference to AdditiveChi2Sampler, added some "see also"s
       new  c03091c   DOC added stuff about chi2 kernel to narrative docs
       new  d681f58   FIX typo bug, more tests. Still more tests coming right up!
       new  c6e5263   DOC added "precomputed" variant to docs.
       new  07a5e36   TST 100% line coverage
       new  6b1d20e   ENH explicit check for zero denominator
       new  3e43ce3   ENH address @ogrisel's comments.
       new  ecc50d0   ENH addressed @kuantkid's comments. Also add myself to pairwise.py authors.
       new  1d318da   Merge https://github.com/scikit-learn/scikit-learn into travis
       new  ed6111d   FIX import assert_greater from testing module
       new  68c7b6a   COSMIT no diff for pairwise_fast.c
       new  bd536ba   DOC credit @sjackman in what's new for BernoulliNB fix
       new  fd82afa   renamed: ncols -> n_cols;          partial_dependence_plots -> plot_partial_dependence
       new  0d6d67e   ADD install guide for archlinux
       new  7f73ce6   test decorator to skip tests if matplotlib cannot be imported
       new  236465b   smoke test for plot_partial_dependence
       new  09d9e2c   fix: doc rename partial_dependence_plots -> plot_partial_dependence
       new  584e74d   Merge pull request #1415 from SnippyHolloW/travis
       new  c9444eb   Merge pull request #1418 from kuantkid/archlinux
       new  9a79b02   ADD spectra_embedding for wrap function spectra_embeeding as an estimator from spectral clustering
       new  c797e01   ENH finish sketch for the estimator wrapper
       new  f4d03a0   ENH add warning for inverse transform
       new  58572bb   ADD test cases for spectra_embedding
       new  b7c9998   ADD empty test scripts
       new  fc91451   COSMIT
       new  d2485c4   FIX typos
       new  fa8df33   FIX inconsistent typos
       new  8cd804a   FIX nearest_neighbor graph build
       new  b54a3ad   ADD add test_examples for pipelined spectral clustering and callable affinity
       new  7b98cdd   FIX remote does not have test file wired...
       new  ec68d8f   MOV move spectra_embedding from decomposition to manifold
       new  707edfe   ENH docs partially updated, happy mooncake festival
       new  3a4cebf   ENH move spectral_embedding as standalone, fixes for tests
       new  ffed733   COSMIT
       new  a47992d   ADD add the laplacian eigenmap to examples
       new  3b3f0ec   ADD test cases for two components, unknown eigenvectors, unknown affinity
       new  441a739   COSMIT
       new  6c50606   ENH test-coverage
       new  32970e7   PEP8 test files
       new  7faaf08   ADD spectra_embedding for wrap function spectra_embeeding as an estimator from spectral clustering
       new  936cfb5   rebase: fixing conflict
       new  e935121   ENH add warning for inverse transform
       new  d0ed921   ADD test cases for spectra_embedding
       new  ab99238   ADD empty test scripts
       new  5e6430f   COSMIT
       new  fe82743   FIX typos
       new  a92565f   FIX inconsistent typos
       new  cd02b8e   FIX nearest_neighbor graph build
       new  1494345   ADD add test_examples for pipelined spectral clustering and callable affinity
       new  d806cc1   FIX remote does not have test file wired...
       new  9fe7369   rebase: fixing conflict
       new  b7c4385   ENH docs partially updated, happy mooncake festival
       new  5572493   ENH move spectral_embedding as standalone, fixes for tests
       new  adbd0bd   COSMIT
       new  ec601c6   ADD add the laplacian eigenmap to examples
       new  243fc75   ADD test cases for two components, unknown eigenvectors, unknown affinity
       new  e873ed5   COSMIT
       new  f53f855   ENH test-coverage
       new  a04a970   PEP8 test files
       new  3f51b59   SYNC doc built error on one machine, sync with another
       new  fa77ce8   DOC docs for spectral embedding
       new  9d03e78   DOC dox fix and misc post-rebase things
       new  daae623   MRG merge with @Gael's PR 1221 and some name changes
       new  45c2085   FIX lobpcg, amg drops the constant eigen vectors by default
       new  e8d8de9   ADD check for symmetric and check for connectivity
       new  34b3967   ADD add test for check_connectivity
       new  cfff55e   COSMIT
       new  9f8d8d59  Change sparse graph to use cs_graph funcs. minor doc changes
       new  0c20549   Minor doc changes
       new  7f5b642   FIX spectral embedding offers choice whether to drop the first eigenvector
       new  75517e2   COSMIT
       new  0ef3aad   RENAME parameter rename in examples
       new  46f1587   RENAME rename eigen_tol and eigen_solver, and warning about using old variable name eig_tol and mode
       new  0e41b14   ADD add a test for discretize function
       new  615494e   COSMIT and Typo
       new  87978d0   FIX backwards support
       new  dbf39a6   FIX doc fix and test fix
       new  d01f3be   COSMIT
       new  9b47cdf   ADD added examples, and eliminate unnecessary imports
       new  d272464   FIX nn-affinity does not support sparse input
       new  97a216d   COSMIT and minor fixes
       new  67fe92b   DOC update whatsnew
       new  8eb74e0   fix: addressing gael's comments
       new  ba23493   fix: set syntax
       new  ffd8f3d   fix: increase robustness of label binarizer test
       new  92ba802   sty: white space
       new  89e1969   fix: change affinity check
       new  6312a9f   doc: clean up style and grammar
       new  e0959c9   ref: change name to indicate semantics
       new  447b9d5   fix: removed unused keyword precomputed and clean up if clauses
       new  bda006d   fix: moved random state check to fit
       new  2978ad9   doc: removed merge diff markers
       new  44253bc   doc: align hyphens
       new  ec4cbef   Merge pull request #1408 from satra/fix/rebase1396
       new  9b6f848   clarify documentation for radius_neighbors
       new  a1bf446   FIX csr conversion in amg code in spectral embedding
       new  fad7143   COSMIT refactor input validation code; skip some issparse calls
       new  a5a01b2   plot fix
       new  e0bf76d   variable name change
       new  fdf4846   ENH add random-seed args
       new  cb810f2   Merge pull request #1425 from arjoly/enh_bench_covertype
       new  df9fc91   Add missing assertion.
       new  236c551   Added strided patch extractor to feature_extraction/image. Extracts patches 16x faster on the MiniBatchDictionaryLearning example
       new  6394b35   Now added extract_patches for random extraction as well
       new  e19d70f   Now replaced max_patches part by fancy indexing
       new  b5f347e   removed stuff i commented out
       new  1f7d3e4   testing for correct output shapes and patch content of the last patch for 1 to 3 dimensional arrays
       new  68a788f   Changes in documentation and notation
       new  8078d70   TST : use assert_true and not assert in test
       new  ac48bbf   Merge pull request #1424 from jaquesgrobler/plot_omg_fix
       new  aff7da7   Merge remote-tracking branch 'upstream/master' into gbrt-interactions
       new  ad4ac3e   better input checking (e.g. for str features) more tests
       new  c9315d5   better handling of multi-class case (w/ symbol labels) more tests for multi-class
       new  f461456   code snippets for narative doc and restructuring
       new  86f590e   fix: random_state got initialized in fit_stage; caused same feature subsample in each tree
       new  7ff28c1   add test for gbrt random_state regression
       new  821f7a0   Merge branch 'master' into gbrt-random-state-fix
       new  f414ca1   Merge branch 'master' into gbrt-interactions
       new  b66a4d6   Add fit_transform to FeatureUnion
       new  a492302   Change / to (…) line continuation
       new  691804a   doctest skip: matplotlib not available on travis
       new  2d0525b   Add test case for FeatureUnion.fit_transform
       new  cb3129a   fix: doctest in ensemble.rst
       new  58070a2   Fallback to fit followed by transform if fit_transform is unavailable
       new  d454c13   FIX #1417: move nosetests configuration parameter to setup.cfg
       new  1e595d9   Remove doctest-options from setup.cfg as not supported in old version of nose
       new  3504e13   Merge branch 'master' into gbrt-interactions
       new  dd0deae   rephrased the one-way PDP description
       new  16cb053   Add test case for fit_transform fallback
       new  d3ce1ca   Merge pull request #1428 from tnunes/feature_union_fit_transform
       new  2b8bef4   Fix pipeline fails if final estimator doesn't implement fit_transform
       new  eb64f18   BUG: Don't test test_k_means_plus_plus_init_2_jobs on Mac OSX >= 10.7 because it's broken. See #636. Closes #1407.
       new  e85dd68   Merge pull request #1430 from erg/issue-1407
       new  82a281a   Merge pull request #1429 from tnunes/fix_pipeline_fit_transform
       new  5773168   implemented predict_proba for OneVsRestClassifier
       new  726b42b   forgot an except clause
       new  742072b   removed unnecessary repeat
       new  3340bf0   corrected doc for predic_proba, also caught few errors.
       new  fd930e3   wrote test_ovr_predic_proba method
       new  c4cbffc   divided test for predict_proba into two functions
       new  18a0a2e   removed check for predict_proba method.
       new  396a16f   [pep 257](http://www.python.org/dev/peps/pep-0257/) and and other doc improvements.
       new  48f0a44   corrected bad test in test_multiclass
       new  166f32d   Flake8 Corrections made
       new  fa0fd40   spell checked
       new  482e576   Spelling is checked, passes Flake8 without errors.
       new  9bec60a   added backtick around self.classes_ in multiclass.py
       new  15e40c9   new example added for manifold learning
       new  8b9c248   Andy`s suggestions
       new  db28229   links for MDS
       new  4fb669e   small changes
       new  9a15d93   final changes
       new  80c7451   pep8
       new  f09a579   heading change
       new  d5ff04a   FIX: amg requires sparse matrices input
       new  cc8bede   missing _set_diag
       new  4161f56   ENH cleanup tests, lower tolerance
       new  9603b07   COSMIT pep8
       new  b7b91ef   Merge branch 'master' into gbrt-interactions
       new  4806bc4   Update what's new.
       new  3be27ff   Cosmits and typos.
       new  3848f09   topics -> topic
       new  0ba8617   Merge remote-tracking branch 'upstream/master'
       new  bf752dd   use Agg backend with warn=False for matplotlib enabled tests
       new  755164f   check in ``if_matplotlib`` if $DISPLAY set
       new  3de4442   FIX and test deprecated import of spectral_embedding from cluster
       new  2c984ee   TST better test-coverage in clustering module
       new  d173ab1   COSMIT in cross-validation tests
       new  00f4fb2   FIX random state in test by @briancheung. Thanks
       new  a375cbf   TST better coverage in dict learning and cross validation
       new  a3af919   TST better coverage in preprocessing module
       new  43e3b7f   DOC add matplotlib version requirement, rephrase
       new  cca3a35   fix spectral related testing errors
       new  663539e   COSMIT and unused lines
       new  a257181   FIX further improve the thresholds
       new  ee21736   Merge pull request #1440 from amueller/matplotlib_requirement
       new  a35bfd5   use subplots_adjust instead of tight_layout
       new  8897c6d   use 100 instead of 800 n_estimators; looks the same but faster; ESLII uses 800
       new  66cce69   FIX discretization test have shape problem, use coo_matrix instead of LabelBinarizer
       new  138bf89   ZipFile context manager is only available in Python >= 2.7
       new  7024a5f   Addressing @ogrisel's comments
       new  db98995   ENH: Simplify the shape of (n_)classes_ for single output trees
       new  829afed   COSMIT Mean Shift docs.
       new  0799ee9   ENH: Simplify the shape of (n_)classes in forest
       new  4b7ad61   PEP8
       new  f39354d   TEST: regression test for shape of (n_)classes
       new  b31696c   TEST: enforce flat classes_
       new  d278241   What's new: API changes
       new  c5ed18c   ENH: better names for variables
       new  74c17c6   Merge pull request #1441 from kuantkid/fix_spectral_test
       new  331155b   What's new: added :class: keyword
       new  de3b09a   FIX: convert predictions into a numpy array
       new  caf69c0   FIX: docstring tests
       new  3708afa   Merge pull request #1445 from glouppe/tree-shape
       new  f6bd2e9   What's new: typo
       new  f118cba   Display the test names to understand which test is triggering the segfault on jenkins
       new  54fb2de   remove reference to removed API, fixes #6
       new  1b06789   COSMIT some fixes in whatsnew rst
       new  c64a6b3   FIX manifold example - sorry, my bad.
       new  0bdf3d9   COSMIT RST in manifold sphere example.
       new  61e2a71   ENH fix random seed in manifold example
       new  25b60c1   ENH Nystroem kernel approxmation
       new  d4eb819   ENH renamed class NystromKernelApproximation to Nystrom (it is in the kernel_approximation module). Also improvements to example docstring
       new  3636358   DOC docstrings for Nystroem.
       new  9e6a904   ENH cosmit, gamma defaults do None, not 0. address some of @mblondel's comments.
       new  f3e8fb8   ENH tests for Nystrom, check that n_components is smaller than n_samples.
       new  5bd197c   DOC added note in example that digits data is to small.
       new  22883c4   DOC narrative doc for Nystroem.
       new  f5a47bd   Merge pull request #9 from benjaminwilson/master
       new  0ed406b   DOC updated whatsnew with Nystroem.
       new  1a83c16   FIX: Do not rely on strides for contiguous arrays
       new  e2ed754   ENH: use broadcasting, not tile
       new  210767d   ENH don't import * in utils __init__.py
       new  2f35fa0   cosmit: remove fourth quote
       new  357f3c1   Call DecisionTreeRegressor instead of Tree
       new  ce7ea53   COSMIT Remove duplicated assignement
       new  6f6d692   Use the check_input argument
       new  14a3f74   DOC : add description of check_input args
       new  8c418bc   DOC explain parameter estimators_
       new  77064a2   DOC explain parameter estimators_ (2)
       new  e8d21a7   ENH Move parameter checking to fit
       new  44ba8ed   COSMIT
       new  ba1d4f5   FIX casting bug
       new  af50266   ENH preserver contiguous property
       new  2c71362   COSMIT
       new  ce735e8   DOC describe reasons for reshape
       new  f4f34d4   PEP8
       new  5d0f417   FIX: perform transition from tree to DecisionTreeRegressor
       new  6c60fe2   FIX feature importance computation + Enable smoke test of feature importances
       new  210f62e   Update whats new
       new  9c94b37   ENH add author
       new  d436563   Merge pull request #1388 from arjoly/issue1047_gradient_boosting_uses_decision_trees
       new  6c82a70   TST better coverage for GridSearchCV, test unsupervised algorithm.
       new  67fccac   Merge pull request #1458 from seberg/contig_strides
       new  7242908   What's new: fix by @seberg
       new  1eecf4c   FIX roc_curve failed when one class is available
       new  80fc25b   COSMIT
       new  ba54ff4   DOC fix
       new  d657fb0   TYPO fixes
       new  90c72a2   DOC address @amueller's comment
       new  8dcc54c   FIX typo
       new  65f522b   Update whatsnew
       new  489aab4   BUG: Fix the random_state on make_blobs() in test_classifiers_classes(). Fixes #1462.
       new  4bb41cb   BUG: Make a RandomState object and use it in test_transformers(). Fixes #1368.
       new  73a1229   TST better test-coverage for image patch extraction.
       new  f52e811   TST better coverage in kernel_approximation
       new  0e77bce   FIX spectral_embedding test erros, ADD spectral embedding to sphere examples
       new  7064b3e   FIX: fixed random_state for heisen doctest failure in multiclass module
       new  2b56ee1   set min_density when growing deep trees during gradient boosting
       new  08dff76   ENH input validation only in ``fit`` in LassoLarsIC, check that error is raised.
       new  8eda5e1   TST document and test verbosity parameter of lars_path
       new  270bc8e   TST some more tests for SGDClassifier input validation
       new  17e8f01   ENH / TST better coverage of supervised clustering metrics, slight cleanup
       new  7cb50d1   DOC make unit test requirements a bit stricter. 80% is sub-par with current code-base
       new  e0c3aa4   COSMIT pep8
       new  22b52eb   COSMIT renaming chunk_size to batch_size in MiniBatchDictionaryLearning and MiniBatchSparsePCA
       new  b9c0c55   DOC add rename to whatsnew
       new  52cc50a   Merge pull request #1468 from erg/random-failures-12345
       new  c193eac   Delete iris.dot in tree.rst doctest
       new  b07608e   FIX: seed blobs dataset to have a stable spectral clustering under OSX 10.8
       new  e12c673   cosmit pep8
       new  c79862a   Merge pull request #1478 from amueller/pep8
       new  d9138f4   Merge pull request #1451 from amueller/chunksize_batchsize_rename
       new  e4980bc   Merge pull request #1470 from kuantkid/fix_spectral_cluster_test
       new  f986d74   Add comment in test_spectral_clustering_sparse
       new  8437993   changed n_folds > min_labels error to warning
       new  3fe598a   removed tests for the old error.
       new  c8ae9a5   added test for warning. Added warning category
       new  2b80d00   removed a carriage return in warning message
       new  02fe944   added space between # and text
       new  7177172   FIX GridSearchCV on lists that I broke in 8b3e4d06c05ac82130176161404f0434b74fe2c7
       new  19cc55d   ENH added test, started on cross_val_score
       new  aac12fa   ENH allow lists in check_arrays
       new  e1972fa   ENH make cross_val_score work, some refactoring in GridSearchCV
       new  9fd6b52   ENH consistency: stuff is not an array if it doesn't have ``shape``.
       new  7dae6bf   TST GridSearchCV raises ValueError when precomputed kernels are not matrices.
       new  ef60f35   ENH Simplify estimator type checking in GridSearchCV.
       new  909e700   Merge pull request #1465 from AWinterman/issue-1017
       new  e9d69d8   first pass at implementing sparse random projections
       new  2bf3468   DOC: better docstrings
       new  8fa242f   DOC: more docstring improvements
       new  5f4bc0c   Remove non-ASCII char from docstring
       new  110842a   use random projections in the digits manifold example
       new  fc1e9a6   test embedding quality and bad inputs (100% line coverage)
       new  1180cb2   typos
       new  48323fe   one more typo
       new  ea43920   OPTIM: CPU and memory optim by using a binomial and reservoir sampling instead of direct uniform sampling in the n_features space
       new  16fa4ef   note for later possible optims
       new  a617cd5   fix borked doctests
       new  8eb8554   make it possible to use random projection on the 20 newsgroups classification example
       new  60a86cc   FIX: raise ValueError when n_components is too large
       new  9701ce7   remove the random projection option from the 20 newsgroups example
       new  c4d5d65   leave self.density to 'auto' to implement the curified estimator pattern
       new  f0ece82   more curified estimator API
       new  0ab2d64   useless import
       new  f2d738d   change API to enforce dense_output representation by default
       new  68f6904   ENH: vectorize the johnson_lindenstrauss_bound function
       new  4ca8e34   started work on plotting the JL bounds to be used in the narrative documentation
       new  b0c1b6c   More vectorization of the johnson_lindenstraus_bound function
       new  8840fb9   More work on the JL example to plot the distribution of the distortion
       new  f13cbf4   WIP: tweaking JL function names
       new  62b9d2b   check JL bound domain
       new  6649d5d   JL Example improvements
       new  2feb067   WIP: starting implementation implicit random matrix dot product
       new  82cc48f   working on implicit random projections using a hashing function
       new  a850892   OPTIM: call murmurhash once + update test & example
       new  2b329b9   first stab at CSR input for hashing dot projections
       new  ba5ea28   implemented dense_output=False for hashing_dot
       new  8e742d5   refactored test to check that both materialized and implicit RP behave the same
       new  cbb2e39   fixed broken seeding of the hashing_dot function
       new  08cb357   leave dense_output=False by default
       new  1e528c4   use the 20 newsgroups as example dataset instead
       new  3621693   make it possible to use a preallocated output array for hashing_dot
       new  510b4a0   missing docstring and s/hashing_dot/random_dot/g
       new  edafb48   COSMIT use sklearn.utils.testing
       new  35c75bf   ENH Let the user decide the number of random projections
       new  add00fa   Clean random_dot features
       new  a069d97   Clean random_dot features (2)
       new  961c27a   Clean random_dot features (3)
       new  2ac9b61   Clean random_dot features (3)
       new  a5ebd1d   ENH let the user decide density between 0 and 1
       new  84bc401   COSMIT
       new  5aef6b2   ENH Strenghtens the input checking
       new  956f3d8   ENH Add gaussian projeciton + refactor sparse random matrix to reuse code
       new  2e29be3   ENH add more tests with wrong input
       new  10aade4   ENH add warning when user ask n_components > n_features
       new  3b96b91   DOC: correct doc
       new  0bb6207   ENH add more tests
       new  7219323   Update doctests
       new  f77185e   ENH cosmit naming consistency
       new  825303c   FIX renaming bug
       new  9a2d76a   COSMIT
       new  743f416   WIP: add benchmark for random_projection module
       new  eccbd90   ENH finish benchmark
       new  75728cc   Typo
       new  f014ed5   ENH optim sparse bernouilli matrix
       new  6cc0997   FIX example import (name changed)
       new  e3dcdd4   FIX: argumetn passing selection of sparse/dense matrix
       new  e69281f   ENH assert_raise_message check for substring existence instead of equality
       new  0d20c6c   ENH add two test to check proper transformation matrix
       new  0843b66   PEP 8 + PEP257
       new  23bdd88   DOC improve dev doc on reservoir sampling
       new  4ea5fb8   COSMIT + ENH better handle dense bernouilli random matrix
       new  2753422   FIX: make test_commons succed with random_projection
       new  d5ed2ff   DOC removed unrelevant paragraph(s)
       new  3fd684b   ENH add implementation choice for sample_int
       new  06da921   ENH add various sampling without replacement algorithm
       new  eff360b   Typo
       new  3a5abab   TST: Add tests for every sampling algorithm + DOC: improved doc
       new  260bdfa   DOC: fix mistake in the doc + ADD benchmarking script
       new  3676a3e   ENH Rename sample_int to sample_without_replacement
       new  63e9510   DOC + ENH: minor add in doc + set correct default
       new  f1776cf   FIX: broken import
       new  474663f   FIX typo mistakes + ENH change default behavior to speed the bench with Gaussian random projection
       new  4729024   ENH Add allclose to sklearn.testing
       new  349af09   ENH improve naming consistency
       new  057b8d9   PEP8
       new  42a4a95   COSMIT
       new  44a7e01   DOC + typo
       new  a9b73d6   DOC set narrative doc for random projection
       new  77914ee   FIX: broken test due to typo correction
       new  32be546   DOC minor improvements
       new  4702f6d   DOC mainly switch from .\n:: to ::
       new  eb67e01   FIX typo mistakes
       new  4cafda2   DOC improve name in example
       new  8deb970   DOC Separate the jl example from references
       new  15d0232   ENH Add jl lemma figure to random_projection.rst
       new  4734f6a   COSMIT (typo, doc, simplify code)
       new  dd4d885   pep8
       new  2b7fd73   Typo
       new  275aada   DOC typo in narrative doc
       new  d1a223b   DOC fix typo in filename
       new  ccea7f1   DOC clarification
       new  a6ed58d   ENH flatten random_projection module + add sklearn.utils.random
       new  980600b   ENH refactor matrix generation BaseRandomProjectiona and subclass
       new  7af5176   DOC improve layout (url)
       new  e511945   Make the JL / RP example use the digits dataset by default
       new  06a477a   FIX broken import
       new  a7ea4ee   pep257 + COSMIT: naming consistency
       new  50368c3   COSMIT
       new  a08b3c1   COSMIT
       new  114e3fa   Remove unused line
       new  a75a3dd   DOC improve doc for jl lemma function
       new  311faa5   typo
       new  12c8d3b   ENH Rename Bernoulli random projection to sparse random projection
       new  616e2b5   ENH Rename Bernoulli random projection to sparse random projection
       new  918d337   DOC add see also
       new  7732a5f   pep8
       new  0788164   COSMIT make everything use the common interface
       new  8931e67   DOC improve + fix mistakes + TST added
       new  db4422f   ENH Simplify assert_raise_message + TST add them
       new  9aa9a10   DOC add utitilies to the doc
       new  a0860b1   FIX don't use assert_allclose. It is not supported in numpy 1.3
       new  81a8146   eps=1.0 is no longer a valid value
       new  a8fec8a   Fix link to plot_lda_qda example.
       new  0eaa6e9   Merge pull request #1483 from aweinstein/fix_doc_example
       new  dbc4689   Typo / fix in JL lemma example
       new  754c2d9   COSMIT pep8
       new  54385c7   COSMIT featuers -> features typo
       new  966e10c   COSMIT PEP8
       new  983b45c   COSMIT pep8
       new  b2cf99f   DOC add version when setting parameters in fit will be removed to docstring
       new  ac2d29d   FIX typo / bug in test_common that ignored the first init parameter.
       new  72248f0   TST make test more stable.
       new  2d4624d   DOC + FIX density to Ping and al. recommandation
       new  8340489   ENH make jl lemma work even with non numpy array
       new  38868a6   DOC add default values
       new  889966d   Updates for DBSCAN clsutering docs
       new  5b822a2   Couple of small changes from comments
       new  cc63e1c   ENH slight improvement of common tests.
       new  0df8fb2   Add failing test and no-op flip
       new  d5857bb   Sign flipping as suggested by @ogrisel, not in place
       new  cf5d34e   Make sign flip in place
       new  9ca8f7e   Test more seeds for svd sign flipping
       new  0755f15   Add sign flip as flag in randomized_svd
       new  160648d   Make sure svd_flip test actually tests something
       new  4040d42   Make randomized_svd flipped by default
       new  5aed364   svd_flip test fails on Travis. Change random seed, see if it helps
       new  641d0a7   Cannot easily ensure non-uniqueness without the fix, just test uniqueness
       new  5b948bb   TEST flipped svd remains correct
       new  f349fcf   DOC slight cosmit in metrics docstrings.
       new  b91f5c0   FIX i should trust my past self a bit more
       new  71d00d9   ENH use an array instead of a dict in RFECV
       new  38b455d   Cosmit pep8
       new  07293bd   TST a little more coverage in unsupervised metrics.
       new  6782658   ENH clean up redundant code in pairwise
       new  42e266c   ENH more test coverage in pairwise distances
       new  805a20e   FIX more robust test for silhouette score
       new  ff3ea95   DOC classifier comparison: plot data without decision boundary first, better (imho) color scheme.
       new  0ac4243   DOC add Nystroem kernel approximation to the references
       new  5a6f7b5   FIX stupid mistake
       new  f0f1e06   FIX: MinMaxScaler on zero variance features
       new  c718bf9   Simpler inline comment
       new  e192ce4   Add one more test for MinMaxScaler on newly transformed data
       new  5fbb227   ENH: issue warning when minmax scaling integer data + test
       new  47305c0   FIX random_state in MDS was not initialized properly
       new  d483c08   COSMIT pep8
       new  083a06e   COSMIT Typo
       new  2430f41   Improving code coverage for datasets module. Moved dataset imports inside test_data_home, because it is preferable for import errors to only affect the tests that require those imported methods.  My first commit to scikit. -bcajes
       new  9364e39   revert to original import placement style
       new  337d776   ENH isotonic regression is now slighty more robust to noise
       new  a716b3c   Added test to check whether the isotonic regression changed y when all ranks were equal
       new  85c0b45   Merge pull request #1504 from NelleV/isotonic
       new  19e8e03   ENH uses the IsotonicRegression classifier instead of the method
       new  9301f14   FIX the mds example did not plot the NMDS
       new  3352bba   Type-os and added great links to learning more about Machine Learning
       new  9d5ed3a   Feedback from @amueller
       new  2ef136c   Add perceptron loss to plot.
       new  f17a517   ENH Add support for multioutput metrics
       new  635a93f   DOC add narrative doc for regression metrics
       new  4061af5   Update what's new
       new  2c22eb5   TST check that ValueError is raised when the number of output differ
       new  ecf7a75   ENH add mean absolute error
       new  9e15d2a   DOC cosmit alphabet order of classification metric in ref
       new  564194e   DOC typo
       new  98a0b47   ENH add multioutput support for dummy estimator
       new  1045780   DOC instance attributes + TST: do not record warning
       new  fd567ac   DOC typo
       new  f5e8c21   ENH preserve output ndim
       new  0d47916   ENH: add the squared hinge loss to the SGD loss example
       new  b44e779   threshold parameter was ignored in SquaredHinge loss.
       new  31cb829   FIX - nmds now uses the same scaling as previously
       new  7b53c9a   Merge pull request #1505 from NelleV/mds
       new  d3262f3   moved class_prior in NB to __init__
       new  c60862a   added deprecation warning to fit function
       new  799bef2   fixed docstring tests
       new  44c4096   fixed typos
       new  0905269   added warnings
       new  537e824   updated based on comments
       new  cc05add   fixed local variables
       new  dc25f23   renamed the new parameter to class_wieght
       new  11a389d   fixed docstring test
       new  795c20a   COSMIT update warning, pep8
       new  127c44b   ENH refactoring class weights for SVM and SGD
       new  e8ec79d   TST all classifiers now have "classes_". adjust test_common.
       new  dddad84   ENH remove class_weight_label from python side of SVM
       new  ff3bb3a   ENH remove class_weight_label from sparse svm
       new  af3ff35   TST move test of "classes_" to the appropriate test in "test_common".
       new  0dd3e13   FIX  remaining doctests
       new  843d795   DOC docstring for compute_class_weight
       new  7d11aaa   ENH remove class_weight_label from LibLinear python side.
       new  2881702   ENH removed unused old function
       new  6f9339a   TST fix import in test
       new  3be8a23   ENH addressed @ogrisel's comments.
       new  956975a   DOC changed docstring to be more clear.
       new  35b1d39   ENH documented changes for SVC classes_ changes.
       new  1c1538a   ENH move utility function into dedicated file, not __init__.py
       new  d0c32a5   TST start on testing consistent class weights
       new  f25412f   ENH nu-SVC doesn't support class_weights
       new  0fbf223   FIX liblinear class weight in binary case, robust testing.
       new  06f4077   cosmit whitespace
       new  ca9dcfd   DOC add comment in liblinear
       new  bad3613   TST  better test for class weights (that actually tests something)
       new  7dc838e   ENH test automatic setting of class weights in common test
       new  c75decd   TST skip RidgeClassifier in class weight test for the moment
       new  7edaaca   DOC added fix to whatsnew.
       new  2d5034d   FIX don't test auto in ridge classifier as it is not supported currently
       new  2d0ef35   FIX tests for auto class weights
       new  6902e07   DOC more concrete whatsnew
       new  30045e1   FIX skip tests for naive bayes for the moment.
       new  76c66a0   BUG Cholesky delete routines wouldn't compile on Solaris
       new  98909ce   Welcome to Wei Li and Arnaud Joly.
       new  86d417b   DOC made myself contact for authors, changed my website to blog.
       new  43a38f7   ADD : cosinus kernel calculation in metrics/pairwise.py
       new  edef908   add cos_kernel in help of decomposition/kernel_pca.py
       new  20909e5   name change: cos into cosine
       new  1fba2b8   change way of calculating cosine_kernel in metrics/pairwise.py
       new  703102f   add test for cosine_kernel in metrics/test_pairwise.py
       new  e96c0e5   correct indent pb and re-edit cosine_kernel help in metrics/pairwise.py
       new  21b939e   fix style issue by running pep8 on metrics/pairwise.py and on metrics/tests/test_pairwise.py
       new  0665e74   remove duplicated test_cosine_kernel() in metrics/tests/test_pairwise.py
       new  8d317e3   change test_cosine_kernel to include normalize from preprocessing.py in metrics/tests/test_pairwise.py
       new  51b3cef   remove duplicated dimension check in metrics/pairwise.py
       new  61742f4   add reference to cosine similarity in cosine_kernel help from metrics/pairwise.py
       new  e26a5a0   modify cosine_kernel func to use normalize from preprocessing.py and change the test_cosine_kernel adding scipy.sparse inputs respectively in metrics/pairwise.py andmetrics/test_pairwise.py
       new  b3eac46   modify test_cosine_kernel to compare result obtain with linear kernel in metrics/tests/test_pairwise.py
       new  0b04561   FIX: add prefix 'np.' to sqrt for test_cosine_kernel in metrics/tests/test_pairwise.py
       new  ad8dc1b   FIX: move import of normalize function into the cosine_kernel call in metrics/pairwise.py
       new  db17315   TST add cosine_kernel to kernel tests, pep8
       new  050b78c   ENH lazy import of metrics in base, not preprocessing in metrics.
       new  83d37ca   Clean up test_pairwise.py.
       new  1574e6d   ENH document attributes in QDA and LDA, rename to adhere to sklearn conventions.
       new  986fdfa   More clean up of test_pairwise.py.
       new  87810a2   Cosmit: break up long line.
       new  a9bdff7   Merge pull request #1517 from amueller/lda_qda_cleanup
       new  704befc   DOC fix shape of coef_ for LDA.
       new  30dfa01   Improving code coverage for datasets module. Moved dataset imports inside test_data_home, because it is preferable for import errors to only affect the tests that require those imported methods.  My first commit to scikit. -bcajes
       new  305b9ec   bring datasets.base to 97% coverage with a few more tests
       new  2e119f6   removing backup file
       new  c7fa346   checking data.shape for each test dataset
       new  38f784b   TST somewhat hacky fix for tests on image loading.
       new  03048dd   ENH more logical behavior, better docstring, tests
       new  de9bcc9   FIX do checks even if allow_lists
       new  7819e18   DOC try to be as clear as possible.
       new  15ded1f   ENH cleanup in check_pairwise_arrays, raise error on sparse input in chi2_kernel and manhattan_distance
       new  69210da   COSMIT doc formating
       new  3a52436   train_test_split: test_size default is None If test_size is None and train_size is not None, test_size is the complement of train_size. If test_size and train_size are None, test_size = 0.25. Docstring is updated. Existing tests didn't seem to cover much of existing behaviour and no updates were added. New behaviour was tested.
       new  c61e1aa   Modified docstrings
       new  5a7f0a8   Modified docstrings and tests
       new  b779b70   DOC updated whatsnew
       new  5f8a9b0   FIX: makes our libsvm port compile under MSVC
       new  e1934db   ENH added class_weight to Naive Bayes docs.
       new  eb9e161   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  24494d3   FIX random seed in FastICA testing.
       new  69eb712   DOC fix docstring of GMM
       new  c36a376   ENH Add "proximity" parameter to MDS.
       new  d519b2d   FIX soime typos, modify test.
       new  791a224   FIX another typo, fix examples
       new  1aacfbe   ENH updated to more examples.
       new  685c652   ENH rename proximity to dissimilarity
       new  eb5b586   COSMIT reorganized functions in the module
       new  60be687   DOC : add doctring in plot_lasso_and_elasticnet.py
       new  58b55ac   DOC: adding Bishop as ref for ARD
       new  4cdf9e4   Merge pull request #1530 from agramfort/doc_lasso
       new  c7c4d53   COSMIT simplify unique_labels in sklearn.metrics
       new  0acce47   Added code to address issue #1403
       new  788a8dd   X is not a constructor parameter.
       new  9095baa   Add missing types to docstring.
       new  4ebd382   In preprocessing.binarize, eliminate zeros from sparse matrices
       new  7c1d643   Added feature for issue #1527
       new  8e423c9   Minor PEP8 fixes for issue #1527
       new  824f6b4   Minor docstring fix for issue #1527
       new  ec2703d   Added tests and docs for normalized zero_one loss
       new  b84bc65   Fixed pep8 spacing issue and floating point doctest issue
       new  9fe2e93   Checkout files from ndawe:treeweights
       new  cbb8df6   FIX: roll back some changes
       new  6c4bb29   FIX: what's new
       new  38c56f2   flake8
       new  546d539   sampling w/ replacement via sample_weights
       new  8dcec88   ENH: early binding + allocate features at tree creation (by @pprett)
       new  bf418c9   FIX: oob test
       new  2494366   DOC: sample_weight=None
       new  25616ce   DOC: what's new
       new  b9df877   DOC: typo
       new  9593b83   DOC: cosmit
       new  d981f85   FIX: use sklearn.utils.fixes.bincount
       new  e949a42   ENH: use random_state.shuffle
       new  d8c9365   ENH: import aliases
       new  a314c9d   ENH: import aliases (2)
       new  68e25b1   ENH: import aliases (3)
       new  fa35dec   PEP8 (some)
       new  dda9c60   TEST: sample_weight
       new  72d767c   TEST: sample_weight (once more)
       new  9cdb787   ignore splits that yield nodes with net negative weight in find_best_split
       new  329c1cd   rm unneeded negative weight logic in Criterion.init_value and Gini.eval
       new  8d9bb14   add note about negative weight treatment in BaseDecisionTree.fit
       new  4bf2866   FIX: iris.target
       new  7945e9d   FIX: raise an exception if negative number of samples
       new  4a5a507   add negative weights test (currently fails): predict_proba should still be valid probabilities
       new  7f745b8   FIX: negative weight test. do not allow any class to have negative weight after a split
       new  f016712   TEST: use rng
       new  8fc565f   FIX: do not overwrite min_samples_split
       new  78fc091   FIX: set min_samples_split=2 by default
       new  64c671a   DOC: document negative weight treatment in the case of classification
       new  f8b8102   DOC: updated docstring
       new  f4ce71e   Typo
       new  b454d56   implement AdaBoost
       new  df4ecef   use weighted mean in ClassifierMixin.score
       new  7e97ca0   ENH: weighted r2 score for regression
       new  0c1ae93   COSMITs
       new  9ff1754   ENH: Added balance_weights
       new  a5e35d9   ENH: added some tests
       new  6b77b82   FIX: test_oob_score_regression
       new  4826fba   FIX: DecisionTreeRegressor.score
       new  f227f9c   FIX: compute weighted oob scores
       new  f6a8313   FIX: NaN problem + Added some tests
       new  5f8049c   TEST: added some more tests
       new  78c484d   EXAMPLE: simplify n_estimators and n_samples
       new  aaaf1cd   TEST: importances
       new  af8384c   TEST: multi-output problems
       new  f188ed6   ENH: WeightedClassifier/Regressor mixins
       new  b1fd7db   DOC
       new  491379d   FIX: import not used
       new  aa5d3d9   FIX: overlapping y-axis labels
       new  f1c9bab   FIX: use generator instead of np.random
       new  4c81a1e   rm doctest in make_gaussian_quantiles
       new  f3b0e65   fix variable naming in weight_boosting
       new  a5723e0   FIX: drop support for multi-output
       new  f3873ea   TEST: errors
       new  0f620c1   ENH: staged_score
       new  81aaede   EXAMPLE: reduce the number of samples
       new  562b60c   EXAMPLE: merge plot_adaboost_iris into plot_forest_iris
       new  59740fd   EXAMPLE: drop plot_adaboost_quantiles
       new  cf0d074   FIX: move balance_weights into preprocessing
       new  2604e97   PEP8 + PyFlakes
       new  bf54f1a   FIX: broken test
       new  91c9e00   FIX: one more bug
       new  678a401   FIX: remove prints
       new  89fe823   DOC: edited some docstrings
       new  b12be6b   DOC: added references into classes.rst
       new  b5d3e93   ENH: rename boost method to _boost
       new  e5ce8f9   DOC: cosmits + narrative documentation (begin)
       new  ac18e8f   DOC: proper citations
       new  e9dbfc4   DOC
       new  e1a76d4   FIX: TypeError for regressor
       new  12580a9   FIX minor comment
       new  15d3baf   FIX: docs, code clean up, learn_rate -> learning_rate
       new  d7524d9   FIX: plot_adaboost_classification.py
       new  43d4275   don't enforce DTYPE at the ensemble level
       new  cea3b90   DOCS: note generator behaviour in staged methods
       new  b628be2   TEST: make test_importances more stable
       new  5d29218   DOC: narrative documentation
       new  18830cf   DOC: What's new
       new  58f1b78   TEST: base_estimator
       new  f5690ab   DOC: classes_ and n_classes_
       new  d2050e7   DOC: put docstrings into subclasses to make them appear in the documentation
       new  86675a6   rename learn_rate -> learning_rate
       new  9799e58   raise ValueError if len(y_true) is less than or equal to 1 add some tests for weighted_r2_score
       new  0eec74e   DOC + Better default parameter values
       new  a5c71c6   DOC: cosmits
       new  bf14e98   DOC: typo
       new  5b293bc   PEP8 and DOC
       new  5f5b627   ENH: use shuffle
       new  e024e81   Make BaseWeightBoosting abstract and other misc changes
       new  a2a4333   Roll back some changes
       new  c280f57   Roll back some changes (2)
       new  e42b608   FIX: what's new
       new  11cdef7   ENH common test that set_params returns self.
       new  00f5a02   DOC: fix typo and formatting around MurmurHash3
       new  9e1fcd9   DOC add narrative overall description of classification metrics
       new  96751e0   DOC add hinge loss narrative doc
       new  754038e   DOC Set reference links in the doc
       new  21a19d5   DOC add narrative doc on zero_one loss metric
       new  be5037d   DOC add narrative doc on zero_one_score
       new  f30261e   DOC add narrative doc for precision, recall and fbeta measures
       new  879b8e1   DOC add narrative doc on roc curve
       new  23c1067   DOC add narrative doc on auc and average precision
       new  b105993   DOC add narrative doc on matthews_corrcoef
       new  ada3532   DOC add narrative doc for explained variance
       new  95fab3e   DOC add reference to multioutput metrics in regression
       new  4d4a1e0   DOC add link to clustering metrics
       new  765173c   Update what's new
       new  e7b8f6e   ENH renamed metrics.zero_one to metrics.zero_one_loss
       new  6cf5ee0   ENH rename zero_loss_score to accuracy_score
       new  8d901ba   ENH ClassifierMixin use a metrics from sklearn.metrics
       new  4db98e2   DOC add classification_report to the narrative doc
       new  46094fa   DOC typo and mistakes
       new  4c49bab   DOC comment from @amueller + several minor improvements
       new  2e911f0   TST + DOC add many examples on sklearn.metrics
       new  e3506ae   DOC typo + minor improvements
       new  b0cfa10   DOC remove redundant comment
       new  4aa4698   DOC better example with dummy estimator + link to appropriate reference
       new  17d5fef   ENH use deprecated decorator
       new  e9bc6ca   FIX DOC missing default behavior change
       new  d9a4cd6   DOC COSMIT pretty math
       new  8a8ea33   DOC clarification of api change
       new  fd5801e   FIX catch deprecation warning
       new  bb1b795   COSMIT (don't change anything see sparse_random_matrix)
       new  787e82b   Added CSC matrix testing for binarize and added type tests.
       new  cd1bc46   Move more minor contributors to what's new file.
       new  47502d3   Remove contact address.
       new  860ce37   revert changes to grid_search
       new  51fe354   FIX: import
       new  3fd0870   revert implementation of sample weights in BaseWeightBoosting.staged_score
       new  18d1209   revert a few spurious changes
       new  6219434   pep8 + pyflakes, use arrays for errors_ and weights_
       new  446fe62   init weights_ to zeros and errors_ to ones
       new  b6d1fa8   add Hastie 10.2 example
       new  f2e0b25   pep8
       new  aa61247   Typo
       new  fa5d064   FIX add doctest ellipsis
       new  f0bbfb3   FIX doctests dtype
       new  239019a   COSMIT remove empty file
       new  68fd7a3   fix: docstring for power_t in SGDClassifier was not correct (0.25 instead of 0.5)
       new  0e5bb0c   cosmit: rephrased doc
       new  37c1d09   COSMIT shut up the build by calling np.import_array in Cython modules
       new  2694e50   Merge pull request #1556 from larsmans/cython-cleanup
       new  305c304   COSMIT wrong path in .gitattributes
       new  ca2b6e0   Update sklearn/metrics/metrics.py
       new  61cb459   Added MinMaxScaler inverse_transform for issue #1552
       new  842d6ae6  Dummy commit to trigger travis
       new  bd5e502   P3K fix incorrect import
       new  74fa5a5   Merge pull request #1562 from kmike/master
       new  e90b4f6   MOD use safe_asarray instead of np.asarray
       new  277c037   MISC update my mailmap
       new  f45afdc   implement SAMME.R algorithm
       new  534831b   update adaboost hastie example and weight_boosting tests
       new  7a2cba9   Merge pull request #1561 from kyleabeauchamp/MinMaxScaler_Inverse
       new  0cf30e2   Merge pull request #1536 from kyleabeauchamp/issue-1403
       new  8291fcb   use broadcasting
       new  4aea9a8   combine real and discrete algorithms under one class
       new  c4eed7e   DOC: AdaBoostClassifier real arg
       new  5fa23ac   Typo
       new  30174c6   update example: fix histogram range
       new  8a27f1e   Merge branch 'master' of github.com:scikit-learn/scikit-learn into adaboost
       new  fe1acac   FIX: broken test
       new  de392c2   FIX: @amueller comments
       new  493f972   Cosmits, code structure and tests
       new  a0b7283   EXAMPLE: better plot_adaboost_regression
       new  16f2f29   Revert changes on plot_adaboost_error.py
       new  fef892c   added links to astropy and scipy workflow guids
       new  1a4ec9b   Merge pull request #1564 from jaquesgrobler/contributor_guide_links
       new  9c907dd   MOD address @mblondel's comments
       new  5e826d1   MOD move generating matrix out of the loop
       new  703cc06   Merge pull request #20 from glouppe/adaboost
       new  94215eb   DOC: Fixed wrong link and formatting in decomposition docs
       new  03f67a5   ENH: set default parameter values
       new  e335f64   Cleanup
       new  db0eade   Merge pull request #21 from glouppe/adaboost
       new  ee5811b   DOC: fixed latex and formatting in SVM docs
       new  1d6d08f   update adaboost example: exposes instability
       new  f22c061   Merge pull request #1563 from kuantkid/sparse_knn_graph
       new  59b5602   DOC more accurate comment in class weight computation
       new  d2b16cb   FIX make sure laplacian in spectral clustering test is really PSD
       new  33d6fbe   displace predict_proba by 1e-10
       new  0324a36   EXAMPLE: give plot_adaboost_classification some love
       new  da5ee3f   DOC: narrative documentation
       new  e3dc80b   Merge pull request #22 from glouppe/adaboost
       new  fe9b89f   DOC: more consistency in metrics docstrings
       new  340311c   DOC: More consistency in metrics and clustering metrics docstrings
       new  dbdc8d3   DOC: more consistency in docstrings for unsup clustering metrics & missing link
       new  f6548c4   DOC: fixed missed details in metrics docstrings
       new  901d933   DOC: addressed more inconsistencies in metrics docstrings
       new  452ca22   ENH: use lgamma function from John D. Cook
       new  b2a3f97   Merge branch 'master' into lgamma_port
       new  5ca8ff4   FIX: adaboost predict_proba
       new  07a0c47   only boost positive sample weights
       new  fb02cc6   FIX: only boost positive sample weights
       new  9be24cb   Merge branch 'master' of github.com:scikit-learn/scikit-learn into adaboost
       new  9a0b44b   Merge pull request #23 from glouppe/adaboost
       new  35949fb   FIX: negative and zero probabilities while boosting with SAMME.R
       new  81b84c1   FIX: doctest
       new  71ff28c   FIX: doctest and slightly larger displacement from zero probabilities (32 vs 64bit doctest instability)
       new  872ea4b   remove weighted_r2_score (leave for next PR scikit-learn#1574)
       new  4d1a652   revert spurious change in metrics.py
       new  79b3527   FIX: use full decision tree in AdaBoost and fix title in plot_forest_iris.py
       new  6ef53e8   DOC: add __doc__ to plot_adaboost_hastie_10_2.py
       new  1ca2401   FIX: reference format
       new  6b13691   FIX: show decision boundary in plot_adaboost_classification.py
       new  15b1d97   zero_one_loss now does normalize on default.
       new  84ed0f8   FIX: refactor plot_adaboost_classification.py and add legend
       new  ad64a0d   rename plot_adaboost_classification.py -> plot_adaboost_twoclass.py and add predict_twoclass method to AdaBoostClassifier
       new  8b41d8c   DOC add recall_score to new classification metrics listing
       new  204b0aa   FIX: only possible split sometimes creating children with negative or zero weight in the presence of negative sample weights
       new  07798e2   FIX issue #1457 KNeighbors should test that n_samples > 0
       new  1c59bbd   Changed a minus sign to a plus sign in the documentation of n_jobs in some files.
       new  deca12c   Merge pull request #1577 from ApproximateIdentity/n_jobs-documentation
       new  e93bd6e   FIX: improve multi-class AdaBoost example (rename to plot_adaboost_multiclass.py)
       new  a248b1e   add author
       new  a717125   typo
       new  778b77b   use metrics module and pep8
       new  cdfbd5a   typo
       new  dcd66a8   fix class ordering in two-class
       new  de87e91   DOC document gamma in chi2_kernel.
       new  040b0b8   update year in copyright notices
       new  97a60de   Fresh fixes of grammar errors
       new  d9236ed   Grammar fixes
       new  e53c1a9   Changed minus sign to plus.
       new  a2d23b6   Merge pull request #1578 from zaxtax/elastic_documentation
       new  6ea2b14   TST add common test to check if estimators overwrite their init params
       new  9c46400   ENH use only a few samples in test.
       new  696dd6f   FIX in tree and ensemble: don't overwrite random state in fit.
       new  c8e7e39   FIX don't overwrite random_state in fit in EllipticEnvelope
       new  fd8a1d2   FIX don't modify random_state in clustering algorithms.
       new  a56d213   ENH make code more clear: MiniBatchKMeans only uses random_state in first run of partial_fit.
       new  d6225ee   FIX in ward: don't overwrite n_components.
       new  c0323d1   FIX remaining parameter issues in GradientBoosting
       new  b2d3435   TST took the safty off the tests ;)
       new  97305ea   Merge pull request #1582 from ApproximateIdentity/doc-n_jobs-parallel
       new  a1448fb   ENH we require a version of sphinx sufficient for "new" numpy_ext to work
       new  da0e34b   FIX instead of appending numpy_doc to the list of extensions, directly add when creating the list
       new  eff42b9   faster sample_weight initialization
       new  6ecfbfe   speed improvements to make_gaussian_quantiles
       new  a23928a   even more speed improvements to make_gaussian_quantiles
       new  c33197d   py3k
       new  49e9391   BUG don't write comments in SVMlight dumper by default
       new  53c3164   DOC: note initialization of sample_weight if None
       new  0a319df   factorize common sample_weight check
       new  9e3ef39   DOC some sphinx fixes
       new  44fe9be   DOC fix in mds example (new interface)
       new  f5cfa64   DOC mds example: suppress warning for explicit initialization
       new  13e78ff   DOC don't use deprecated parameter rho in the lasso / enet examples
       new  7aa26a8   COSMIT typos in hierarchical clustering warning
       new  dee73f9   DOC more sphinx fixes
       new  c59af39   BUG hotfix for issue #1501: sort indices in SVMlight i/o
       new  03bf125   Merge branch 'master' of github.com:scikit-learn/scikit-learn into adaboost
       new  2904265   FIX: some nitpicks
       new  80930b1   Merge pull request #24 from glouppe/adaboost
       new  9da3468   EXAMPLE don't use deprecated interface in lasso model selection
       new  4f1bc46   COSMIT pep8 in examples
       new  d5ce8dc   COSMIT pep8
       new  0cb8963   DOC more sphinx fixes
       new  c75dd39   FIX sort indices in CSR matrix for SVM
       new  70ae016   DOC fix travis URLs in README
       new  578034e   TST sorting CSR matrix indices in SVMlight file handling
       new  d4bed81   dpgmm: setting the weights to something reasonable
       new  36e080a   TST add regression tests for Alex' fix.
       new  47b1ba0   COSMIT: pep8
       new  18352da   ENH rename cosine_kernel to cosine_similarity. Also make the test actually do something.
       new  a37445a   DOC improve cosine similarity docs
       new  e4b215f   TEST: fixing randomly failing test
       new  dfcc235   DOC fixed problem in citations in spectral_embedding
       new  ac23a7d   COSMIT typos
       new  88ea77e   ENH don't use deprecated class_prior fit parameter for NB in test
       new  79a79b2   ENH in spectral_embedding: do input validation before anything else
       new  8df28b0   TST in testing deprecated load_filenames catch deprecation warning
       new  fd117f1   TST catch expected warning in sparse coordinate descent test.
       new  9ddaac3   remove 1000`s of warnings from example
       new  19ea420   Merge pull request #1592 from jaquesgrobler/master
       new  1ffa66c   P3K: division should produce integer.
       new  57aea36   P3K: Fixed print related Python3 errors
       new  743035f   P3K: avoid iteritems / itervalues when feasible
       new  8cc7c96   P3K: decode error message in svm wrapper
       new  d24caae   DOC : missing alpha doc in LassoLars
       new  9bbca58   DOC cosmit fix column span alignment errors.
       new  b0e6681   FIX example uses old parameter name
       new  b0ac1fc   COMPatibility more careful deprecation of mode and k in SpectralClustering
       new  7fb10b9   COMP more careful deprecation of seed in SGDClassifier
       new  f0e8abe   COMP add deprecated property rho to ElasticNet
       new  1024662   COMP keep seed as init parameter of Perceptron, only deprecate
       new  1a57dab   COMP add deprecated ``labels_`` property to LinearSVC
       new  7637261   FIX deprecated properties in ElasticNet
       new  fa5a698   COMP in SVC rename self.label_ to self._label (it is redunant now but I don't want to refactor the rest of the day) and add a deprecated property label_, that points to classes_.
       new  6f3d7aa   FIX in Perceptron and doctest
       new  4ff7970   FIX in common tests: don't test init parameters that are deprecated. They might be changed.
       new  4461aae   FIX some doctests for SGD
       new  5e0dd46   COSMIT typo thanks @jaquesgrobler
       new  68bfc90   ENH don't return deprecated parameters by get_params.
       new  da336ac   FIX typo in spectral clustering deprecation
       new  04bedd8   TST catch deprecation warning when testing SVC label_ attribute, also test new classes_ attribute.
       new  39b3d41   ENH: output processing speed in MB/s for vectorizer example
       new  d3a0aaa   Initial work on hashing vectorizer
       new  f10cb47   Add fit_transform support using the TransformerMixin + missing ABCMeta marker
       new  9c66bb7   Improved the clustering example with HashingVectorizer
       new  ecb706a   Remove TransformerMixin from vectorizers and do a direct fit_transform alias for HashingVectorizer instead
       new  e27c199   Improve module docstring of document clustering example
       new  d7b2f92   COSMIT make BaseVectorizer a mixin
       new  add5689   cosmit
       new  72e3054   Updated whats_new.rst
       new  e57c6b2   DOC: Started section on hashing vectorizer in narrative section
       new  fe38fb9   DOC: narrative doc for HashingVectorizer
       new  de9d734   DOC: typos
       new  26d8bf7   DOC: merged the whats new entries and add links to the narrative doc
       new  592c0f7   DOC: address @mblondel's comments
       new  0c2a87f   ENH: measure feature extraction speed in document classification example
       new  d564e66   DOC: typos
       new  5ae14af   DOC copyedit HashingVectorizer docs
       new  fb8ae6a   DOC reorganized whatsnew a bit, put new estimators on top.
       new  56b4776   DOC added user guide links to all estimators on the whatsnew page
       new  c708381   DOC some more fixes for whatsnew
       new  5f06795   EXAMPLES add header to hash_vs_dict_vectorizer.py - otherwise it won't show in the html docs.
       new  bf1aac6   COSMIT pep8
       new  6f8b2a5   ENH undo renaming of class_prior to class_weight in naive bayes
       new  d37cead   Merge pull request #1529 from vene/lgamma_port
       new  4269bf2   DOC some more minor fixes to syntax / links
       new  c1ffb19   Merge pull request #1598 from amueller/naive_bayes_class_prior_rename_revert
       new  637d891   DOC fix indentation typo
       new  906290e   ENH: rng local to tests
       new  c050d28   TEST: add a test of sample weights
       new  94404fd   TST: improve last test
       new  ceb78a7   BUG: fix sample_weights in ridge
       new  e3dd35f   BUG: shape bug in tests
       new  cc22dc8   BUG: fix sample weights in ridge
       new  83a9151   BUG: Ridge: sample_weights in intercept
       new  8c41be3   TST/BUG: test_common sample_weights in ridge
       new  ada1578   DOC added commit counts for 0.13 to whatsnew, added website for Rob Zinkov aka zxtx
       new  1a030f0   ENH: random reassignements in MiniBatchKMeans
       new  22ba844   ENH: fine tuning to the random assignement
       new  2b61805   DOC: example of dict-learning with KMeans
       new  70a3090   DOC: improve online KMeans example
       new  4ba01b6   DOC: dict learning with kmeans in narrative doc
       new  134df77   DOC: fix typo
       new  c83adb5   BUG: check n_clusters == len(cluster_centers_)
       new  10d4cab   PEP8
       new  5d97a11   DOC: change the example to lighter dataset
       new  6009c05   ENH: more control on reassignment in MiniBachKMeans
       new  cab6e72   DOC: link to example
       new  aff0868   DOC: add comment
       new  a68a340   DOC: complete whats_new
       new  a27c0db   TST: random reassignment in MiniBatchKmeans
       new  d2341b5   TST: test verbosity mini_bach_kmeans
       new  2a8db773  ENH: control random_state in MiniBatchKMeans
       new  bd5671e   Added n_jobs to multiclass.py
       new  d0feab0   Revisions due to previous pull request.
       new  7a595b8   COSMIT: simplify parallel code in multiclass
       new  ed169d9   First draft of new Affinity Propogation description in docs.
       new  4ec84f1   Who doesn't love equations?
       new  d498833   Spelling
       new  3fbd098   Update doc/modules/clustering.rst
       new  0028fb4   DOC: put math at back, simplify formulation
       new  4d73592   MISC: fix rst in whats_new
       new  10c690f   MISC: index arrays with integers
       new  780c9b9   DOC: voronoi + kmeans picture
       new  e52e247   COSMIT pep8
       new  53a9b64   Add temporary survey banner
       new  3e89aa5   DOC updated commit counts.
       new  19beac8   REL change version to 0.14-git everywhere, update news, support page.
       new  9cf67f9   REL changed version to 0.13 everywhere.
       new  a9e6d5c   add decision_function and staged_decision_function and refactor some code
       new  f973c27   Merge remote-tracking branch 'upstream/master' into treeweights
       new  86e8b0d   website: fix for survey bar
       new  74b2382   website: fix for survey bar
       new  7267e4e   COSMIT rm deprecated svm.sparse module
       new  69d1650   COSMIT rm deprecated attrs from [LQ]DA
       new  b69fb64   BUG last references to svm.sparse
       new  f8a7274   DOC typo: Pereptron -> Perceptron.
       new  c770e2f   Merge pull request #1604 from darkrho/doc-linear-model-typo
       new  f7e0023   ENH : add reconstruction_err_ for NMF with sparse input
       new  07c56d7   COSMIT rm deprecated stuff
       new  feee36d   FIX: Cast floats to int before slicing in robust_covariance
       new  a186db4   BUG fix failing doctest
       new  58723f6   BUG one more failing doctest
       new  90dc434   Merge tag '0.13' into releases
       new  b4f780b   Merge branch 'releases' into dfsg
       new  0749be2   Merge branch 'dfsg' into debian
       new  6462dcd   changelog + refreshing patchset
       new  226f77a   debian/rules - removing exclusion of tests previously failed
       new  351a30b   added patch changeset_567460a602b4e2fc6029b2d063988408061b7974.diff to "cherry-pick" 567460a602b4e2fc6029b2d063988408061b7974 (BF: explicitly mark train_test_split as not the one for nosetesting)
       new  123a991   BF: explicitly mark train_test_split as not the one for nosetesting
       new  3d89ae1   ENH: remove boost_method parameter and use a string as switch
       new  2497856   ENH: weights_ -> estimator_weights_
       new  bae2d1b   BUG move label_ from BaseLibSVM to BaseSVC
       new  31f412a   COSMIT decouple regression and classification in SVMs
       new  7ed0063   Merge pull request #25 from glouppe/adaboost
       new  f0fb4f4   pep8
       new  123c318   use scipy.linalg in test_nmf.py
       new  a119337   adding comment on why sparse frobenius is ok as done
       new  d263ee4   DOC: typo in warning
       new  5be04bf   fix: map labels to {0, 1}
       new  83026a8   fix: deviance computation in BinomialDeviance was wrong (ignored cases where y == 0) - thanks to ChrisBeaumont for reporting this issue
       new  caeb299   raise ValueError if division through zero in LogOddsEstimator
       new  1a09e20   add loss function for gradient boosting binomial deviance
       new  1ba8f07   remove the equaldistance code warning, replace with doc warnings
       new  e47bd5a   typo fix
       new  c83ce41   remove warning
       new  0985264   warning removal
       new  82bcda2   update warning box
       new  329aa4e   deprecation warnings, indent fix
       new  abee93d   andys suggestions and test
       new  44360a3   pep8 and assert_equal instead of assert
       new  e369c43   COSMIT remove unused imports, pep8
       new  9ddb9ca   correct docstring
       new  11b126f   Merge branch 'master' into gbrt-deviance-fix
       new  c0d4015   BUG allow outlier_label=0 in RadiusNeighborClassifier
       new  2dfe13d   BUG in RadiusNeighborClassifier outlier handling
       new  30a3a0d   Merge pull request #1576 from mrorii/fix_kneighbors
       new  74d9b66   ENH rewrite radius-NN classifier's outlier handling
       new  e1b583e   WIP: doc hyperlinks, fixed size thumbnails
       new  0e160aa   gzip support, whats_new
       new  d365c1b   use Sphinx searchindex.js
       new  4c871de   no_image.png for examples w/o thumbnail
       new  a10365b   fix paths for Windows
       new  38af9c5   use unique from sklearn backports (return_inverse)
       new  23cb9ea   FIX: pprett comments
       new  5bfe75a   DOC: Added a References section in _samme_proba
       new  2aa64fa   COSMIT: flake8
       new  3cc2054   ENH: weight -> estimator_weight
       new  6751e52   ENH: weight -> estimator_weight (2)
       new  2e8a70e   ENH: weight -> estimator_weight (3)
       new  1eb07eb   EXAMPLE: better x-axis label
       new  5d5e376   EXAMPLE (2)
       new  efb382e   Merge pull request #26 from glouppe/adaboost
       new  a23a98e   Update travis config to remove -qq flag for scipy
       new  965b6e4   links for scipy, cleanup
       new  cea09f1   Merge pull request #1607 from agramfort/reconstruction_err_nmf_sparse
       new  1271b4d   DOC: small fix in the regression's score method documentation
       new  c56a915   update adaboost regression example and use estimator_errors_
       new  79d4175   rm n_estimators argument from predict methods
       new  8cb6e18   DOC: fix docstring for make_gaussian_quantiles
       new  c44d63c   changed wording in linear model docs about Normalized. It was frustrating me haha
       new  bdf2d4b   TST some more tests for multi output lars
       new  707c0f4   FIX: alpha=.5 and use more difficult dataset in two-class example. Add mean and cov arguments to make_gaussian_quantiles
       new  f431060   FIX: learning_rate default value consistency
       new  e421c01   FIX: TypeError message if base_estimator does not support class probabilities
       new  54ee881   FIX: comments from @ogrisel
       new  029b29c   BUG: reassignment_ratio == 0 in MiniBatchKmeans
       new  c06444c   make learning_rate=1 default for classification
       new  19c6a57   only sum sample_weight once
       new  2699cff   Minor bug fix in metrics.adjusted_rand_score
       new  f319bb9   Added tests
       new  c638437   rm sphinx/docutils formatting in exception messages
       new  ab27f37   inline comment about learning_rate in hastie example
       new  8450880   add note about SAMME.R converging faster than SAMME
       new  68d2388   COSMIT translate lgamma replacement to C and clean it up
       new  16244a0   BUG: sparse center reassignment MiniBatchKMeans
       new  8214100   BUG: sparse vs non sparse centers
       new  a2e37eb   COSMIT add lgamma to gitattributes
       new  39a431a   add note about y coding construction
       new  e21399f   add description of dataset in two-class example
       new  71f81eb   fix missing parenthesis in make_hastie_10_2 dataset
       new  c6db8e6   add warning for no internet
       new  42b2995   BUG: fix test to use sparse array
       new  4a797e5   FIX: make_hastie_10_2 reference docstring
       new  b56674a   DOC: add a short dataset description in hastie example
       new  2eeb724   DOC: narrative documentation
       new  6017e4a   DOC fix typo in LinearSVC error message
       new  c35d509   DOC: reference for discretise option
       new  e7ad4fd   Merge pull request #27 from glouppe/adaboost
       new  5cde80b   DOC update SMART notation in TfidfTransformer docs
       new  837093d   P3K: use print as a function in the examples
       new  3485b2f   FIX make error message work when return_path=False. Btw I feel that getting "references" for numbers out of numpy arrays is pretty ugly.
       new  23dfc0c   TST fix random states in all dict learning tests, make test independent of test sequence.
       new  ba777f4   DOC: make distinction between evaluation and pairwise metrics.
       new  1a1c4ca   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  4ed0edb   Revert "trying travis cfg with system-site-packages"
       new  de898ee   COSMIT pep8
       new  f3b6dd3   Adding support indices in svm for sparse matrices
       new  ba5c99e   COSMIT PEP8
       new  418b6a1   Adding test to check support_ is equal in dense and sparse matrices
       new  f9d15a3   COSMIT PEP8
       new  a91fd1d   Recompiled base
       new  2831ee2   DOC add return values of cross_val_score and train_test_split to docstrings.
       new  a7db2af   import pylab as pl
       new  c23dd74   remove check for fit_predict
       new  f04e368   fix importance test and test both SAMME and SAMME.R algs
       new  8835435   don't show class B probabilities in two-class example
       new  07e6834   two-class decision scores -> decision scores
       new  e8a8101   clarification on two-class decision scores plot
       new  491319d   explain decision scores in two-class example
       new  536e50d   fix AdaBoost.R2 and update example
       new  29fa16c   DOC: loss_function
       new  63d524e   fix failing tests
       new  5e49c13   fix failing doctest
       new  79f606f   FIX: doctest
       new  ed93183   EXAMPLE: add AdaBoost to plot_classifier_comparison
       new  06220ee   Merge pull request #28 from glouppe/adaboost
       new  ab4cd62   ENH added test, started on cross_val_score
       new  6eb2c1b   ENH adding SomeScore objects for better (?!) grid search interface.
       new  e8fa4c3   ENH refactor, taking @GaelVaroquaux's and @ogrisel's suggestions into account
       new  b4ef84c   ENH deprecated ``score_func``, introduced ``score`` parameter in GridSearchCV
       new  be375cb   TST test giving score as string in GridSearchCV
       new  e58e135   FIX rename ``score`` to ``scoring`` because of the name-clash with the ``score`` function.
       new  a347fbf   FIX two score objects, adjust tests to new interface
       new  7bcfa36   ENH remove old interface completely from tests.
       new  1db5e70   DOC fix docstring
       new  801e265   ENH working on cross_val_score, trying to simplify unsupervised treatment.
       new  50542dc   ENH better testing of old an new interface. Still a bit to do for unsupervised grid search, though.
       new  773c799   FIX usage of scores for unsupervised algorithms.
       new  1ec4a21   ENH use new api in permutation_test_score, don't use old api in testing.
       new  e6484c4   ENH fbeta score working, more tests
       new  46c2649   DOC-string for AsScorer
       new  053b6e7   ENH renamed ap and auc, added RecallScorrer
       new  e2e8d0b   DOC narrative docs for scoring functions. Put them next to GridSearchCV. Should they go into metrics?
       new  ffa0b9d   ENH update example, minor fix.
       new  8111ec0   DOC improve cross validation and grid search docstring
       new  53f8973   FIX rename error
       new  7aaee87   DOC add whatsnew entry
       new  b256d13   DOC fixed formatting in user guide
       new  98b5d10   FIX example
       new  aee8e8d   DOC added a new template to sphinx so view the "__call__" function.
       new  2734bef   COSMIT address @ogrisel's comment.
       new  c04da31   FIX rename ZeroOneScorer to AccuracyScorer
       new  a728ee2   DOCFIX for zero_one_score / accuracy_score renaming
       new  fb8285a   DOC add narrative about score func objects to the model_evaluation docs.
       new  8511f48   ENH rename scorer objects to lowercase as they are instances, not classes
       new  b118b93   DOC minor fixes in pairwise docs.
       new  4fc3f43   ENH/DOC add "score_objects" function for documenting the score object dict.
       new  394f87b   DOC add metrics.score_objects to the references
       new  e1d9376   DOC use table from score_functions docstring in model_evaulation narrative.
       new  a92adad   DOC move scoring function narrative above dummy estimators, fix tables, some refinement.
       new  0d7e1a6   DOC minor fixes in score_objects documentation.
       new  e9556eb   DOC better table of score functions in grid-search docs.
       new  2d9cb81   ENH GridSearchCV and cross_val_score check whether the returned score is actually a number, not an array (otherwise cross_val_score returns bogus).
       new  43e1c39   TST improve coverage of permutation test scores
       new  f650ad4   TST slightly better test coverage in cross_val_score
       new  374f81a   COSMIT built-in typo
       new  385e581   DOC some improvements as suggested by @ogrisel
       new  99f6a56   TST add test for pickling custom scorer objects
       new  4847227   DOC more improvements by @ogrisel
       new  b5cc3ba   COSMIT rename AsScorer to Scorer
       new  31254c2   COSMIT :: in rst is easier for syntax highlighters
       new  03aa748   DOC: minor formatting in model_evaluation.rst
       new  8db5750   DOC: minor rst issues
       new  9d8e6e9   DOC: misc rst formatting
       new  6711dfc   API consistency with gradient boosting: loss_function -> loss
       new  ee96e7d   MISC moved score_objects.py to scorer.py, added module level doc string and license note.
       new  6f79e47   DOC add kwargs in Scorer to docstring.
       new  51c3678   ENH add ``__repr__`` to Scorer
       new  318278e   DOC addressed @ogrisel's comments.
       new  0730e73   COSMIT text reflow
       new  e375677   MISC pep8: rename scorers to SCORERS, remove score_objects getter
       new  dbfd837   DOC remove duplicate table, add references to appropriate user guide section to docstrings of cross_val_score, GridSearchCV and permutation_test_score
       new  266fd0c   DOC add note on deprecation of score_func to whatsnew
       new  3d3a305   FIX imports for Scorer and SCORERS
       new  31fb403   DOC fixes in whatsnew, typo
       new  8ca3647   TST smoke test repr
       new  8ecc504   FIX: some of Gael comments
       new  92be923   ENH import six package for Py2/Py3 compat in a single codebase
       new  9b183a8   Merge pull request #29 from glouppe/adaboost
       new  41f3f39   minor edits in docs
       new  cde9fbf   P3K: support for py3k in dict_vectorizer module
       new  19ed9b0   ENH refactor univariate feature selection
       new  57b86fc   P3K use six.string_types and six.PY3
       new  819b322   P3K one more iteritems
       new  be2394a   DOC: notes about examples and minor edits
       new  fb95fc2   PY3: Fix stdout capture in graph lasso test
       new  0693734   P3K More python 2 / 3 compat in tree exports
       new  db52852   COSMIT rm Python 2.5 and Jython compat from six
       new  68d0423   BUG fix import problem in preprocessing
       new  f4a283a   COSMIT: prettify code and figure in example
       new  1f95b18   COSMIT
       new  f03e7f3   Merge pull request #1644 from jaquesgrobler/doc_url_error
       new  c235ef4   Merge branch 'treeweights'
       new  171f682   What's new: Adaboost
       new  7d4034a   COSMIT removed unused imports, fixed error message in test of boosting
       new  293c78b   ENH break ties in OvO using scores
       new  0449d51   TST test for breaking OVO ties
       new  f7c79fb   test case for unfitted idf vector
       new  fcca1df   raise ValueError for unfitted idf vector
       new  1adfed9   Merge pull request #1656 from rlmv/idf_diag
       new  f7f5424   Remove compute_importances parameter
       new  f229d38   What's new
       new  c11f943   FIX docstring deletions
       new  72cce35   ADD test coverage for _check_stop_list
       new  357db85   FIX comment typo
       new  789d1d3   ADD test cases to fill out VectorizerMixin coverage
       new  9a012e8   ADD another VectorizerMixin test
       new  f259bf4   ADD test for get_feature_names
       new  96f79fd   ADD test for tfidf fit with incompatible n_features
       new  9a5c70c   ADD test for TfidfVectorizer attribute setters
       new  6bd1c75   ENH: Remove compute_importances in AdaBoost
       new  5c3dc9a   ENH: Update feature_importances in GBRT
       new  12c9262   ENH: remove "mse" method and simplify
       new  3963311   COSMIT
       new  7130c99   DOC: feature importances
       new  a54809e   Merge pull request #1657 from glouppe/feature-importances
       new  b3e6751    MV Mixin tests to CountVectorizer tests
       new  0520ecc   RM CV import
       new  c8e8614   MV _check_stop_list tests to CV get_stop_words
       new  0e97a7f   Merge pull request #1660 from rlmv/fe_tests
       new  2c44dd6   DOC: add balance_weights to reference
       new  0b2f92d   EXAMPLE: compute_importances=True is no longer required (1)
       new  46f4e91   EXAMPLE: compute_importances=True is no longer required (2)
       new  4fb6c88   DOC: narrative documentation on feature importances
       new  2297306   COSMIT pep8
       new  9f1c346   FIX: variable naming inconsistency in NMF
       new  4a76b06   ENH: precompute X_argsorted when possible
       new  ec49cba   DOC: X_argsorted
       new  ce1bc1b   Flake8
       new  1967a0b   P3K use six to have a python 2 & 3 compatible code base
       new  0fe516e   P3K StringIO vs BytesIO
       new  1c0d422   DOC fix failing doctest due to unicode_literals
       new  b445537   Cosmit: more explicit xlabel.
       new  135636e   Cosmit: more explicit label.
       new  759e6fa   ENH: use isinstance instead
       new  b126421   Merge pull request #1668 from glouppe/adaboost-tree
       new  4c60052   DOC whitespace in doctest
       new  c809368   BUG revert P3K changes that broke mldata tests
       new  041abe3   rm gender classification example
       new  b99dbab   ENH get rid of imports in test_common by checking by names, not classes.
       new  a3b55f4   ENH fix test_estimators_overwrite_params to also test regressors and transformers. Then fix all the regressors and transformers ... meh!
       new  6920295   ENH set the random state to avoid heisenfailures
       new  7d57b9e   COSMIT pep8, removing unused imports
       new  85ec0fd   P3K death to the print statement
       new  c481ff8   P3K fix broken doctest and add forgotten print_function import
       new  cc401c9   DOC no more need for compute_importances in trees
       new  9f6c8bb   DOC copyedit FeatureHasher narrative
       new  16e595a   ENH move covtype loading to sklearn.datasets
       new  81428f7   TST covertype loader
       new  f63cdbb   BUG: update joblib to 0.7.0d
       new  50dbadb   DOC copyedit FeatureHasher narrative further
       new  5419878   P3K range vs. xrange
       new  645216f   Merge pull request #1524 from amueller/break_ovo_ties
       new  d949425   chi2 and additive_chi2 raise error if input are sparse matrices
       new  f01c506   Added same for additive_chi2_kernel
       new  0f39df7   Fixed pep8 issues
       new  829d3fc   pairwise_distance_functions renamed to PAIRWISE_DISTANCE_FUNCTIONS
       new  bec1226   Made more changes renamed pairwise_kernel_functions, kernel_params to allcaps
       new  0ca0e15   DOC pretty math in kernel docstrings
       new  1c60618   BUG MinMaxScaler missing from preprocessing.__all__
       new  c0d1d88   FIX remove dtype from covertype, add fetch_covtype to init, add missing docstrings.
       new  60b3181   Only reassign centers if to_reassign.sum() > 1
       new  a58ac8c   TST: add a test for empty reassignment in MBKmeans
       new  86ab060   Added test for fit_transform(X)==fit(X).transform(X)
       new  3cc8158   Fixed pep8 issues
       new  0edc94b   BUG: highly-degenerate roc curves
       new  d821417   BUG: fix change of behavior in last commit
       new  d51fa05   DOC: add example and ref to lars_path in lasso_path
       new  42c7496   Merge branch 'master' into gbrt-deviance-fix
       new  a7c6b06   BUG in KernelPCA: wrong default value for gamma
       new  ab92433   Merge pull request #1688 from hrishikeshio/fit_transform
       new  5cfd45e   FIX doctest kernelpca
       new  7dd1ade   ENH speed up RBFSampler by ~10%
       new  2d63eb2   BUG oops, removed validation by accident
       new  070528b   added patch changeset_0edc94b528b630eae63f3409c4c4d95a2e6184a4.diff to "cherry-pick" 0edc94b528b630eae63f3409c4c4d95a2e6184a4 (BUG: highly-degenerate roc curves)
       new  b46daf4   added patch changeset_d821417ebf7570397a344a8d670045cd2d2e590a.diff to "cherry-pick" d821417ebf7570397a344a8d670045cd2d2e590a (BUG: fix change of behavior in last commit)
       new  2a47d8d   Merge branch 'master' into gbrt-deviance-fix
       new  2aef122   changelog for 0.13-2
       new  186858e   ENH get rid of most imports in test_common
       new  5e3230c   TST stronger tests for arbitrary classes. make explicit what works and what doesn't.
       new  2353796   WEB rm survey bar
       new  6aabf3b   DOC typo: Pereptron -> Perceptron.
       new  9f39791   BUG move label_ from BaseLibSVM to BaseSVC
       new  1dc8526   COSMIT decouple regression and classification in SVMs
       new  66e7482   DOC: typo in warning
       new  cc4b85d   DOC: small fix in the regression's score method documentation
       new  664f866   changed wording in linear model docs about Normalized. It was frustrating me haha
       new  b416113   TST some more tests for multi output lars
       new  cfef316   BUG: reassignment_ratio == 0 in MiniBatchKmeans
       new  5f7460f   Minor bug fix in metrics.adjusted_rand_score
       new  a21f564   COSMIT translate lgamma replacement to C and clean it up
       new  197e3c0   BUG: sparse center reassignment MiniBatchKMeans
       new  12264a1   DOC fix typo in LinearSVC error message
       new  d4d8d66   DOC: reference for discretise option
       new  6441e6e   DOC update SMART notation in TfidfTransformer docs
       new  b12ee6e   DOC: make distinction between evaluation and pairwise metrics.
       new  f0f953b   FIX make error message work when return_path=False
       new  944a7ef   TST fix random states in all dict learning tests, make test independent of test sequence.
       new  e14543c   DOC add return values of cross_val_score and train_test_split to docstrings.
       new  5f7e81d   COSMIT :: in rst is easier for syntax highlighters
       new  204ff55   ENH refactor univariate feature selection
       new  1f1f385   raise ValueError for unfitted idf vector
       new  f2333e4   FIX docstring deletions
       new  ff78bf9   ADD test coverage for _check_stop_list
       new  096610e   FIX comment typo
       new  446e675   TST improve feature_extraction.text coverage
       new  2fd0147   DOC: add balance_weights to reference
       new  9d49b6d   FIX: variable naming inconsistency in NMF
       new  c59b2f6   ENH get rid of imports in test_common by checking by names, not classes.
       new  5b9a65e   DOC copyedit FeatureHasher narrative
       new  bff549f   ENH move covtype loading to sklearn.datasets
       new  437644e   BUG: update joblib to 0.7.0d
       new  7379aae   ENH break ties in OvO using scores
       new  0757503   chi2 and additive_chi2 raise error if input are sparse matrices
       new  34ef825   DOC pretty math in kernel docstrings
       new  6c2b64d   BUG MinMaxScaler missing from preprocessing.__all__
       new  b52269f   BUG MBKmeans only reassign centers if to_reassign.sum() > 1
       new  84eaed3   BUG: highly-degenerate roc curves
       new  5cde50c   DOC: add example and ref to lars_path in lasso_path
       new  c656b10   BUG fix broken grid search example
       new  a6b3d24   BUG fix broken grid search example
       new  f4bd2ca   fix: map labels to {0, 1}
       new  2da9f01   fix: deviance computation in BinomialDeviance was wrong (ignored cases where y == 0) - thanks to ChrisBeaumont for reporting this issue
       new  4064301   raise ValueError if division through zero in LogOddsEstimator
       new  4a9ecf0   BUG in KernelPCA: wrong default value for gamma
       new  91c9fab   TST fit_transform(X)==fit(X).transform(X)
       new  aefa925   ENH speed up RBFSampler by ~10%
       new  0adb851   ENH fix test_estimators_overwrite_params to test regressors and transformers
       new  4194ffc   COSMIT update mailmap
       new  1746100   COSMIT update mailmap
       new  a7e1231   BUG: Build random forests the same way regardless of n_jobs and add a test for this. Don't predict in parallel since the cost of copying memory in joblib outweighs the speedups for random forests. Fixes #1685.
       new  61447db   COSMIT: Fix up a loop.
       new  274d10a   COSMIT: Better assert.
       new  704ee89   DOC: Update new magic numbers in docs since random forests train differently now.
       new  9c30c92   FIX: sklearn.ensemble.forest: Refactor to remove references to parallelism in predict() functions.
       new  60374ee   Added test_regressor_pickle to tests.
       new  2ba21ec   Added test_classifiers_pickle to tests.
       new  940b516   Finished adding pickle tests.
       new  e681ec5   Removed the use of StringIO, using pickle.dumps instead.
       new  17bd1ed   FIX rebasing trouble in common tests: the meaning of dont_test changed
       new  1d958b2   FIX don't compare strings with "is". that is really not robust!
       new  94318a8   ENH in transformer pickle test, only test transformers that provide a 'transform' method. and only test that.
       new  6d9935b   ENH in common tests, use long variable names for all tests
       new  db6494c   FIX remove all unseeded random variables from common tests.
       new  e109175   COSMIT: fix excessive indentation.
       new  51da03a   Merge pull request #1695 from mrjbq7/issue-1694
       new  544e949   BUG: Fix performance regression on large datasets in random forest. The crossover point seems to be around 5k rows. To get a fast predict time on fewer than 5k rows, set n_jobs to 1 after fit and before predict. Alternatives would be to split n_jobs into two parameters, for fit and predict, or to have a complicated heuristic based on n_estimators, n_samples, n_features, tree depth, etc.
       new  edb7592   DOC: Emphasize that n_jobs is for fit and predict methods in random forests.
       new  7c05c24   Merge pull request #1700 from erg/rf
       new  e6c85a3   decision_function forces dense output (in the case of sparse coef_)
       new  ff937b4   FIX: use DOUBLE_t type
       new  ace0d6a   Merge pull request #1705 from glouppe/tree-fix
       new  c9877e9   COSMIT pep8: blank line contains whitespace
       new  73d3d3f   DOC added sentence about oob_decision_function_ containing NaN to docstring. Still need some narrative about oob score.
       new  6bf4c90   ENH sparsify method for L1-reg linear models
       new  dd338b1   BF: explicitly mark train_test_split as not the one for nosetesting
       new  9ae3094   TST stronger tests for arbitrary classes. make explicit what works and what doesn't.
       new  5a2b047   BUG allow outlier_label=0 in RadiusNeighborClassifier
       new  2f2a382   BUG in RadiusNeighborClassifier outlier handling
       new  3ff1a15   ENH rewrite radius-NN classifier's outlier handling
       new  1befd65   ENH set the random state to avoid heisenfailures
       new  6f293a2   COSMIT pep8
       new  f4bc9a2   DOC some copyediting
       new  8075887   REL increased version number, added whatsnew
       new  68d378d   DOC developer guidelines for unit tests and classes_
       new  bc33004   DOC add 0.13.1 changelog to whats_new.rst
       new  483e0b6   DOC dev guide: random_state_ + @amueller's remarks
       new  71d4e42   DOC r2_score may return negative values
       new  4d85d16   DOC dev guide: deprecation
       new  0bc4c72   ENH: support float value for max_features
       new  72a8017   DOC: if float, then max_features is a percentage
       new  c41ccbb   ENH: Defer parameter checking of trees
       new  495b6c3   DOC: GBRT max_features
       new  d8d4ede   TEST: added test
       new  b8f1b13   ENH: use numbers
       new  d49af04   FIX: numpy integers
       new  2adbee6   Merge branch 'sparse-coef'
       new  472a1f5   COSMIT callable instead of hasattr __call__
       new  b9a8e08   BUG update graph_laplacian to upstream SciPy version
       new  746c8bd   DOC rm failing doctest on graph_laplacian
       new  f4d4b4c   DOC fix text vectorizer docs and add NLTK example
       new  ec47a6b   DOC fix broken doctests for feature_extraction.text
       new  018d6ad   StratifiedKFold: remove pointless copy of labels
       new  b099d46   DOC add random_state parameter to docs of LogisticRegression and LinearSVC
       new  b83f112   TST/FIX set random_state in logistic regression tests
       new  2afe4d5   TST/FIX always use "almost equal" for floats.
       new  8453dff   ENH CountVectorizer using arrays instead of lists
       new  cc4c3c3   BUG restore empty vocabulary exc in CountVectorizer
       new  9c19eb3   ENH prevent copying of indices in CountVectorizer
       new  74fb91d   DOC credit @ephes
       new  f150a06   Merge pull request #1713 from larsmans/vectorizer-memory-use
       new  a2ed7a3   FIX MinMaxScaler bug.
       new  c24d0e4   COSMIT use callable instead of hasattr
       new  192802e   Merge pull request #1727 from amueller/min_max_scaler_fix
       new  62ba53a   FIX : fix kfold balance due to int rounding
       new  b8d23fa   FIX : test due to KFold change
       new  1098c3c   FIX : better fix of KFold balance
       new  adb85bb   fix doctest
       new  1920b88   Removed an unnecessary if statement in KFold __iter__ method.
       new  470e01c   Improved the test that checks the balance of sizes of folds returned by KFold.
       new  7394c1c   DOC Corrected the docstring of KFold about the sizes of the folds.
       new  a8b4712   TST : improve test_kfold_balance test
       new  5d9ce74   update what's new
       new  5bcf25f   ENH multilabel metrics: accuracy, Hamming, 0-1 loss
       new  2d5ae28   BUG broke the what's new while rebasing
       new  2797aae   TST : improve again  test_kfold_balance test
       new  5058fac   Merge pull request #1726 from agramfort/round_kfold
       new  8ab798b   DOC FIX foating point issue
       new  f7cf1ed   Merge pull request #1730 from arjoly/doc-feature-selection
       new  db6f005   TST FIX random state for LibLinear sparse tests
       new  0c94b55   ENH add randomized hyperparameter optimization
       new  57260da   DOC fixed links in whatsnew
       new  3bf53c8   Ball Tree, KD Tree, and tests
       new  268ea8e   PEP8
       new  845fc5d   Merge pull request #1712 from glouppe/tree-maxfeatures
       new  4229707   What's new: float values support for max_features
       new  cd998e2   What's new: fix indentation
       new  bde8b7a   Update coordinate_descent.py
       new  bb9d11e   Fix tests for scipy <= 0.9
       new  523ec76   Merge pull request #1736 from jamestwebber/patch-1
       new  3d45191   ENH set min_df in fe.text back to 1
       new  4690743   DOC improve mini-batch k-means narrative
       new  5a8114d   COSMIT Moved the test_roc_curve_one_label test where other ROC curve tests are.
       new  fcad90d   Merge pull request #1740 from tjanez/move_roc_curve_test
       new  e42e530   COSMIT pep8
       new  48ed945   DOC FIX links on grid search narrative
       new  2006e27   FIX numpy 1.3 issues with multilabel metrics
       new  50e22dd   Update load_svmlight_file docstring.
       new  4376f73   Merge pull request #1741 from arjoly/metrics-fix-np-1.3
       new  6bf90fe   TST compute_class_weight in utils
       new  dc2c5c4   FIX + TST + DOC compute_class_weight
       new  9633d9d   ENH use bincount in compute_class_weight
       new  f39d86a   BUG use fixes.unique
       new  ffad26a   FIX compute_class_weight edge case
       new  2b6532c   BUG in SVM tests
       new  da1ce26   FIX make_classification now outputs integer labels
       new  88b1ba5   DOC formatting (k_means)
       new  29ed168   BUG fix compute_class_weights issue in SGD
       new  0702cc6   Merge pull request #1753 from NelleV/FIX
       new  2cb8b5e   P3K some more fixes in random places
       new  6c609f6   DOC OpenBLAS is more dangerous than I thought
       new  8af659f   DOC oops, typo
       new  12d48d0   FIX KFold should return the same result when indices=True and when indices=False.
       new  e1789e2   ENH Function auc_score should throw an error when y_true doesn't contain two unique class values.
       new  4964474   changed warnings in grid_search.py related to loss_func and score_func being passed
       new  d238e40   speed up KD tree construction by ~25%
       new  9fa27c8   add author & license information to pyx files
       new  7975ccb   COSMIT get rid of undocumented attributes on SVMs
       new  41ae851   PEP8 and allow non-bool truth values in CD
       new  4610103   BUG + ENH: removal of components in kernel PCA
       new  cc37fd6   Merge pull request #1758 from larsmans/kernelpca-fix
       new  21ba76d   DOC some sphinx / rst fixes
       new  edd1555   MISC minor fixes in examples
       new  01a745e   DOC FIX column span alignment problem in NMF ^^
       new  09f2d90   P3K make feature_extraction.text work
       new  e4ee70d   ENH - 3x speedup in the isotonic regression
       new  30ccbd1   Fix comment: returns fbeta_score, not f1_score
       new  ce72c0b   BUG failing doctest
       new  bb7da14   COSMIT typo
       new  32325eb   DOC IsotonicRegression wasn't in the changelog at all
       new  19591be   P3K all of feature_extraction passes tests on Py2 and 3
       new  d2254e4   DOC clarify column ordering in SVC scores
       new  a360542   DOC fixing some more rst / sphinx errors :-/
       new  3bb86ab   DOC more sphinx stuff.
       new  cacf6f9   FIX: X was converted twice.
       new  8ee4513   COSMIT DictVectorizer.inverse_transform readability
       new  9e44e3a   fix the kwarg name
       new  444bfd2   updated the .c file
       new  992e982   remade the cython with 0.18
       new  0053aa5   Merge pull request #1767 from rmcgibbo/balltree_docstring
       new  b4b3b1b   DOC CountVectorizer does NOT do stopword filtering by default
       new  b80f6a8   ENH allow SelectKBest to select all features in a parameter search
       new  968b337   ENH don't recompute distances in MBKMeans
       new  b37791b   ENH cut MiniBatchKMeans memory usage in half for large n_clusters
       new  bb6c7b1   DOC installation instructions: MacPorts, fix types, stdeb instructions
       new  35d4417   DOC Allowing a list of param_grids means GridSearchCV is more than grids
       new  5edfce0   DOC clarify relationship between pos_label and average parameters for PRF family
       new  804bf9e   Merge pull request #1772 from jnothman/comment_exhaustive_search
       new  cb73742   Merge pull request #1773 from jnothman/prf_docstring
       new  91d9dc9   BUG StandardScaler would ignore with_std for CSR input
       new  21d1e39   DOC add roll your own estimator docs
       new  68de819   TYPO fix
       new  b1f5c0e   bring text classification somewhat closer to current API
       new  7997993   FIX for iid weighting in grid-search
       new  65d1ef5   DOC FIX finite precision
       new  0fe17b2   typo
       new  d61f36c   Clarify docstring in lars_path
       new  f95dfe7   BUG SGDClassifier and friends did not forget labels_ in re-fit
       new  d96a3c2   ENH optimizations in sklearn.cross_validation
       new  fd90edc   DOC clarify C parameter on LogisticRegression
       new  44d31cb   TST + DOC + COSMIT refactor ParameterGrid and test it
       new  b36d221   ENH len on ParameterGrid and ParameterSampler
       new  d6e96dd   BUG deprecation of grid_scores_ in GridSearchCV
       new  f8aacaf   BUG always do cross-validation in GridSearchCV
       new  264a1f6   DOC fix clone and get_params documentation
       new  09abb18   TST grid search/randomized search on non-BaseEstimator
       new  a768b16   TST actual sparse input in sparse k-NN tests
       new  cdb9ef2   COSMIT prevent a copy in randomized LR
       new  ec0fc91   TST speed up comment tests by ~20%
       new  2909234   TST radius-neighbors regression test not entirely stable
       new  5b31612   BUG additive_chi2 missing in KERNEL_PARAMS
       new  66eb410   BUG + DOC fix Nystroem for other kernels than RBF
       new  5f6c0db   TST added test of fit and transform for kernels for nystroem
       new  1228490   example title change
       new  0016ea0   gallery effects,icon change,cleanups
       new  a71f616   typo fix and heading changes
       new  b0baab6   fix indentation error-cause lots of build warnings
       new  071d08d   4 thumbs per row/hover effect/some cleanup
       new  3b8d2ec   fix for iris dataset
       new  6dedeba   line_count sort added, some changes reverted
       new  6625317   move comment out of list
       new  c2bebbc   remove comment, undo change
       new  b41eae2   COSMIT rm repetitive __main__ blocks from tests
       new  ae70008   ENH allow additional kernels on KernelPCA
       new  1f6701b   TST fix broken doctest
       new  fc804b8   add median of 3 pivoting to quicksort
       new  cbd3e5e   P3K developer docs
       new  0cd16d1   PY3: Disable lib2to3
       new  3379504   PY3: fix urlopen in mldata and california housing loaders
       new  71d1a7c   PY3: used six.u to fix unicode variables in svmlight
       new  6d9ab96   PY3: fixed basestring in crossvalidation.py
       new  4d326b2   PY3: six.moves.cStringIO to fix StringIO import
       new  5993ef5   PY3: use b() convenience function for string literals
       new  5a5268b   PY3: fix remaining cStringIO imports
       new  2c9a815   PY3: ensuring file stream is read as binary
       new  9e93ba5   PY3: fix None comparison (when not in OS X) in test_k_means.py
       new  a81df3b   PY3: fix for string literals in datasets' test_base.py
       new  155c6ea   PY3: print function in coordinate descent doctest
       new  095085a   PY3: used six.moves.xrange to fix xrange
       new  63562b0   PY3: replaced izip
       new  966e0f1   PY3: convert string literal to bytes using six in cython file
       new  7da46e0   PY3: used six.iteritems to fix dict iteritems in module pipeline.py
       new  48cbd5e   replacing numpy array with range for use in random.sample
       new  4d1d8b9   PY3: changing None to 0 to ensure comparability in py3
       new  461b585   PY3 fixing utf8 comments in svm through try/except and six.b
       new  177b26b   PY3: forcing execution of map by using tosequence
       new  090ad85   PY3 fix comparison of ndarray and string
       new  58b96aa   DOC FIX: multi-target linear model attribute shapes
       new  5cc8032   ENH add normalize option to accuracy_score + FIX bug with 1d array
       new  e3dac0d   PY3 array.array wants str in Python 2.x and 3.x - give it a str
       new  154007b   fixing bug in linear_model.SGDClassifier for multi-class warm start
       new  9b66038   removing accidental space
       new  b556759   Merge branch 'pr/1790' -- Python 3 support from PyCon sprint
       new  3579a87   PY3: record is a kwarg argument for warnings.catch_warnings
       new  0863d3d   PY3: long is no longer a type in Python 3
       new  b75f2f5   add pydist code
       new  8d5d40e   fix binary tree sort bug
       new  473c716   add pydist: user-defined metric
       new  176063a   Update outdated comments in sklearn.hmm.
       new  6d10280   Merge pull request #1803 from kmike/hmm
       new  4f9a073   DOC update example path
       new  01fce76   add haversine distance
       new  b470370   add exception passing to C functions
       new  1faff8b   rename dist conversion funcs
       new  d9023c7   Merge pull request #1804 from AlexanderFabisch/fix_example_path
       new  fcf69bd   Cosmit: remove needless blank lines.
       new  bf0c1a3   Merge branch 'master' into pr/1798
       new  dbf1e54   rename example title
       new  79ea5fb   PY3: fix exception syntax in tests/test_common.py
       new  7bad6ae   PY3 fix test_cross_validation
       new  2af2c73   PY3 fix OneHotEncoder doctest ( "<type 'float'>" is "<class 'float'>" in Python 3.x)
       new  6d47ec0   Merge pull request #1812 from kmike/testing-fixes
       new  99f88db   make setup.py executable
       new  cb7ba3a   Merge pull request #1816 from ndawe/master
       new  b1a97de   DOC describe SVM probability calibration (and advise against it)
       new  5eb035c   DOC further comments on SVM probabilities
       new  0180188   ENH multiclass probability estimates for SGDClassifier
       new  b796520   BUG digits grid search was passing cv to the wrong method
       new  87dcc8b   DOC typos in grid search docstrings
       new  bc67dbf   ENH Added custom kernels to SpectralClustering
       new  46292a1   PY3 + TST decouple test_metrics from random module
       new  6a47c6f   BUG: Use Py_ssize_t to index into numpy arrays to help Python handle big data. Indent a few copy/pasted function declarations for consistency. Fixes #1466.
       new  4db660e   MISC: Update _tree.c with cython.
       new  0f1950c   BUG: Use ``Py_ssize_t`` in a few more places for strides. Add the c file again.
       new  83f67ee   ENH/FIX make best_estimator_'s predict functions available in parameter search
       new  364db7a   FIX make *SearchCV picklable
       new  a1d4d18   COSMIT pep8
       new  2dd99c0   DOC fix macports package name
       new  c87d45d   Merge pull request #1823 from erg/issue-1466
       new  972e7cf   BUG: ElasticNectCV choosing improper l1_ratio
       new  df74a8a   PY3 allow multiple base classes in six.with_metaclass
       new  eae18a3   PY3 fix metaclasses. See #1829.
       new  307afc8   DOC correct / simplify dbscan examle
       new  1201376   Merge pull request #1839 from amueller/dbscan_example
       new  382f74c   Merge pull request #1836 from kmike/master
       new  af674ac   get rid of ``rho`` in sgd documentation - has been replaced by ``l1_ratio``
       new  5ece0e1   Cosmit: more idiomatic way of clipping to zero.
       new  d3bc5ac   Demystify magic values in NNLS implementation.
       new  8a473b6   BUG: fix replacement for _neg.
       new  6f8da5c   Fix random state where appropriate.
       new  9cf9e9f   Fixx doctest.
       new  f21f64a   DOC: document attributes fitted by DictVectorizer.
       new  67ece78   COSMIT various typofixes
       new  2ad3448   Merge pull request #1852 from slattarini/typofixes
       new  5d90520   ENH: minor changes for numpy versions
       new  3c501ec   COSMIT typo. the French again ;)
       new  aff7089   FIX setting k in KMeans and MiniBatchKMeans was silently ignored. Left over in 07c56d7cd2ddfe71e7a4399d74fc367d6000d854 Damn, that was nasty :-/
       new  0d756f9   FIX Moved copying of labels in LeaveOneLabelOut and LeavePLabelOut to __init__.
       new  0372c20   TST Added test that checks if LeaveOneLabelOut and LeavePLabelOut work normally if the labels variable is changed before calling __iter__.
       new  bf3f601   DOC Fixed doc test to work with the fixed versions of LeaveOneLabelOut and LeavePLabelOut.
       new  9b1aabc   pep8
       new  79190e2   BUG Add lambda_ attribute to ARDRegression after fit
       new  ff0e882   ENH reduce size of files produced by dump_svmlight_file
       new  e084727   ENH: moved export_graphviz to sklearn/tree/export.py
       new  813d1a4   ENH: add max_depth to export_graphviz
       new  2b01f3a   ENH: output criterion name instead of "error" in export_graphviz
       new  919951a   ENH speed improvements in HMM
       new  164d36f   DOC distributions produced by hashing trick depend on input
       new  65bdff8   ENH Release GIL when entering LibSVM/Liblinear code
       new  81e0091   ENH actually use scikit-learn's PCA class in plot_pca_3d.py
       new  223f2e6   Implement correct d-dimensional kernel norms
       new  ade2db9   DOC Add labels and some explanation to confusion matrix example
       new  2f2f102   Switch off survey banner
       new  fc014c2   newline at end of file
       new  456b0d0   Merge pull request #1581 from jaquesgrobler/example_gallery_cleanup
       new  8824979   temp disable line-count-sort for gallery while fixing bug
       new  e1f5f42   DOC multiclass: typo and use case
       new  1a6b9e8   DOC: put feature extraction before pre-processing.
       new  8e9f452   sort-by-line-count bug fixed
       new  e019b6b   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  dbd3109   Update LIBSVM_CHANGES
       new  3444bfb   ENH Rewrote CountVectorizer fit_transform to be ~40% faster
       new  ed52df3   fix numbering for tutorials page
       new  a85c896   Add test for PatchExtractor (float value for max_patches)
       new  1816dee   Fix float value support for max_patches in PatchExtractor
       new  2611179   add metric mappings to dist_metrics
       new  75a7047   binary tree: make valid_metrics a class variable
       new  2ecf5e1   dist_metrics: allow callable metric
       new  55f409c   add chebyshev distance to kd tree
       new  b72e944   add functionality to NearestNeighbors estimators
       new  c808ec0   DOC PR means pull request
       new  a344f10   FIX BytesIO and urllib usage in fetch_olivetti_faces
       new  4ea8fa6   Roger-Stanimoto -> Rogers-Tanimoto
       new  12f2292   Update hmm.rst
       new  9cfb34c   Add bit more instruction on writing docs
       new  b6b0713   big O/tilde add in
       new  d448a2a   removed old complexity info
       new  16e682e   DOC: remove typo
       new  1d3e480   image and html file added
       new  a999972   link fixes
       new  574b9ed   add further links
       new  d6d2a5b   last links fixed
       new  467c904   jquerys added
       new  6115114   FIX gen_rst.py was something using an undefined variable
       new  e6ed507   intigrated to tutorial index
       new  8238a86   update tutorial page
       new  8df53e1   make links relative
       new  6261c4b   rename image/html
       new  707887d   add instructions for editing Readme, and script needed for that
       new  a39b547   Merge pull request #1886 from NelleV/DOX_fix
       new  d562b46   ENH refactor and further speed up CountVectorizer
       new  d88c38b   DOC I didn't mean soft-O by "tilde notation"
       new  1e6e337   DOC: libatlas3-base in requirement
       new  021f515   remove svg2html script,toctree section added,doc page for ml_map created
       new  8ef17f7   Release GIL around sparse liblinear training
       new  a11abb5   calculate kernel norm only once
       new  44f3944   compute kernel norm only once
       new  34f7cdc   TST: compare gaussian KDE against scipy version
       new  ed643e1   sidebar added
       new  cd9cffa   layout fixes and top paragraph
       new  a4071f9   TYPO fix
       new  72cce4c   ridge regression uses compute_class_weight()
       new  3df7309   Re-add deprecated class_weight parameter.
       new  1bf52ec   removed class_weight parameter from RidgeClassifier.fit()
       new  ecbb563   Change dual splits to single splits in query_dual
       new  29b10e8   stochastic_gradient: fix mistake in _init_t docstring
       new  0f00baf   stochastic_gradient: describe all losses, fix epsilon description
       new  4a98b3d   Merge pull request #1893 from dougalsutherland/sgd-docs
       new  6a489cc   ENH: Avoid computations in ElasticNetCV
       new  cf0311c   ENH: improve memory usage in ElasticNetCV
       new  850a8f9   DOC: docstring of private functions
       new  3376704   BUG: fix sparse support in ElasticNetCV DOC: fix docstring: sparse matrices are accepted
       new  341f6f7   COSMIT: address @agramfort's comments
       new  6224c4f   DOC describe API, not internals, for AdaBoost
       new  d0e2281   DOC replace "arithmetical order" in AdaBoost docs
       new  e5cc09d   Added random_state=0 for AdaBoostRegressor
       new  10ee464   TST strengthen AdaBoost tests
       new  719ca98   FIX SVR complaining about a single class in the input
       new  58e8789   COSMIT do np.unique(y) once in SVC
       new  11902ae   update what`s new
       new  fcf911c   deleted unnecessary thumbnail
       new  6a058aa   check_pairwise_arrays() preserves dtype==numpy.float32 (squashed)
       new  f1df9ab   Replaced 'for i' with 'for _' at place where i is not used. fix pep8 for weighted_boosting.py
       new  465e053   COSMIT pep8
       new  6b47d76   DOC improve description of cross validation
       new  c0f7b89   DOC rewrite description of k-fold CV
       new  c04aec9   DOC: add random_state parameter to StratifiedShuffleSplit doc string
       new  f7ad2bc   Merge pull request #1907 from aflaxman/stratified_shuffle_split_rand_state_doc_str
       new  0225c36   ENH speed up TfidfTransformer using spdiags
       new  e9a1410   mailmap entry for @lqdc
       new  05a7766   resized image
       new  64e9290   REFACTOR combine train_wrap and csr_train_wrap
       new  8615aba   Added sponsors to the about.rst page
       new  bdc410b   Spelling mistake
       new  fd052ff   DOC add 2012 GSOC students
       new  cff20f6   DOC define validation before cross validation
       new  f924d3e   COSMIT: labels in plot_lasso_coordinate_descent_path
       new  dacfd8b   DOC: Replaced all BSD style licenses with "BSD 3 clause"
       new  c8a5574   DOC typos in cross-validation description
       new  4b6a13d   Extended test_confusion_matrix_binary to incorporate non-integer labels
       new  5b185f5   Extended test_confusion_matrix_multiclass to incorporate non-integer labels
       new  c15f141   BUG: Fix for non-integer datatypes in confusion_matrix
       new  c33c04b   ENH: faster preallocation and integer type for the accumulators
       new  ddaa494   clean up mailmap/deduplicate contributors
       new  0859653   FIX: use the mldata mock in docstring as well
       new  5d51e42   STY: one-lined lines that where less than 79
       new  a5b978c   Merge pull request #1913 from Jim-Holmstroem/refactored_precision_recall_fscore_support_to_count_with_integer_type
       new  922df5a   Add SVD-based solver to ridge regression.
       new  9c84263   Remove unnecessary code in ridge svd
       new  68d0bc5   disable sidebar using cookies to remember last position
       new  daf5277   BUG disable memory-blowing SVD for sparse input in RidgeCV
       new  a8fe355   BUG: solver was not passed to computational method in Ridge object
       new  34022b5   FIX scores calculation in ovo multiclass
       new  2ffb484   COSMIT pep8
       new  83ed57d   FIX jenkins error on numpy 1.3.0
       new  7fff871   Merge pull request #1884 from jaquesgrobler/ml_map
       new  1dc2a26   DOC documented n_init parameter of MiniBatchKMeans. Closes #1900.
       new  fb89031   FIX broken scorer, add non-regression test.
       new  e9a9592   FIX WARN about **params being not used in GridSearchCV.fit. Closes #1815.
       new  e163f8b   FIX DictVectorizer behavior on empty X and empty samples
       new  e4b9fd2   FIX bug in callable kernel decision function - Sorry, I think that was me.
       new  b2082b5   DOC added link to scipy lecture notes to tuts
       new  e94b774   FIX: restore numpy 1.3.0 compat with np.divide fix
       new  8582a7a   MAINT: let the result type be infered by coo_matrix, possible since np.ones already integer typed
       new  8c7b900   FIX test error in test common for KernelPCA that doesn't respect its n_components.
       new  26a86bd   DOC return_path argument, prettier references
       new  ff884e0   Merge pull request #1924 from jaquesgrobler/FIX_sidebar_on_index_page
       new  0a2dc13   Merge pull request #1911 from Jim-Holmstroem/generalize_label_type_for_confusion_matrix
       new  ff69cc5   ENH more pythonic way to treat list of list of labels
       new  be842f2   ENH add jaccard similarity score metrics
       new  1c9a6c9   COSMIT: txt -> rst
       new  c80ca91   FIX compatibility issue with np 1.3 py 2.6
       new  7af4fd9   ENH call asarray on returned scores and pvalues
       new  40cf5c9   TST ensure SelectKBest and SelectPercentile scores are best
       new  0480a66   FIX ensure SelectPercentile only removes tied features in case of ties
       new  313193f   ENH _BaseFilter.inverse_transform should respect dtype
       new  14f22a9   DOC Fix comment for _BaseFilter.inverse_transform
       new  4fd975c   ENH sparse _BaseFilter.inverse_transform
       new  35b4045   FIXTST fix errors introduced to feature selection tests
       new  1b07773   DOC comment feature selection sparse inverse_transform
       new  a9b93a2   DOC: latex beautification
       new  9663654   DOC: latex beautification
       new  a7e5eb9   DOC: cosmit - fix latex typo
       new  84b2326   Merge pull request #1944 from jnothman/selectpercentile_limit_bug
       new  2b57e4e   Merge pull request #1935 from jnothman/base_filter_inv_transform
       new  489d294   ENH Feature selection should use CSC matrices
       new  aaaa95c   COSMIT Remove redundant code in CountVectorizer
       new  5d8a4d9   TST test CountVectorizer.stop_words_ value
       new  735a814   Use Cholesky solver by default, but use SVD as fallback
       new  65b1924   fixed typo
       new  0fc812b   COSMIT: better notation in CountVectorizer.
       new  dc38e6b   COSMIT: same changes in transform method.
       new  dcebf9e   COSMIT: more robust condition in inverse_transform.
       new  9589e04   FIX typo in test for RdigeCV
       new  235ee25   ENH Use csr_matrix.sum_duplicates instead of tocoo
       new  0cf8e3d   Use ValueError for non-existant solvers
       new  64760f2   Import gzip and bz2 only if necessary.
       new  84ceb74   DOC typo in RandomizedSearchCV docstring
       new  d0f9cab   DOC small typographical fixes in grid_search documentation
       new  8e9719d   COSMIT refactor roc_curve and precision_recall_curve
       new  6736a21   FIX bug where hinge_loss(..., neg_label=1) produced incorrect results
       new  cb701b0   MAINT: refactored metrics.auc to use np.trapz
       new  d6e9598   Merge pull request #1880 from NicolasTr/patch_extractor_float_max_patches
       new  928541c   Fixed precompute issue (again) in ElasticNet and enet_path
       new  8c71062   Fix setup.py to resolve numpy requirement during pre-install. This should allow 'pip install numpy cython scipy scikit-learn' to install scikit.
       new  4e2bb00   ENH: add pre_dispatch option to cross_val_score
       new  890b11a   DOC Fix estimator unsupervised fit method signature
       new  962e2e1   maintenance scripts added for machine learning maps - needed for modifying the map in future
       new  07c8bf7   WMinkowskiDistance corrections to error messages and docstring improvement
       new  d78f43a   Merge pull request #7 from jhale/new_ball_tree
       new  8e79203   DOC rename lambda to alpha in plot_lasso_coordinate_descent_path. (Re)-Closes #903.
       new  f3107ad   DOC clarification of parameter search
       new  7c16dad   initial commit
       new  ae7adc8   AdaBoost: use estimator weights in predict_proba
       new  ac38016   TST + DOC AdaBoostClassifier.predict_proba fix
       new  91111d8   COSMIT refactor AdaBoost code
       new  b0f1d46   DOC fetch_20newsgroups returns the text, not text files. see SO question: http://stackoverflow.com/questions/16615523/using-scikits-kmeans-to-cluster-ones-own-documents
       new  76bb083   DOC Fixed documentation of kernel parameters: sigm uses gamma, but not degree. Closes #1972.
       new  d33634d   ENH add multilabel support to PRF metric family
       new  096c8a0   DOC fix typos
       new  3483cf5   COSMIT shorten long line for pep8
       new  9e4518b   Upload to unstable, move joblib to Depends from Recommends (Closes: #709056)
       new  f180328   ignore PDFs
       new  9a64bc1   DOC move old tutorial out of the way for merge
       new  fda420e   Merge branch 'tutorial'
       new  9a8845e   ENH Create FeatureSelectionMixin for shared [inverse_]transform code
       new  7ebfd57   DOC rewrite descriptions of P/R/F averages and define support
       new  bb6e65f   DOC/COSMIT fix typos in What's New
       new  5b0afce   DOC add some contributions to What's New
       new  8668d02   BUG: set random state in LogisticRegression
       new  27cae1a   DOC clarification in Scoring objects: Its not a good sign if I don't understand my own wording.
       new  89f29ba   DOC Fix references to missing examples
       new  901e9d7   fix incorrect reference
       new  b84eb30   Merge pull request #1986 from jaquesgrobler/DOC_reference_fixes
       new  331738f   remove placehoder
       new  d8da862   Merge branch 'tutorial' into DOC_olivier_tut
       new  89ba448   remove duplicates for merging larsmans branch
       new  d69566c   initial tutorial re-added with it`s commit history
       new  820aab5   Placeholder to tutorials page
       new  b647705   TST Use assert_almost_equal in test_symmetry
       new  a13437f   COSMIT prefer partial over lambda in test_metrics
       new  4b5ae71   TSTFIX use name, not metric, in test_metrics error messages
       new  dafc663   DOC correct note on handling 0-denominator in P/R/F
       new  623d52a   add optional banner to index page to advertise code sprints
       new  40adb7a   link updated
       new  531b2d8   Changed the setup instructions in the README to properly install the package in the user home.
       new  9eec732   Merge pull request #1998 from kgeis/fix-setup-instruction
       new  b7e9ee6   Merge pull request #1996 from jaquesgrobler/DOC_sprint_sponser_banner
       new  752e9aa   added first version of out-of-core example
       new  161d19b   ENH: avoid MemoryError on manhattan_distances
       new  c649440   BUG: old versions of numpy
       new  6204ed4   BUG: old versions of numpy
       new  3931951   Corrected a few things on the Mutual Information doc pages.
       new  c3cac4b   moved tutorial out of tutorial folder
       new  f2e065d   tutorial index initial touchups
       new  402f53e   TST Fixed test_pipeline_methods_preprocessing_svm: pca was unused
       new  7b46a4f   revision round #1 (move to examples/applications, 1 file, auto-download dataset)
       new  8f85b0f   Merge pull request #2005 from kmike/test_pipeline_methods_preprocessing_svm
       new  85811ce   Fixed typo in metrics.py
       new  ecf135a   ENH faster unique_labels for big sequences of sequences
       new  38c31d4   Fix L2 regularization order in sgd_fast
       new  b0449b6   DOC explain labels parameter to confusion_matrix
       new  724652e   corrected doctests after moving L2 penalty application in SGD
       new  c0abef1   hover removed from nature, jquery more recent version, containerexpansion on mouseover add
       new  de7c758   image resizing added
       new  884306c   Zoom bug fixed
       new  c068493   added docstring space to popup block
       new  15c8a5e   docstrings embedded into example hovers
       new  194fa35   ENH speed up sklearn.feature_selection.chi2
       new  6c7d0ba   Merge remote-tracking branch 'upstream/master' into pr/2016
       new  7840620   added SGD L2 fix to whatsnew
       new  f1dee6e   fix: add missing str formatting operator
       new  4d4a288   Final visual effects added to hovering
       new  c417837   Nelle`s review fixes addressed
       new  01c4705   Cross browser shadows covered
       new  b036dbc   remove forgotten print
       new  a196462   DOC dependency installation with yum (Red Hat, CentOS)
       new  b486875   FIX bug (swapped args) in chi2
       new  1bee1b8   Merge tag '0.13.1' into releases
       new  9ad265e   Merge branch 'releases' into dfsg
       new  bce5737   Merge commit '0.13.1-178-g9ad265e' into debian
       new  efba14e   "Fresh" bugfix release: dropping all cherry picked changesets
       new  bc2acb5   FIX yet another chi2 bug
       new  eb755db   ENH remove pos_label argument with multilabel binary indicator format
       new  71f3f38   ENH remove warnings at testing time
       new  4a4c2cf   Documentation ERROR: mixture.DPGMM.precs_
       new  de560bf   Move balance_weights out of preprocessing.
       new  71db3d4   Add categorical_features option to OneHotEncoder.
       new  4c4f1cf   Minimal spanning tree backported from scipy 0.13
       new  8ad6b7b   Added test
       new  67cdef8   Moved mst to a subfolder and added a README file
       new  81ad2a7   Added new files (from previous commit)
       new  e8d3460   Support both masks and arrays of indices.
       new  aff9276   shorten displayed dosctring to 95 chars
       new  6e2aee2   Typo.
       new  ff8c77c   DOC Detail on parent-child relationship in tree
       new  a24a29d   Merge pull request #2031 from jnothman/tree_comments
       new  017afab   Rename _apply_transform to _transform_selected and make it a function rather than a method.
       new  fe03357   FIX #2032, FIX #2033: ensure module names consistency with __all__
       new  82323a3   Remove redundant test that was checked in by mistake
       new  b9d14f9   fix white space inconsistency between header and docstring
       new  9ad829b   Merge pull request #1914 from fabianp/ridge_svd
       new  6fc3499   example docstring fixes
       new  c453118   logistic regresion example fix
       new  4b15cb5   Tutorial Setup
       new  fd3bd2d   FIX/COSMIT helper to identify target types
       new  210cefa   FIX cannot use set notation for Py2.6
       new  785c596   FIX need explicit dtype for array of sequences in numpy 1.3
       new  65805de   COSMIT remove redundant target size check
       new  1b3b0c2   FIX numpy 1.3 has no float16; use float32
       new  48a87b5   FIX/TST np.squeeze in numpy1.3 fails with array of sequences
       new  bf15dcf   FIX numpy 1.3 throws error with array of arrays
       new  5d7aaae   FIX inconsistent cv_scores_ generation for randomized search and re-add example
       new  c012d4f   FIX use Python 2.6-compatible str.format
       new  9d9ea41   pep8 / pep257 compliant formating
       new  e2b46c4   get rif of feature dicts, leverage HashingVectorizer class directly
       new  a8ed40e   plot as both a function of time and n_examples
       new  ecd7792   using print() function
       new  ddd0560   improve explanations on out-of-core learning paradigm
       new  9dec5b8   improve explanations on example structure
       new  5fe68bc   fixed use of docstrings + added section in whats_new.rst + added data dir to .gitignore
       new  c6d6558   more robust data location
       new  5afc1c9   COSMIT refactor cross-validation strategies
       new  9f17401   Include LeavePLabelOut in refactoring
       new  721be7f   A further refactor
       new  ce74daf   COSMIT Base class for KFold/StratifiedKFold validation
       new  2278764   COSMIT make BaseKFold abstract
       new  8b761b8   COSMIT pep8 in cross_val_score
       new  de18f40   COSMIT Base class for [Stratified]ShuffleSplit
       new  ab03f01   DOC clarify LeavePOut's combinatoric explosion
       new  524e5ad   DOC similar note in narrative docs
       new  246938f   DOC More explicit note
       new  b06f08f   use same, separate held-out data to estimate accuracy after each mini-batch
       new  cb91680   Merge pull request #2056 from jnothman/leavepout_clarify
       new  cacb14e   ENH add latent semantic analysis/sparse truncated SVD
       new  5ab4b4e   ENH use rnd SVD in TruncatedSVD by default for speed
       new  34815a3   COSMIT omit unused parameter/return value in svd_flip
       new  8202df5   TST strengthen TruncatedSVD tests
       new  8675dcb   DOC + MAINT deprecate RandomizedPCA scipy.sparse support
       new  a7e0246   FIX and link LSA clustering example
       new  52a1002   DOC explain normalization in LSA KMeans example
       new  df889f8   Merge pull request #1716 from larsmans/truncated-svd
       new  ebb6367   ENH: Added input checks in confusion_matrix
       new  f1e9116   DOC fix docstring headings
       new  a73fd1a   ENH: removed leftover condition to get a wider application of the import all consistency check
       new  347dc7c   added first version of out-of-core example
       new  89e6b0d   revision round #1 (move to examples/applications, 1 file, auto-download dataset)
       new  bfa939d   pep8 / pep257 compliant formating
       new  d392666   get rif of feature dicts, leverage HashingVectorizer class directly
       new  956aa39   plot as both a function of time and n_examples
       new  0853c68   using print() function
       new  d16aa68   improve explanations on out-of-core learning paradigm
       new  775bfc1   improve explanations on example structure
       new  d8cf157   fixed use of docstrings + added section in whats_new.rst + added data dir to .gitignore
       new  9a58f0a   more robust data location
       new  03e9dbe   use same, separate held-out data to estimate accuracy after each mini-batch
       new  ef872fb   FIX metrics/scoring bug with LeaveOneOut CV
       new  c849b32   MAINT remove deprecated gprime handling from FastICA + refactoring
       new  0db0d05   COSMIT make helpers private with underscore
       new  2d8842e   COSMIT make BaseKFold private with underscore
       new  ebff9a3   TST additional tests for preprocessing.Binarizer
       new  7bc5d1a   Merge pull request #2055 from jnothman/cv_refactor
       new  1324f45   fixed conflict in whats_new.rst
       new  d8eaed6   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn into out-of-core-examples
       new  d63535e   factorized instance extraction + plots
       new  8ed796e   added note on test set creation rationale
       new  b20accb   cosmit : inline extract_instance
       new  80e1d27   Enforce n_folds >= 2 for k-fold cross-validation
       new  fd3a80f   Merge pull request #2067 from jnothman/test_binarizer
       new  febe9be   COSMIT add underscore prefixes where forgotten in cross_validation
       new  11023d1   Fix wrong argument name in RFECV docstring
       new  517ceac   Merge pull request #2071 from djv/patch-1
       new  1880e22   COSMIT much simpler agglomeration inverse_transform
       new  8c42565   TST stronger test for agglomeration transforms
       new  cd9dbbc   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn into out-of-core-examples
       new  9dc2b46   more structured iteration using islice + wrappers; renamed chunk for minibatch as the latter seems more common in hte literature
       new  24acc5e   Merge pull request #2075 from jnothman/agglomeration_simplify
       new  4063d9e   FIX : use unique from fixes
       new  233c7dc   enhanced (hopefully) DBScan documentation; killed some whitespace along the way...
       new  f84e646   Merge remote-tracking branch 'upstream/master' into dbscan-doc-enh
       new  d0b6433   DOC minor fixes to Ward docstrings
       new  2245907   DOC fix docstrings for AgglomerationTransform
       new  3d2a44c   DOC detail Ward.children_ and fix n_components_ type
       new  8d5303c   DOC comment on Ward algorithm
       new  9abb016   DOC clean pooling_func arg type
       new  7ead6c9   Merge pull request #2074 from jnothman/ward_docstring
       new  176c9ce   DOC no more mention of the Bunch in the narrative docs
       new  f102bac   DOC copy comment describing hierarchical clustering children
       new  54540ad   Merge pull request #2076 from pprett/dbscan-doc-enh
       new  e31c4a7   added sub section on out-of-core scaling in the narrative docs
       new  e89ecd8   FIX don't rely on Bunch behavior with fetch_covtype
       new  7063ad2   DOC much more readable formula in chi2 kernel doc
       new  3139f1d   Remove invalid todo comment
       new  9c7817b   Merge pull request #2080 from ahojnnes/dist-todo
       new  aa66b62   Merge pull request #2054 from ogrisel/invalid-n-folds
       new  05bfd4d   Merge branch 'master' of git://github.com/scikit-learn/scikit-learn into out-of-core-examples
       new  5fcc104   some more language corrections
       new  77402c2   more pep257 fixes (not for ReuterStreamReader as it is not really the interesting class here)
       new  d2fe5dc   Merge pull request #2004 from oddskool/out-of-core-examples
       new  c2e025d   fix: needs_threshold not plural in repr
       new  bba4225   DOC fix some docstring/parameter list mismatches
       new  558cf06   DOC fix RandomizedPCA docstring for n_components=None
       new  0c3ae9c   ENH allow empty grid in ParameterGrid
       new  006ec19   MAINT ignore kernprof.py reports
       new  22e1f69   BUG delete/delete[] error in Liblinear
       new  0efd767   DOC fix some docstring/parameter list mismatches
       new  fe2aad9   DOC ParameterGrid on lists
       new  a7d34ca   Merge pull request #2082 from larsmans/empty-parameter-grid
       new  873c6f3   FIX avoid spectral_embedding naming conflict
       new  71f7d2c   FIX : missing y=None in FactorAnalysis
       new  434ca77   Merge pull request #2085 from agramfort/fix_y_score_fa
       new  0714468   Add missing doc string printing for examples
       new  23c55da   Merge pull request #2087 from ahojnnes/examples-print-doc
       new  54931e3   renamed weight to sample_weight in sklearn/isotonic.py
       new  8f78ab8   Merge pull request #2090 from kanielc/fix_weight
       new  c540b30   FIX: make doc auto-linking support any Unicode / UTF-8 content
       new  6d4914d   Make the out-of-core example plot work when launched by the sphinx extension
       new  6bf4b8e   COSMIT move deprecated parameter to end
       new  f73fbde   firefox bug fixed
       new  054fee9    classifiers comparison fix
       new  636f793   DOC fix V-measure docstring
       new  63df99d   FIX: do not print to many messages to stdout when generating the documentation
       new  85e4a5f   Merge branch 'master' of github.com:scikit-learn/scikit-learn into select_categorical
       new  eb19f89   Removed pl.axis('tight') and set the plot limits with pl.xlim(), pl.ylim(). pl.axis('tight') appears to be adding whitespace around the colormesh
       new  67487b6   add notes on implementation details to binary_tree.pxi
       new  7f9b783   remove scipy cKDTree support from neighbors
       new  351ca97   add neighbors module changes to whats_new
       new  7c4126e   Merge pull request #2104 from kastnerkyle/master
       new  e11873f   Address @jnothman's comments.
       new  8861833   DOC spellfixes
       new  d40caea   DOC spelling and clarification
       new  6feda82   BUG: fix precision issues in kernel_density; remove buggy dual-tree KDE versions
       new  df98f79   add KDE Estimator class
       new  cdf6b36   Change 'charse_error' to 'charset_error' in load_files.
       new  474d3a0   Revise documentation about handling text and bytes.
       new  b7b4f93   Add a documentation section about decoding text.
       new  b9877a7   Move the new "Decoding text files" doc section
       new  49bcd9c   DOC fix in the hierarchical clustering
       new  8043403   Merge pull request #2118 from NelleV/DOC_fix
       new  57b2a96   DOC Acknowledge sponsors for the Paris sprint
       new  fc7aeee   DOC missing stuff in randomized_l1 module
       new  c01ed96   PY3 xrange, np.divide, string.uppsercase, None comparison
       new  60d39a2   MAINT dedup Clay Woolam's contribs (>100 commits!)
       new  d270249   FIX/ENH mean shift clustering
       new  35260c1   DOC typo
       new  9c162a8   ENH micro-optimize RFECV
       new  dc28e23   FIX mbkmeans benchmark bug (k instead of n_clusters)
       new  7419d2c   FIX off-by-one error in neighbors benchmark
       new  ca2142e   ENH lots of benchmarks fixes
       new  a115e91   COSMIT refactor LibSVM wrapper for safety and readability
       new  6fc6d10   add kwargs to PyFuncDistance
       new  5423477   DOC fix some broken URLs
       new  fbdc2a4   PY3: New test for the get_params handling of deprecated attributes.
       new  a58ceed   DOC: Clarify docs on preprocessing.Binarizer.
       new  ed204ea   FIX charset -> encoding in load_files
       new  2e812d9   DOC typo
       new  78a80ce   Revert "FIX charset -> encoding in load_files"
       new  b70953d   FIX verbose output from k-means
       new  1327057   FIX remove params from RandomizedSearchCV
       new  1556d8a   ridge multi target with individual penalties written. To be tested
       new  0913d85   old tests passing
       new  1efc49a   new multiple target tests added, functionality confined to direct usage of ridge_regression function
       new  2ca4fb2   Ridge estimator works with individual penalties
       new  a92ae2d   test for ridge estimator
       new  388bfa7   ridge doc string
       new  066a491   ValueError for wrong shaped input instead of assertion failure, in order for sklearn/tests/test_common.py, line 238 to pass
       new  1f45378   docstring in Ridge estimator
       new  20b8bc6   added individual penalties function for all other solvers. Tests passing for all of them
       new  0f11642   always make alpha into an array
       new  92418d6   updated tests
       new  9c51438   tests passing
       new  4ce2038   removed elaborate testing in ridge.fit, not necessary anymore
       new  371dea2   Test exception is raison when number of targets and penalties don't match.
       new  547c8f5   Simplify ridge solvers (ongoing work).
       new  3c91661   Extract sparse_cg and lsqr solvers.
       new  38463c1   Extract dense_cholesky solver (linear case).
       new  4d8f5e8   Extract dense_cholesky solver (kernel case).
       new  52b972e   Clean up.
       new  d4494a0   Extract SVD-based solver.
       new  e850ffd   Clean ups.
       new  e83056f   Remove copy option.
       new  e0b26bc   Cosmit in docstring.
       new  1dba0d3   What's new.
       new  5852998   Remove if statement.
       new  d461db2   simplified _solve_svd
       new  c09d90f   Cosmit.
       new  b9e94eb   Fix failures in grid search.
       new  41b1f9a   FIX charset -> encoding in load_files
       new  436c405   FIX search bug introduced in 1327057f4258f41712ecab5c94770aac5ff01982
       new  e9b9896   FIX inconsistent attributes shapes in naive Bayes
       new  9a452d5   Do not set sample_weights unless need to.
       new  ebc0634   Add warning when fall back to other solver.
       new  61f2310   FIX test failure in naive Bayes
       new  7cff2e0   FIX failing doctest for CountVectorizer
       new  6e1ca5d   Remove unused variable.
       new  163a68e   Fix failure in svd-based ridge solver w/ old numpy.
       new  d420aaf   TST + PY3 various fixes
       new  4fe51ba   Merge pull request #2027 from mblondel/select_categorical
       new  d25f12b   FIX copy in OneHotEncoder and _transform_selected
       new  e0b46f8   DOC: document the new neighbors functions & KDE
       new  c7c523f   undo change to clustering example
       new  eaac7aa   fix conflicts with master
       new  7725cde   import KernelDensity from neighbors module
       new  3be1e6c   adjust math formatting in neighbors docs
       new  518c289   fix NearestNeighbors to pass common tests
       new  60ddef9   add KernelDensity to class list
       new  cb38853   set random seed in KDE example
       new  c7dc28b   skip KDE test to prevent failure due to older SciPy versions
       new  9b5efbc   fix typo: SkipTe -> SkipTest
       new  7409639   fix doctest in neighbors
       new  ebdc90e   BUG: return proper algorithm in KDE
       new  8c1a75e   add species KDE example
       new  bd6cec3   PEP8: neighbors module
       new  c7548cc   DOC: rearrange KDE examples
       new  3778bf8   TST: increase test coverage in neighbors module
       new  86ed8b7   DOC: pep8 & formatting in neighbors docs
       new  7b762d8   DOC: make doc tests pass
       new  85ec129   DOC fix docstring typos in cluster/mean_shift_
       new  55bdefc   Merge pull request #2135 from fhs/meanshift-doc
       new  e40a7f5   FIX unique_labels in corner case
       new  e3ca5c1   FIX issue with comparable but different dtype
       new  adcb0a8   ENH don't allow mix of input multilabel format
       new  825b2c0   ENH simpler check for mix of string and number input
       new  45847a7   COSMIT better name
       new  46d4718   Typo
       new  91e4b9a   ENH use type_of_target within unique_labels
       new  61d9f41   ENH improve documentation with allowed label types
       new  aef47a7   ENH check that we don't mix number and strings
       new  7856a72   Flatten label type checking
       new  33016ed   TST add smoke test for all supported format
       new  6c68cac   COSMIT
       new  7699100   PY3K use six.string_type
       new  109f2fb   OPTIM + ENH simplify mix string and number check
       new  a9197f3   FIX bug with indicator format
       new  a62abe6   ENH use a comprehension over imap
       new  eff660b   ENH optimize KMeans for sparse inputs
       new  6a6c2e9   FIX KMeans bug; argsort result apparently not always C-contiguous
       new  8371647   DOC what's new: faster KMeans
       new  c2784e7   DOC fixed small mistakes in the pls module
       new  7be8a1e   Merge pull request #2138 from NelleV/kCCA
       new  5482a73   DOC more explicit description of degree param on SVMs
       new  9f30384   Better status for the Py3 port
       new  d51b5ab   @arjoly and @glouppe thanks their funding FNRS and DYSCO
       new  90f3f58   Adding covariance regularization to QDA
       new  43fda22   Update qda.py
       new  c6dfe7e   Update qda.py
       new  a2f2d6d   Missed a space!
       new  7450058   Merge pull request #2140 from arjoly/ajoly-glouppe-sponsor
       new  49de26a   Updating to ensure pep8 compliaance
       new  9abb39f   reg_param is a float
       new  445db73   Update qda.py
       new  7b067b5   Update test_qda.py
       new  d147ad5   Update qda.py
       new  eee3c7c   Merge pull request #2142 from sergeyf/master
       new  09a64c5   DOC fix small mistakes
       new  bc54f01   DOC fixed some formatting in kernel approximation
       new  60b66a2   DOC fixed some formatting in the multiclass module
       new  765ac5d   Merge pull request #2145 from NelleV/kCCA
       new  fb10548   clearer decision surface plots and classifier final predictions for the ensembles
       new  d041aed   Donate buttons added `About us` and front page
       new  fe52584   Got all the doc-tests working
       new  4e1bc3c   improved formatting
       new  c619e7c   Merge in master
       new  def6c35   More python3 fixes (and just plain bugs)
       new  14f8fdb   use ELLIPSIS in doctest to deal with numpy changes.
       new  ef2cff1   Forcing the deprecation warnings to happen while in get_params.
       new  73e0492   Force warning to be heeded in deprecated args check. Possibly fixed a test bug (but maybe I just got it wrong)
       new  91ececa   Make a test not dictionary order dependent.
       new  61226a2   Fix up last doc tests.
       new  f1a40a4   Make the fixes 2.6 compatible
       new  e27743f   ELLIPSIS around a unicode issue.
       new  964a530   Fix y vector. We wanted round off division so that y == [0 0 1 1 2 2 ...], not [0 .5 1 1.5...]
       new  777f57a   A little more of those unicode helpers
       new  2bd695a   Another ELLIPSIS
       new  42af130   add 1D KDE example
       new  be0ba9d   DOC: small fixes to neighbors doc
       new  0fcd0ba   DOC: move KDE discussion to separate page
       new  162934e   add some notes and doc strings to neighbors cython code
       new  77695ee   COSMIT pep8
       new  3840fec   ENH order *does* matter for sparse matrices
       new  e53a2d5   donations paragraphs added
       new  43c55b5   FIX get rid of the last few asanyarray calls
       new  c942bcf   DOC fix erroneous docstring on preprocessing._transform_selected.
       new  331c450   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  8b9b4e4   BUG: replace elif by if in Ridge solver selection.
       new  daa3445   MISC: details about the donations
       new  0a549bf   Merge pull request #2146 from ianozsvald/clearer_iris_decision_surfaces
       new  958de49   add more documentation to ball tree and kd tree
       new  a42e4f1   DOC: tweak kde examples and move density docs
       new  ff11845   BUG: fix tophat sampling in KDE
       new  db888b3   Xplot -> X_plot
       new  814ad9b   Test for singular matrices in Ridge regression
       new  91e880f   MAINT: dedup @jakevdp and @jnothman in mailmap
       new  aeeda9f   COSMIT refactor document frequency implementations
       new  c3f9739   ENH print number of fits in BaseSearchCV._fit
       new  5bff742   COSMIT simplify printing of number of fits in grid search
       new  3e6121e   COSMIT fix a docstring in feature_extraction.text
       new  6121367   Pop off the recently added filter after testing for deprecation warnings.
       new  cfe0eae   merge in origin
       new  23a1a56   Comment change
       new  7697a6f   Merge more Py3 fixes
       new  16e971c   PY3: refcounting change introduced a regression on the use of resize in LARS
       new  ae9f7ef   BUG: type conversion in spectral_embedding
       new  3001fb6   bt->tree; dm->dist_metric
       new  cad8629   Additional implementation notes in binary tree
       new  40f4e94   MISC: remove unused imports
       new  b9673ab   example yahoo stock issue fix
       new  9c7f446   Fix two remaining python3 bugs.
       new  0dc094c   BUG: restore Python 2.6
       new  1f8ed89   COSMIT: two empty lines between functions
       new  9bd98c3   Merge branch 'pr_1732'
       new  98151cb   P3K developer docs
       new  b0d2649   FIX: pep8 and Py3 support in sklearn.neighbors.base
       new  00af6a2   FIX: Python 3 support for the neighbors doctests
       new  4fc42b9   FIX: pep8 + Py3 fixes in test_dist_metrics
       new  7f0803c   FIX: pep8 and Py3 support in sklearn.neighbors.dist_metrics
       new  4938dec   FIX: Py3 / pep8 fixes in test_ball_tree / test_kd_tree
       new  22dbecc   Update Python 3 support status
       new  e433d20   FIX use float64 in metrics.r2_score() to prevent overflow
       new  1b9ad77   TST r2_score float32 overflow fix
       new  d1bc252   Revert "TST r2_score float32 overflow fix"
       new  0bf3988   updated docs to fix formatting errors
       new  c490691   Merge pull request #2163 from ianozsvald/fix_plot_forest_iris_docs
       new  5190583   PY3 use urllib2 or urllib.request, based on Py2/3
       new  6713132   DOC let OneHotEncoder, DictVectorizer and FeatureHasher refer to each other
       new  1a8b797   DOC correct class_weight description for LogisticRegression
       new  376c570   BUG: use correct algorithm for callable metric
       new  7d489f3   TST: set random state in callable_metric test
       new  a7dc0fb   Style
       new  93deda5   More readable condition and more precise error message
       new  9c2ec56   FIX memory usage in DictVectorizer.fit
       new  bbf74a9   DOC/FIX twenty_newsgroups.rst should use TfidfVectorizer
       new  2fbd66c   Traversal in and tested. Next step is to remove references to old code
       new  1d4d664   Removed reference from spectral_clustering to old csgraph
       new  e30103a   csgraph updated from hierarchical.py
       new  09b6ce9   Removed actual _csgraph file, tests still all pass
       new  9259c39   Turns out sparsetools wasn't needed either
       new  3ebd9ee   Missed a spot
       new  4f125c7   Reference to graph components updated in dev docs
       new  baae05e   Two more spots. I think that's it
       new  4e1e1ab   Now that the folder has more than just mst in it, rename to sparsetools, which should help with referencing it.
       new  9e5e287   BUG: fix sparsetools tests in old scipy
       new  f778a8d   WIP: new Cython interface for decision trees
       new  ecd4f29   WIP: comments on the Cython interface
       new  51780b1   WIP: Criterion interface and base class
       new  5355988   WIP: ClassificationCriterion (reset, update)
       new  251ed5f   WIP: Gini criterion
       new  c0e1485   WIP: entropy criterion
       new  5a74082   WIP: remove n_left and n_right attributes
       new  28ee133   WIP: MSE criterion
       new  9eccd4f   WIP: tree class
       new  5a0cd20   WIP: tree algorithm
       new  92695fd   WIP: add_node
       new  f0d8b15   WIP: node_value
       new  bd71188   WIP: node_value
       new  6c39af3   WIP: predict + apply
       new  a488b3c   WIP: Random Splitter
       new  64c5af5   WIP: splitter
       new  6c832ba   WIP: Best Splitter
       new  83e7e89   WIP: sort features
       new  d27763b   WIP: first pass on tree.py
       new  794a1e2   WIP: some debug
       new  03a3763   WIP: some more debug
       new  23eef48   WIP: debug in progress...
       new  9a4cae5   WIP: debug (tests still don't pass...)
       new  4ea0d11   WIP: one more bug fixed
       new  8c82dcd   WIP: cleanup
       new  f2ca99a   WIP: one more test fixed
       new  1912a28   WIP: more bugs fixed :)
       new  0b3cf31   WIP: 19 tests passed
       new  172b021   WIP: test_tree.py now passes \o/
       new  da811ec   Cleanup
       new  bbc3c23   WIP: feature importances
       new  6306c4c   WIP: discard samples with weight = 0
       new  3200735   WIP: fix export functions
       new  abaee1a   Cleanup
       new  76670a1   WIP: first pass on ensembles
       new  79ee321   WIP: use heapsort
       new  6378925   WIP: small optimization to heapsort
       new  7e03c7d   removed min_density example - dropped param
       new  4742806   gbrt now works with new DecisionTree implementation
       new  c0a4a9a   import classes - now they work!
       new  bf8c786   fix: proper dtype for SIZE_t
       new  7344b8e   add GBRT to covertype benchmark
       new  9f0e56e   WIP: remove asserts
       new  44171c8   WIP: use C-based random number generator
       new  0e420b7   WIP: set n_classes as ndarray
       new  7880b7d   FIX: fix test_random_hasher
       new  5e5dd90   WIP: fix adaboost
       new  5fd21a1   WIP: small optim to regression criterion
       new  faeea4b   WIP: optimize tree construction procedure
       new  6de7739   WIP: optimization of the tree construction procedure
       new  561a6b5   cleanup
       new  2debc7a   recompile _tree.pyx
       new  167bbec   FIX: export_graphviz test
       new  c806cdf   FIX: set random_state in adaboost
       new  314ffce   FIX: doctests
       new  93806cf   FIX: doctests in partial_dependence
       new  ead9631   FIX: feature_selection doctest
       new  2d69d4c   FIX: feature_selection doctest (bis)
       new  e66224f   WIP: allow Splitter objects to be passed in constructors
       new  ca3cf08   FIX
       new  563f907   Some PEP8 / Flake8
       new  7f9a595   Small optimization to RandomSplitter
       new  9b62f3c   FIX: fix RandomSplitter
       new  fa34848   Cosmit
       new  950622b   FIX: free old structures
       new  8f18d96   WIP: Added BreimanSplitter
       new  3f540da   WIP: small optimizations
       new  b398083   WIP: fix BreimanSplitter
       new  e5d0416   Cleanup
       new  730d6b0   WIP: optimize swaps
       new  2ee32a2   Regenerate _tree.c
       new  a3a7fc8   WIP: some optimizations to criteria
       new  a30fa20   WIP: add -O3 to setup.py
       new  13bb04b   WIP: normalize option for compute_feature_importances
       new  2027f89   WIP: Added deprecations in tree.py
       new  0bd3f50   WIP: updated documentation in tree.py
       new  5170345   WIP: added deprecations in forest.py
       new  79953a0   WIP: updated documentation
       new  7785287   WIP: unroll loops
       new  eff6d64   WIP: setup.py
       new  42af22d   WIP: make sort a function, not a method
       new  2e71ba3   WIP: Cleaner Splitter interface
       new  e2363c9   WIP: even cleaner splitter interface
       new  46d43e8   WIP: some optimization in criteria
       new  5d37cbd   WIP: remove some left-out comments
       new  b1848d5   WIP: declare weighted_n_node_samples
       new  ad8b6b9   WIP: better swaps
       new  247172e   WIP: remove BreimanSplitter
       new  8499b11   WIP: small optimization to predict
       new  9b5cafe   WIP: catch ValueError only
       new  a2ea591   WIP: added some documentation details in _tree.pxd
       new  52b193a   WIP: PEP8 a few things
       new  01081b5   Benchmark: use default values in forests
       new  f84e463   WIP: remove irrelevant and unstable doctests
       new  8f0fd21   WIP: address @ogrisel comments
       new  10a4155   WIP: address @ogrisel comments (2)
       new  d1e72c5   WIP: remove partition_features
       new  95f5df0   WIP: style in _tree.pyx
       new  cf19e1b   WIP: make resize a private method, improve docstring
       new  4c032ac   WIP: use re-entrant rand_r
       new  4e3d8f0   FIX: doctest in partial_dependence
       new  24cb60d   WIP: break or shorten some long lines
       new  570bca0   FIX: doctest in feature_selection
       new  b564098   WIP: break one-liner if statements
       new  0c79587   WIP: revert use of rand_r
       new  c7266ec   ENH back-port rand_r from 4.4BSD
       new  3b81aee   FIX move rand_r to tree module for now
       new  a398577   FIX: broken tests based on rng
       new  0be3a70   DOC: update header in rand_r.c
       new  c1b93ed   TEST: skip test in feature_selection (too unstable)
       new  a42701e   FIX: one more doctest
       new  dc94d14   WIP: Faster predictions if n_outputs==1
       new  cb994bf   WIP: Break comments on new line
       new  cc3f2cf   WIP: make criteria nogil ready
       new  bbdccca   WIP: enforce contiguous arrays to optimize construction
       new  e0dcedd   WIP: avoid data conversion in AdaBoost
       new  f03095a   WIP: use np.ascontiguousarray instead of array2d
       new  4e4316c   TEST: add test_memory_layout
       new  ed9d503   FIX: broken test
       new  2c3861f   WIP: Make trees and forests support string labels
       new  1bd7c8d   WIP: refactor some code in forest.fit
       new  757de67   TEST: skip doctest in feature_selection (unstable)
       new  8f9f2e4   WIP: better check inputs
       new  df8eedd   Make callable svc test more robust for MacOSX.
       new  80dba53   WIP: check inputs for gbrt
       new  239054d   Merge pull request #2131 from glouppe/trees-v2
       new  c0b35eb   What's new: new implementation for trees
       new  cbd6544   Added decision_function support to OneVsRestClassifier and a test, test_ovr_single_label_decision_function, in test_multiclass.py
       new  5088875   Updated fixes for #2012.
       new  baffa30   Strengthened tests for OneVsRestClassifier decision_function
       new  7844861   Cleaned up tests, and removed unused multilabel parameter in decision_function_ovr
       new  2d6093a   Inlined extraneous function call from decision_function and added a check that the base estimator has a decision_function attribute
       new  d504675   PEP8
       new  f1f0a1a   FIX Minor stuff in document_classification_20newsgroups output
       new  abddaa8   ENH Add filters on newsgroup text
       new  a75949a   DOC 20news filtering with smaller set and MultinomialNB
       new  c61b560   misalignment fix
       new  0c6c773   FIX: remove debug message
       new  33069d1   added pxd to Manifest (to be included in source tarball)
       new  0535d27   Merge remote-tracking branch 'upstream/master'
       new  55377a1   PY3 fix string literal syntax error
       new  613d025   TST skip Graphviz export docstring in trees
       new  12d9bf2   FIX: remove -funroll-all-loops
       new  0a23ff7   FIX: ur strings are not supported in Python 3.3
       new  cbfe851   example fixes to clean first docstring paragraph of rst code
       new  0454a0e   ENH remove _is_1d and _check_1d_array thanks to @GaelVaroquaux
       new  813203d   flake8
       new  94227aa   ENH raise ValueError with row vector if multilabel or multioutput is not supported
       new  83b4836   ENH being less permissive thanks to @jnothman
       new  da7fe89   DOC add example is_multilabel
       new  0d22d8c   ENH handle properly row vector
       new  a406a5d   Flake8
       new  46e201d   ENH better error message
       new  7967c34   FIX switch to the new format syntax
       new  636033f   ENH prettier error message for _binary_clf_curve with bad input shape
       new  7ab0eb6   ENH use ravel instead of atleast_1d and squeeze whenever possible
       new  d2c3608   ENH coherently input checking for regression metrics
       new  243e633   ENH dryer thanks to @jnothman
       new  669e7d2   TST stronger test for _column_or_1d function
       new  31b056b   Cosmit
       new  b285bd0   ENH better error message when estimators don't specify their parameters in the signature.
       new  b60aa08   TST use TruncatedSVD in random forest tests
       new  ac40a85   Merge branch 'pr_2002'
       new  65cce1b   fix merge conflict
       new  b892db9   COSMIT refactor random forests
       new  ca40d2f   COSMIT refactor forests, part 2
       new  71de0fe   FIX ^ is a symetric difference
       new  1a21518   border added for IE
       new  b5950ad   Style fixes
       new  cfe0c56   [DOC] missing parameter description
       new  a9354a6   MAINT Set random_state, modernize tests
       new  5fafad9   TST max_features for more tree estimators
       new  73e24d2   TST remove unused tests
       new  d276e55   ENH add missing pxd of utis.random
       new  1c052f3   ENH Use file configuration
       new  d000060   FIX signature
       new  df7e135   cosmetic: Changed all instances of nonlinear to non-linear
       new  6094feb   Merge pull request #2187 from FedericoV/non_negative_style
       new  798f442   FIX faulty import in 20news docs
       new  043aaa1   Fix as_float_array behaviour when copy=True
       new  c91504a   Add test of the as_float_array behaviour when copy=True
       new  edb2ff7   COSMIT sphinx fixes
       new  e488704   COSMIT pep8
       new  5546681   Fix broken link to web designer
       new  61a16b9   Fix broken link to web designer
       new  92a5197   Add fit_transform to FastICA.
       new  fe8790c   Add inverse_transform to FastICA.
       new  62795e0   Add docstrings to methods in FastICA.
       new  9ec0127   Address @dengemann's comments.
       new  7c30e5a   Add test.
       new  12c41e7   Push failing test.
       new  4200c96   FIX transform tests
       new  7ad789d   FIX: remove inplace mod
       new  0e55dfe   ENH fit_inverse_transform for FastICA
       new  1708214   DOC document mixing_ attr on FastICA
       new  5ee41cc   COSMIT attribute checking in FastICA
       new  04459cc   COSMITS
       new  0000826   FIX: inverse transform + add mean_
       new  8342b7c   COSMITS
       new  9b0a96d   FIX: syntax typo
       new  17aff7d   FIX: tutorial
       new  643a00f   COSMITS + DOC
       new  772102d   COSMITS
       new  f7f2db1   ENH: improve tutorial to be more clean.
       new  27cc389   ENH + FIX: remove inverse-t kwarg + fix mean_
       new  0e58157   FIX: address @agramfort 's comments
       new  5f8aa2e   FIX: address remaining issues
       new  340dd33   FIX : finish get rid of fit_... param
       new  2c5b03a   ENH : avoid one copy in FastICA code
       new  3ff1f65   misc
       new  1e02032   ENH: speed up logcosh
       new  91f588b   update ICA examples
       new  71b5a16   FIX: Finish package rename from mst -> sparsetools. Fixes #2189.
       new  db7d0bf   DOC: Fix backwards docs on thresholds for preprocessing.
       new  a00a50e   adding comment
       new  ec1bdbc   Merge pull request #2196 from erg/labelencoder-docs-fix
       new  d6b0f36   make new classes for lasso_path/enet_path and deprecate old
       new  1c784b9   ENH : massive refactoring of CV models in coordinate descent. Now the algo core is in path functions
       new  a164db3   update what's new
       new  52ed83b   FIX: Py3 print statements to print functions
       new  3953c47   Merge pull request #2195 from erg/bug-2189
       new  ed6cd9d   ENH added an option to do an isotonic regression on decreasing functions
       new  03bd2bf   rel_canonical prelim
       new  7749896   Merge branch 'master' into ENH_docstrings_in_gallery
       new  650e643   syntax fix
       new  192c558   TEST: added a small test for fitting an isotonic regression on a decreasing function
       new  31d0f45   DOC : fixes in covariance module
       new  8b31ddb   DOC : more fixes in covariance module
       new  3fdb4fc   COSMIT explicit None check in naive Bayes
       new  fcb8f30   TEST tests the class instead of the function for the decreasing isotonic regression
       new  4dd30a5   Merge pull request #2202 from NelleV/isotonic_reverse
       new  d743520   cleaned up-ready
       new  34b3f03   add OOB improvement and set oob_score deprecated
       new  eecca4b   example for oob estimates in GBRT
       new  9b20081   plot cv error as well
       new  1073699   rm print stmt
       new  1fd0090   rn: plt -> pl
       new  9c58296   fix: oob_improvement_ with trailing _
       new  8376e12   more docstrings
       new  81ffae5   cosmit: use train_test_split - tuned params for nice plot
       new  04e8d2b   narrative documentation for oob improvement.
       new  f1d0451   more tests
       new  898dbc7   cosmit: better links and a note on efficiency using max_features
       new  c2e9949   comments
       new  017b605   cosmit: n -> n_samples
       new  725020a   cosmit: rs -> random_state
       new  b8f018d   more doc for OOB example
       new  42ff16c   use new style str formatting
       new  bf063b1   rearanged some code
       new  e75d9bb   rn: ACC -> Accuracy
       new  08d795a   rephrased max_features doc
       new  cebaeeb   moved to new pyplot import
       new  b5c055b   more narrative documentation for oob in gbrt
       new  1efd666   regression tests for oob_improvement_
       new  63c8848   example doc string
       new  30c08d4   Merge pull request #2017 from jaquesgrobler/ENH_docstrings_in_gallery
       new  f727efa   Small docstring changes for plot_ward_structured_vs_unstructered example, as mentioned in PR #2017
       new  67ac5f7   Merge branch 'gbrt-oob-improvement'
       new  6de5e41   covertype benchmark: use C-style input as default (most models require it as input)
       new  50de54d   fix: use asserts from sklearn.utils.testing fix: duplicate test names
       new  a1cbe78   TST error message for _check_clf_target
       new  407a43e   COSMIT
       new  e93cba3   FIX TST given cosmit
       new  5c92dff   COSMIT don't need set
       new  913e815   DOC explain the code
       new  bff544e   COSMIT product(..., repeat=2)
       new  1720cef   nitpick fixes, pep8 and fix math equations
       new  76ba751   fix: python3.3 warning fix
       new  4e7e637   Merge pull request #6 from jaquesgrobler/cov_doc_fix
       new  20f1554   Merge pull request #2203 from agramfort/cov_doc_fix
       new  393b996   ENH simplify the Scorer API
       new  5597d6f   FIX bug in scorers that take probabilities
       new  c9c7a61   removed old_version block test
       new  720bc2e   Merge pull request #2205 from jaquesgrobler/ENH_rel_canonical
       new  9f365ee   ENH added Restricted Boltzmann machines
       new  75c7cc5   30% speed-up thanks to in-place binomial
       new  3388b66   ENH 12% RBM speedup with ingenious ordering of operations
       new  96a2834   rename h_samples to h_samples_
       new  3068e61   added URI for RBM reference
       new  78b128e   improved docstring for transform
       new  97c9760   renamed _sigmoid to _logistic_sigmoid
       new  7ca5e30   use double backquotes around equations
       new  0b23c07   logistic_sigmoid moved to function
       new  a4b4b43   transposed components_, no performance penalty
       new  90c6fe8   only compute pseudolikelihood if verbose=True
       new  fd2867b   more accurate pseudo-likelihood
       new  e0490a4   use iteration terminology instead of epochs in RBM
       new  b784c4c   default n_components from 1024 to 256
       new  b22c0ca   clarify some method names (ex: mean_h -> mean_hiddens)
       new  0651905   added epoch time
       new  2f17015   ENH RBM example
       new  13a57a1   switched to digits
       new  cf4b9c3   moved rbms to neural_networks module
       new  a465523   add tests for rbm
       new  ec58f98   trim whitespace
       new  971360a   COSMIT RBM test in usual nose style + moved to proper module
       new  4212444   BUG + COSMIT + ENH RBMs
       new  a9a7ae7   use train_test_split
       new  41fedd6   Added RBM to whats_new.rst
       new  e8b0f95   DOC Added skeleton for RBM documentation
       new  2d86899   ENH Rename RestrictedBolzmannMachine to BernoulliRBM
       new  cd79faf   FIX: make BernoulliRBM doctest pass
       new  71e058c   FIX: BernoulliRBM check random state in fit, not in init
       new  49c4a7a   FIX: validation in `BernoulliRBM.transform`
       new  db373c2   DOC: first attempt at RBM documentation
       new  36681ab   Link to RBM docs from the unsupervised toctree
       new  0835af4   FIX: uneven RBM image
       new  82cd51d   DOC: PCD details and references
       new  3cafa9f   neural_networks -> neural_network
       new  712f3cc   ENH rename n_particles to batch_size in RBM
       new  44b7be1   TST added more RBM tests
       new  2c905f2   BUG: fix unsafe casting
       new  2ed3192   DOC: improve RBM example
       new  ef16f95   Fix typos in example
       new  0dd53f4   PEP8 and indentation
       new  2b17b4a   DOC add plot and example to docs
       new  b7e7a68   DOC rewrite BernoulliRBM example description
       new  08a2a12   Set seed through params, not globally
       new  8e7fd22   FIX handling of random state, hide some of API
       new  d3a7491   nudging the digits dataset for BernouilliRBM example
       new  e9e799d   Pep8 example
       new  e2ea85a   Update example params by grid search, and docstring
       new  356ec10   One space after dot
       new  9ba37e5   DOCFIX neural networks module
       new  71ff60b   DOCFIX spacing and clarification in RBM docstring
       new  8795ab3   More stable implementation of logistic function and its derivative by @fabianp
       new  38fca5a   Use gen_even_slices instead of homebaked code
       new  110297d   TST added a 'fit [[0],[1]] + gibbs sample it' test for RBMs
       new  59631d6   replaced test_gibbs by a smoke test for NaNs
       new  d3c34fd   check for pseudo_likelihood clipping
       new  e488649   ENH Add fast and stable logistic sigmoid to utils and RBM
       new  00cf80c   ENH Support sparse input in RBMs
       new  0f11698   COSMIT refactoring rbm
       new  89d0eaf   ENH Prevent memory copying in RBM's _fit
       new  050fe37   Do not touch uncopied memory
       new  62763ea   RBM example now verbose
       new  ac1e1a9   squeezing logistic_sigmoid result only on 1D arrays
       new  31cac01   adding a test for sparse matrices in RBM
       new  f05c6c5   changing free_energy to private in RBM
       new  52eeb95   added neural_network to setup
       new  f37668f   Nudge images using convolve, slower but more readable
       new  beb4d73   Clarify narrative docs
       new  c7f9682   Clarify and python3 RBM example
       new  9b40246   Periods and other docstring issues
       new  539bb39   MISC: remove unecessary dtype
       new  5a1097e   Remove redundant test
       new  d497ec0   Python3 support in RBM
       new  3b1bd59   TST RBM smoke-test verbosity
       new  5bba0a8   Add a copy parameter to safe_asarray()
       new  de4b923   Merge branch 'pr/1954'
       new  e3d2ecf   MAINT _logistic_sigmoid.c is "binary"
       new  de6e1a3   cosmit : protect attributes in RBM for sphinx
       new  cec3061   Imp readability
       new  24cb1f0   PY3 fix RBM test
       new  8336f78   GraphLassoCV works with alphas given as list.
       new  27e2b6e   Simplified GraphLassoCV code.
       new  f0858e3   Put back cov_init parameter in graph_lasso_path_
       new  f1eecf1   Speed up some tests
       new  5c5890a   Removed unused import
       new  0626d73   ENH: improve ICA memory profile by 40%
       new  9ebdd08   pep8
       new  e73b0b0   better coverage
       new  a10bfd1   fix doctest
       new  86775a6   ENH : use warning instead of print
       new  9eb9470   update what's new
       new  4cb8e88   Merge pull request #2212 from dengemann/ica_memory
       new  21615b2   DOC/FIX affinity_propagation damping default value.
       new  2b91f39   Merge pull request #2213 from cmd-ntrf/master
       new  81c7a7a   FIX missing class attribute in ICA. Common test was failing
       new  20f5985   DOC: some documentation for the Tree Cython structure
       new  bc572f6   FIX: fastica function dictionary default value
       new  b5d89de   DOC copyedit RBM docstrings
       new  87a18ae   Merge pull request #2216 from glouppe/tree-doc
       new  02ae145   Deprecate FastICA.sources_
       new  419c2be   Rename LabelBinarizer.multilabel to .multilabel_ + DOC
       new  e44eb54   WIP: partial fit for discrete naive Bayes models
       new  c9bd6ea   Remove the class_prior partial_fit param
       new  02cdd92   WIP: started to factorized the raw count collection
       new  dea7f3f   Incrementally is useless now
       new  ef17c8b   Add reference to the Manning text + restaure previous smoothing
       new  6f6a6d2   FIX shape issue when y has only one single class + some missing doc
       new  84397fb   Factorize common classes checks in partial_fit implementations
       new  7362e28   Add note on a possible future performance optimization
       new  b6d59fe   Add a note on performance tradeoffs in the docstring of partial_fit
       new  87d2fca   More informative error message. Also CV now use integer indices by default now.
       new  fa47f2c   Use floats everywhere to get rid of warnings when using sample_weight
       new  4719670   More input checks
       new  3a0214f   Better test name
       new  950bbfd   Remove redundant shape check already done by check_arrays
       new  1505af9   Add missing test for sample weight with partial_fit + fix issue classes passed as a list instead of an array
       new  ae8a0c9   One more input check test
       new  ba69ce4   Add missing test for deprecation warning
       new  1bbdeec   Found a bug: add a failing test
       new  aa44a21   Use unique_labels more consistently in the multiclass model
       new  bbba787   Fix broken partial_fit test
       new  fc8e88c   Factorize label_binarize for binarizing a sequence of labels with fixed classes
       new  9fd25eb   Add a new whats_new entry
       new  30464c7   Add some doc for the new partial_fit method
       new  f3b3a87   wording
       new  e81b9f6   Avoid raising a deprecation warning on label_binarizer_.multilabel_
       new  2f42fd2   Fix docstring and add some usage examples
       new  7d21296   FIX: do not update feature_log_prob_ in _update_class_log_prior
       new  4914657   TEST remove deprecated stuff from fastica tests
       new  14f2b29   Document the deprecation
       new  b00f120   Add one more tests to check the performance on digits
       new  edbeb64   DOC pep257 + c/e in sklearn.base
       new  6244542   FIX bug in test
       new  5d8caab   Make test_deprecated_fit_param pass under python 3 as well
       new  0a42d7a   MAINT moved the pls file based module to a folder
       new  67191ac   TEST fixing pls tests failing:
       new  38e84c5   MAINT Move the pls to the cca to a cross_decomposition module
       new  f93659b   MAINT renamed pls to cross_decomposition in the documentation
       new  0cb1bc3   FIX the example plots of the pls module did not import pls methods from the correct module
       new  6963725   FIX removed the cca and pls modules
       new  d045be0   FIX added the new module to the setup.py installation
       new  8b13a1f   DOC improved docs/docstrings on cross_decomposition
       new  b732f4d   MAINT deprecated the pls module, moved CCA to cca_
       new  a4efec5   FIX init methods of ABCMeta class also need to be abstract
       new  987cc91   Address wording and typos identified in review
       new  5cc9a63   Better parameterization for test_check_accuracy_on_digits
       new  0a6c439   Add a whitespace in parameter docstring item
       new  908ab6e   Merge pull request #2217 from vene/ica_fit_transform
       new  b82601f   More accurate documentation for class_count_ and feature_count_
       new  e8414db   Merge pull request #2182 from NelleV/pls_refactor_2
       new  08de819   Rename helper partial_fit function
       new  a702a35   DOC+ENH: fixes in least_angle + one vectorization
       new  0c461e4   ENH: better error message on scoring
       new  2ef7ece   DOC: reorganize model_evaluation
       new  c9a1808   MISC: address comments and test failure
       new  3d0fc04   DOC: address remarks by @NelleV
       new  0da3477   DOC: Address @larsman's comments
       new  c80b3e1   DOC: @amueller's comments
       new  2b8dc2d   DOC FIX typo on fbeta, closes #2219
       new  a1b5553   FIX on py3k, we need explicit relative imports
       new  040ac76   DOC : better doc of array shapes in fastica
       new  6252883   Merge pull request #2175 from ogrisel/nb-partial-fit
       new  d651a91   MISC : use linalg from scipy
       new  f49f6ad   Update mailmap
       new  4872496   doc: hedge the use of OOB estimates
       new  b0bd203   ENH : removing warnings from tests in cd linear models
       new  b628b82   Merge pull request #2194 from NicolasTr/as_float_array_copy
       new  c38caac   Refactored verbose output in GBRT - output much more nice Added tests for verbose output in GBRT cosmit: use str.format instead of % updated whats new with GBRT changes
       new  3eccb93   DOC add missing datasets helper
       new  6fc30af   ENH remove deprecated
       new  cb113b9   Benchmark: use specified dtype
       new  20b4e4a   fix whitespace around new tree.pyx docstring
       new  b86071b   Missing value imputation
       new  64efcd5   TEST: cosmit on err_msg
       new  81857bf   Raise an exception if rows are full of missing values
       new  0df316c   FIX: doctest
       new  fabca84   Better error message
       new  fe5e2bf   Fix tests
       new  6d79e59   Fix tests + doc improvements + renaming
       new  ffc0d20   Add test with default value of copy + doc improvements
       new  feea9a5   Imp readability
       new  a0b0dc7   Fix use of as_float_array
       new  e3a07ac   pep8
       new  8c1f77d   Imp variables names
       new  3a0589c   Del use of as_float_array + naming and documentation improvements
       new  343ce5e   Fix use of mask
       new  5ad8bff   Fix import names
       new  aac47a2   ENH remove deprecated things (2)
       new  9efd24c   Update what's thanks @NicolasTr
       new  e15f768   FIX: use range instead of xrange
       new  6faafac   Merge pull request #2223 from arjoly/doc-datasets
       new  f1ac2e0   ENH add support for string input with classification metrics
       new  90f062a   ENH use the new format syntax
       new  a8d4956   ENH remove inspect
       new  541e44e   TST fix string labels in metrics tests
       new  d87c4f7   COSMIT
       new  c165b39   Update what's new
       new  62018d9   DOC state that string is possible
       new  fa03e72   TST with labels arguments
       new  0d0a932   FIX what's new...
       new  e1fd0dc   use new virtualenv features of travis, so we don't have to kill the virtualenv
       new  96a24e5   DOC copyedit preprocessing docs
       new  08b2a81   Merge pull request #2228 from amueller/travis_virtualenv_stuff
       new  7f5462c   Trying to enable python 3.3 too.
       new  060bab9   Update .travis.yml
       new  025c25f   MAINT ignore profiling results from kernprof.py
       new  55d2668   DOC copyedit KernelCenterer docstring
       new  9d89354   DOC minimal kernel centering narrative docs
       new  1da5058   fix: newest numpy doesn't like all-indexing non-existing dimension (reported by erg #2233)
       new  6c89eb8   ENH remove bad examples
       new  1a6383f   DOC let some example for prf metrics
       new  aabdbb0   Merge remote-tracking branch 'upstream/master'
       new  cb868ec   FIX: Newer numpy causes scipy to issue a DeprecationWarning. Ignore it. Fixes #2234.
       new  8436201   ENH allows make_multilabel_classification to return label indicator f…
       new  0bce8a1   TST grid_search_cv works with multioutput data
       new  7f95b85   TST cross_val_score with multoutput data
       new  bd6c741   FIX missing deprecation release information.
       new  38b8755   DOC recommend understanding NumPy in the tutorial
       new  3f0751f   DOC expand feature selection docs with an example
       new  89012aa   DOC minor copyedit to FS docs
       new  19033bd   ENH: Add the hungarian algorithm
       new  16e48ab   TEST: Increase testing of hungarian
       new  c5417e7   MISC: cosmit in hungarian
       new  dab7e50   ENH: Speed up in hungarian
       new  1e06e49   ENH: More speedups in hungarian
       new  5fdfede   ENH: More speedups in hungarian
       new  7026c7f   ENH: Still more speed ups in Hungarian
       new  47c643c   ENH: More speedups on Hungarian
       new  941e60c   This should make the hungarian algorithm accept rectangular cost matrices. Also enabled the tests. NOTE: Only tested on rectangular matrices of shape nxm such that m > n.  Tests need to be expanded to test m < n.
       new  00249b1   An additional check needed in case where there are fewer columns than rows. All assignments are made, but the algorithm wants to keep going because there are some rows left.
       new  4157f20   Added support for hungarian assignment problems where one dimension of the cost function is zero-length. The result in such a situation should be an empty array.
       new  bd53545   Created an alternative hungarian solver for rectangular matrices that does not involve matrix padding. Also modified the test to test the transpose of the cost matrix as well.
       new  e59ae98   hungarian() now returns a 2-D array of indices instead of a 1-D array. Also modified the find_permutations test to accomodate.
       new  fd74a46   Some minor changes to docs, and small simplification in code. Still need to update docs for hungarian() and find_permutation().
       new  e14fff0   API: scikits.learn -> sklearn
       new  4fc652bb  Updating namespace usage from scikits.learn to sklearn
       new  e0b6402   implement spectral biclustering and spectral co-clustering
       new  789d8d1   wrote tests
       new  ca902ad   wrote methods for generating bicluster data
       new  cb66caa   added option to return piecewise vectors
       new  39e65cc   cast data in fit()
       new  125ee1a   made internal functions private
       new  55d54fc   use random state in test
       new  2fc53d3   removed pickle test
       new  ab93c0a   shorten first lines of test docstrings
       new  53ac215   use random state in preprocess tests
       new  9f5bad3   duck typing, minor corrections: spacing and typos
       new  316c1b6   fixed exceptions and their messages
       new  216b71b   updated svd()
       new  f3a4c24   better array validation
       new  ddf82ee   use random state in data generator
       new  7f025cf   tests reuse data generators
       new  ce5496e   user may select svd method
       new  2c96083   Added to docstring
       new  0eb5c06   split spectral biclustering into two classes
       new  5ebd11b   removed unused code
       new  4ae9c2c   test bad arguments
       new  268b7ee   now supports sparse data
       new  76143b2   check n_clusters parameter more thoroughly
       new  d51697a   made base class an abstract class
       new  34caec6   checkerboard panels may have arbitary values.
       new  c1df6bc   fixed exception type
       new  d53f470   removed empty mixin
       new  b218edc   started biclustering documentation and examples
       new  0248fc7   shorter array slicing
       new  89fbea5   made some methods into private methods
       new  9856687   cleaner use of check_arrays()
       new  a421e5e   named arguments
       new  64ce386   use safe_sparse_dot()
       new  8204731   use np.random.RandomState directly
       new  9510c20   do not do any checks during __init__()
       new  6b92b4c   do not use mutable default arguments
       new  642be48   added new tests for sample data generators
       new  36daa8b   fixed bug in make_checkerboard(), so tests pass again
       new  2e78d10   use assert_all_finite
       new  d6d635a   skip permutation test for now
       new  836f3f0   fixed some errors reported by pyflakes
       new  bfaf21d   raise exception instead of converting sparse arrays to dense
       new  18c305d   expanded biclustering documentation
       new  0533dcc   corrected k_means in docstring
       new  fe5f3cd   rearranged imports from general to specific
       new  395b231   moved and renamed _make_nonnegative() and _safe_min()
       new  3ca7b70   added option to use mini-batch k-means
       new  470b493   use dia_matrix
       new  5694d34   renamed 'preprocess' to 'normalize'
       new  ae3d61a   use sklearn.utils.extmath.norm
       new  05dcd10   base class __init__ is no longer abstract
       new  135ee9b   added more information to error messages
       new  ec2254a   also use norm in _project_and_cluster()
       new  d34cb31   make test more sparse
       new  9c20a9e   made 'bicluster' a submodule of 'cluster'
       new  08e5c35   removed svd_kwargs argument
       new  9d32afa   added n_svd_vecs parameter
       new  ba68233   tests use ParameterGrid to avoid deep nesting
       new  202e6ee   replaced kmeans_kwargs with some useful k-means parameters
       new  3be95af   updated documentation
       new  83a2850   keep biclustering algorithms in submodule
       new  8b2b8e5   renamed examples; added to example docstrings
       new  871d92b   re-added bicluster mixin, this time with some functionality
       new  1978d2f   wrote newsgroup biclustering example
       new  a591ff9   fixed a few things in examples, documentation, and docstrings
       new  5ad0dfb   wrote bicluster scoring using jaccard index and hungarian matching
       new  45dafac   removed some parameters to speed up test
       new  ecc96d9   added default arguments to base class's__init__ to make test pass
       new  4657ea2   test_make_checkerboard was wrong after api change
       new  f3cf630   added documentation for bicluster evaluation
       new  96e0264   moved shuffle functionality to utility function
       new  e748018   added consensus score to bicluster examples
       new  b0a9f92   renamed example to get output to work
       new  9308d4a   made bicluster utilities for dealing with indicator vectors
       new  108d1bc   index in one go. added sparse test.
       new  4fda1a2   documentation and docstring fixes
       new  510d860   merged newsgroup example with Vlad's
       new  cbdd485   moved bicluster examples to their own category
       new  269eeb0   reduced noise in spectral coclustering example
       new  b95b9c4   updated newsgroups example
       new  8984ce4   added n_discard parameter to _svd()
       new  2137016   check value of n_components and n_best
       new  6a35a93   a fix for nan values in singular vectors.
       new  3c23903   wrote tests to ensure svd works on perfect checkerboard
       new  1893477   redundant phrase in docstring
       new  479d327   put biclustering section after clustering section in reference
       new  2fa8586   misc. fixes
       new  0dadefe   changes to newsgroups example: - only print the best biclusters, ranked by ncut; - increase example batch size
       new  f9efd1b   fixed some docstrings: backticks and missing parameters
       new  89be3d5   updated setup.py
       new  2442b84   added myself to authors; added biclustering to whats new
       new  af8abde   examples use matplotlib.pyplot instead of pylab
       new  f9670d8   consistency changes: - renamed maxiter to max_iter - named arguments to apply_along_axis()
       new  6789f59   Clean up and rename Hungarian algorithm
       new  bbccabc   Clarify and clean up example
       new  0749e57   Remove print in Hungarian tests
       new  41438e0   Consistency for floats in consensus score
       new  56d9728   Add warning in private _Hungarian docstring just in case
       new  0442d00   removed plot_ from newsgroups example file
       new  76ae2fa   import biclustering methods in sklearn.cluster and sklearn.metrics.cluster
       new  c0a0e31   ENH make spectral clustering test more stable to random seed
       new  95d9c63   FIX: imputation example
       new  0bbd57b   Merge pull request #2229 from larsmans/kernel-center-narrative
       new  04c132b   remove negative indices from neighbors cython code   - use X.ndim - k instead of -k for indexing
       new  88fd230   MAINT charset is deprecated in favor of encoding
       new  6f030d9   ENH add return_path in orthogonal matching pursuit
       new  31c6ff5   TEST for omp path feature
       new  372ca22   ENH OrthogonalMatchingPursuitCV estimator
       new  cc4b4a1   DOC : docstring fixes
       new  bad0538   DOC : more docstring fixes
       new  e417221   use pre_fit in OMP
       new  4ce4863   API : deprecate a lot of extra parameters in OMP object
       new  1e408aa   API : deprecations in orthogonal_mp
       new  a6cb736   ENH : update example of OMP
       new  c4df8ed   FIX respect conventions in OMP init
       new  4d3a7ff   update what's new + classes.rst
       new  8c9dbc9   TST added tests for encoding/charset deprecation
       new  7dae356   Merge pull request #2230 from pprett/neighbors-segfault-fix
       new  0498af7   DOC better deprecation warning messages.
       new  a29392e   TST catch deprecation warning in feature_extraction.text
       new  11a0c28   Merge branch 'pr/2246'
       new  8367727   COSMIT
       new  ef1d3bd   Added GraphLassoCV changes to whatsnew.rst
       new  d634db8   [DOC] Corrected errors in clustering documentation
       new  a4fa80f   [DOC] fixed a typo in an warning message.
       new  de86e97   Merge pull request #2241 from arjoly/grid-cv-multioutput
       new  9ad4e8f   One more typo fixed
       new  7dd738a   Merge pull request #2247 from pgervais/docfixes
       new  2109b1b   ENH consistency mse=> mean_squared_error ari => adjusted_rand_score
       new  bc7571c   FIX docstring
       new  0da99d4   FIX hopefully fixing travis.
       new  a4e53a6   FIX hopefully fixing travis.
       new  86d7afb   FIX OrthogonalMatchingPursuit normalized twice
       new  a8f5b04   DOC correct/copyedit linear model docstrings
       new  a48207c   One more Python 3 fix in feature_extraction.rst
       new  c9df3ad   Update what's new
       new  c1c9e6a   BUG: fix some numpy 1.3 compat issue
       new  e63bc7a   BUG: numpy 1.6 compat
       new  3abd51b   :
       new  daa051e   Py3 fix
       new  dd097e7   DOC improve svm sample weight example
       new  c257543   DOC improve documentation of sample_weight, add to docstring.
       new  b54ecb6   TST small improvement of test for sample weight in svm
       new  006c434   BUG: fix kde tests
       new  f7555a1   DOC add missing links to the scorer and classication section
       new  9f8eb14   TST better testing of the PLS module
       new  9d85c99   FIX PLSSVD now returns the correct number of components
       new  f08ee94   COSMIT small documentation tweaks
       new  dc7091b   Add pycharm files in .gitignore
       new  01e99d0   cosmit typo
       new  d8e00eb   ENH add multioutput support to KNeighborsRegressor
       new  ee514a9   ENH add multioutput support to RadiusNeighborsRegressor
       new  85f4905   ENH add multioutput support for KNeighborsClassifier
       new  9a1c010   ENH add multioutput support to RadiusNeighborsClassifier
       new  aeaffe5   DOC + example with multioutput regression face completion for knn
       new  c6fa3a0   ENH allows make_multilabel_classification to return label indicator format
       new  351784a   ENH TST grid search with multioutput
       new  7e3f7ba   ENH TST random search with multioutput data
       new  60267d2   DOC gridsearch support mulioutput data
       new  9fbc072   TST cross_val_score with multioutput data
       new  2be42f0   DOC more information about which classifier support multilabel
       new  629483d   DOC unveil that some estimators support multilabel classification and multioutput-multiclass classification
       new  f4eded6   DOC overall improvements
       new  def7af6   pep8
       new  387951d   DOC credit + fix typo + wording + use mathplotlib.pyplot
       new  e25abf7   ENH take @glouppe comments into account
       new  1d6ce2c   FIX small title issue
       new  2f71e90   DOC update what's knn and radius-nn support multioutput data
       new  44f454c   Merge pull request #2258 from NicolasTr/ignore_pycharm_files
       new  f4fe3fd   Show 95% confidence interval, not 40% confidence ^^ Closes #1940.
       new  7b6a808   FIX whoops sorry!
       new  4a60dcf   fix pycharm file ending
       new  ec6057a   Imp splitting of preprocessing.py
       new  b8201b4   Imp splitting of test_preprocessing.py
       new  15984d9   Del unused imports in preprocessing + pep8
       new  bc36ad0   Fix imports
       new  9cb320d   Imp move OneHotEncoder to preprocessing/data.py
       new  f713996   pyflakes and pep8
       new  0db387f   MAINT: update copy_joblib script
       new  107a0c8   ENH: update joblib to 0.7.1
       new  7ba0d27   MAINT: misc change to copy_joblib
       new  1c133e8   ENH add "make_y_1d" to utils, use it in estimators where needed.
       new  1cc1147   fix make ``make_y_1d`` save for lists.
       new  d8f90d1   use column_or_1d, move it to utils
       new  32c1fc9   Fix self.statistics_ souldn't be set if axis==1
       new  2c02ab0   FIX inline rand_r to fix build on Windows
       new  30f06b1   DOC add an extremely simple classifier code example to dev docs
       new  b661cb7   DOC ignoring gen_rst's parsing errors
       new  8d5e142   Fix use of self
       new  66acf36   Merge pull request #2262 from NicolasTr/fix_statistics
       new  620ce4a   ENH added multiclass_log_loss metric
       new  808799c   ENH rewrite multiclass_log_loss, rename log_loss, document it
       new  d0474ae   ENH Scorer object for log loss
       new  377c61e   ENH add log_likelihood_score as -log_loss
       new  7164e33   BUG: add new preprocessing module to setup.py
       new  8f4d076   Merge pull request #2264 from jakevdp/setup_fix
       new  8fe7121   ENH: make bdist_rpm work
       new  32eb8b5   ENH rename eval / pseudolikelihood to score_samples
       new  857f918   PY3 new overfit prevention stuff in 20newsgroups loader
       new  f072108   FIX: remove blank lines
       new  5796359   fixing ridge and label binarizer... I'm pretty sure that worked before?
       new  090ba25   FIX make neighbors y prediction shape consistent
       new  fd8bef9   TST add regression test for label_binarizer
       new  2c5ea34   More explicit tests in test_label_binarizer_column_y
       new  0bad1ab   fix: check for impurity ties
       new  572600b   added 32bit 64bit equality test case
       new  961ac0a   adapt OOB regression test to change in tree module
       new  71e59b7   Use epsilon=1e-7
       new  25b5b96   FIX bug in f_score with beta !=1
       new  d6d876e   FIX formula inversion for sample-based precision/recall
       new  16a409a   FIX set same default behavior for precision, recall and f-score
       new  94ba088   ENH raise warning with ill define precision, recall and fscore
       new  cdfaa82   Backport assert_warns and assert_no_warnings from np 1.7
       new  d62b21c   TST test warning + ENH Add warning average=samples
       new  76bc601   FIX TST with warnings thx to @jnothman
       new  b9593f8   flake8
       new  5a1c565   ENH set warning to stacklevel 2
       new  1fad3e8   TST silence warning
       new  fc22665   ENH use with np.errstate
       new  2660a3b   DOC TST correct comment
       new  1fb634c   FIX warning test
       new  43bff76   FIX warning tests in preprocessing
       new  3a0af03   FIX: partial dependence test
       new  0c82974   FIX/ENH make StandardScaler convert int input to float and warn about it, instead of warning and rounding for dense and crashing for sparse.
       new  8b54e23   DOC adjust docstring as suggested by @gvaroquaux
       new  47c9a07   addressing @ogrisel's comments: catch warnings in test, no unneeded digits
       new  5a1fde1   COSMIT fixing some unused imports, adding stuff to __all__,  and light pep8 (not all whitespace to make rebasing less painful)
       new  12d1b66   TEST: skip test_oob_multilcass_iris for now
       new  b9cc3a7   Catch expected warning in sklearn/tests/test_naive_bayes.py (part of #2274)
       new  b3ab5fa   Merge pull request #2277 from glouppe/tree-fix-32bits
       new  206d227   Use projected gradient solver in transform to support sparse matrices
       new  ae74c9d   Use same parameters when solving the transform
       new  7aac367   Use scipy.nnls.optimize for dense data
       new  a9a51e2   Revert "Catch expected warning in sklearn/tests/test_naive_bayes.py (part of #2274)"
       new  6ff5d97   PY3K remove __pycache__ in make clean
       new  60f40e4   neighbors numpy1.3 compat: fix typedefs, regen with cython 0.19
       new  b793231   numpy 1.3 compat: use explicit type definitions
       new  318955d   numpy 1.3 compat: make neighbors/dist_metrics compatible
       new  525398c   COMPAT: empty_like does not have a dtype in np 1.3
       new  fb12a48   COMPAT: make NeighborsHeap compatible with numpy 1.3
       new  97c9252   COMPAT: make NodeHeap compatible with numpy 1.3
       new  ca60af5   COMPAT: make BinaryTree class compatible with numpy 1.3
       new  56f5f19   COMPAT: make BallTree & KDTree compatible with numpy 1.3
       new  926a7fd   COMPAT: last few BallTree/KDTree numpy 1.3 issues
       new  e97c266   COMPAT: fix arpack and pls on old scipy/numpy
       new  3edb5b1   COMPAT: string formatting syntax in Py 2.6
       new  8ec572d   COMPAT: median and nans in old numpys
       new  62cd4a2   COMPAT: no assert_warns in np 1.3
       new  0ef0804   FIX PY3K warning.catch_filter set record
       new  ef2a17b   FIX PY3: list and tuples cannot be compared in Python 3
       new  dd23eea   Py3: fix version comparison in imputation module
       new  8dd37ab   DOC fixing some sphinx stuff.
       new  de3866c   DOC SGDClassifier has multiclass predict_proba
       new  8d47341   DOC minor copyedit to narratives
       new  52efd9e   FIX don't use old scoring API in randomized search
       new  e647ac8   Add supported python versions to the classifiers + fixes
       new  88b7e84   Merge pull request #2280 from larsmans/randomsearch-scoring
       new  429ec5e   Merge pull request #2281 from ogrisel/improvements-to-setup-py
       new  f2643c8   Add failing test for libsvm random state proba
       new  89a58cd   FIX support random state in libsvm
       new  44dd09f   DOC document changes in LIBSVM_CHANGES
       new  2d97949   DOC update docstrings to reflect libsvm random_state
       new  d65de5e   Fix libsvm seed when predict_proba in tests and examples
       new  ed4c11b   Clarify and make libsvm random seed more consistent
       new  c1bab14   Comment predict params in libsvm
       new  3009a85   more sphinx fixes
       new  1904494   DOC reference and rename cross decomposition module
       new  f218fce   ENH: add failing test exposing bug in RandomizedPCA
       new  86d96ce   FIX: only center if copy == True
       new  09da6b5   ENH: get it right.
       new  ed0fba3   FIX: inverse_transform; tests
       new  45046fb   DOC better doc message
       new  94dad56   API: get rid of **params in PCA estimators.
       new  d229ac7   DOC: more doc string fixes in pca.py
       new  c5f4fa7   DOC: more fixes in pca.py doc strings
       new  544cba4   STY: get rid of unnecessary identifiers
       new  3453739   FIX: X.copy() test now works
       new  27370de   STY: removing unnecessay import
       new  bbb7b25   COSMITS
       new  0682711   COSMIT: typo in examples/imputation.py
       new  505bc21   BUG: fix Py 3
       new  b669f82   Sample compiler config for windows
       new  7271d53   Force stdc++ link for the windows build
       new  7ad2bf2   Regenerate pairwise_fast.pyx with recent cython for windows build
       new  43b3698   Fix atomics definitions under windows for sklearn._hmm.pyx
       new  b65e882   typo
       new  c9f3e01   Refactor loss_func and score_func warnings in grid_search
       new  31ae3ca   Add score_overrides_loss to _deprecate_loss_and_score_funcs
       new  93ef065   Add deprecation warnings in Ridge
       new  11dfe44   Add deprecation warnings in rfe
       new  eae4fab   Add catching of the deprecation warnings in rfe and ridge tests
       new  c16da08   Refactor loss_func and score_func warnings in cross_validation + replacement in two examples
       new  45d65c1   Fix 'scoring' docstrings
       new  e77e210   Imp documentation
       new  2918015   Fix tests
       new  55ca021   Fix grid_search.py example
       new  b898edd   Fix tests
       new  83e952d   DOC fixed the optional arguments
       new  6e9498f   Update multiclass/multilabel documentation
       new  c1d7666   DOC overall improvements in the multiclass documentation
       new  9ec7d05   DOC take into account @vene and @ogrisel + specify format for multioutput-multiclass
       new  b79005a   DOC rewording
       new  57ec891   Typo
       new  886754b   DOC ENH take into account @NelleV comments
       new  eb068a7   DOC more comments from @NelleV
       new  1d846c8   first try at bootstrap-based website
       new  f0ef902   "fix" sidebar stuff - this was not my idea
       new  32d0f67   remove gray boxes around h3 on the two new pages
       new  5ac91a5   put banner into header, make it spread over whole page
       new  00edac0   FIX added some descriptions to each categories in the main webpage
       new  d3b75d5   FIX spelling mistake
       new  1e30f72   FIX the css in the API
       new  c8771b6   Fix link to flowchart, add text descriptions.
       new  f7f01dd   Minor fixes in front-page text, css
       new  2f3edc6   rework front-page box texts
       new  1857762   fix typo, missing p
       new  4bba215   fix and refine some css and html tags
       new  4b3c3fb   ENH added the fork me ribbon to the website
       new  a1f308d   add example banner image
       new  bd20a8d   Mr. Proper, act 1
       new  282fa3b   add section, estimator and model links on the frontpage
       new  4204586   fix styling of rst links
       new  f4c306f   Banner improvements
       new  0388b32   add links for examples
       new  5b92394   Banner style
       new  a32e345   fix css that I just broke with the sphinx links
       new  211e538   flatten the tutorial / doc structure as proposed by @ogrisel
       new  d2fe4ae   Boxes on front page
       new  cc0ce1a   add js for collabsible toc tree in the user guide.
       new  7d18010   minor typo thing
       new  1e1336e   don't have old version warning on install, as that will be shared across all versions.
       new  7fe5383   sidebar fix - sidebar.js was called before jquery. works fine under new version jquery too
       new  7169c45   DOC: invert priorities bootstrap <-> nature.css
       new  80c8f3c   DOC: sidebar lighter
       new  6ed33dc   Load bootstrap first
       new  a504463   FIX: footer character encoding
       new  294125c   CSS tweaks
       new  96bd061   CSS tweaks (2)
       new  c1c093d   sidebar/toctree harmonie, must still fix toggle
       new  1b86a84   WEB added testimonials
       new  a821ccb   jquery reverted to 1.7.2 version. sidebar/toc-collapse works
       new  81a363a   Lower part of the index
       new  3ae6c62   DOC fixed the previous/next button
       new  0a117e4   CSS tweaks
       new  9ea3755   More css tweaks
       new  b5ac978   Better alignment in the sidebar
       new  bce901b   CSS tweaks
       new  654edd2   More css kungfu
       new  3b0b889   CSS stuff
       new  f8f3f9d   Remove testimonials for now
       new  f9943af   DOC fided the collapsable sidebar
       new  5066f1e   added "show source" link to footer, made dimensionality reduction examples link to decomposition
       new  96c9e51   slightly hackish way of inserting a whatsnew link. I really don't want all the sphinx containers here, though. Asked on stackoverflow about it btw.
       new  e491472   a little less ugly footer. @glouppe should maybe have a look ;)
       new  653a2ac   DOC dropdown menu works
       new  da33442   CSS tweaks
       new  f23cb42   Donate button + citing
       new  678b595   Enhance contrasts
       new  750989d   Contributin
       new  2f6a670   Remove toc on the API page (it is already in the sidebar)
       new  a833e7a   make links to old versions actually do something (currently link to the user guide as the other versions are not rebuild yet). Add developers guide section.
       new  6e30189   replaced lorem ipsum in news. still a draft but whatever.
       new  c3b52f7   nicer dates
       new  f9ad532   FIX use category and stacklevel=2 for {loss,score}_func
       new  c95842d   Use extra_link_args for -lstdc++
       new  db5b867   Ignore compiled shared library files generated in the source tree under windows
       new  2531fb4   ENH speed up BernoulliNB's predictions
       new  a3c3057   FIX raise tolerance in svm predict_proba test
       new  d4e9b8e   FIX minor edits on the website
       new  5c3fa68   DOC "creating features" -> "feature extraction" + minor stuff
       new  71f3b4b   DOC fixed z-index on the website
       new  0292f14   Try to raise and test warnings.
       new  652a873   ENH: add a new DataConversionWarning
       new  ae23b19   Make common PLS tests more stable
       new  89886e5   Revert "ENH add log_likelihood_score as -log_loss"
       new  c57858c   DOC added website to whatsnew, added link to github for Nelle
       new  c7c7ade   FIX don't use old API in examples
       new  be62aa9   more fixes for docs, deprecated interfaces
       new  d1c51e3   FIX made the building of the docs slightly more robust. readme files in folders without examples kill it otherwise.
       new  04f7d49   DOC copyedit example docstring
       new  add1e12   DOC XHTML fixes (unclosed tags, type="text/javascript")
       new  b3eddb7   FIX website layout on small screens
       new  d29719f   FIX improve display on small device
       new  7e1bb89   DOC fix dropdown menu
       new  b5e0674   try to fix the toctree in a semi-meaningful way.
       new  26eb3c6   DOC/EXAMPLES fix more documentation errors, deprecated api usages.
       new  75af99e   EXAMPLES remove non-existing example from doc, don't trigger deprecated interface in enet_path, lasso_path
       new  716fbb0   much better input validation, test that warning is raised on (n_samples, 1) y
       new  2c10151   FIX: sidebar.js
       new  880762a   Move Google javascript near </body>
       new  e0e900f   Merge pull request #2293 from amueller/warning_input_shapes
       new  74ec752   ENH speed up logistic_sigmoid (using less code)
       new  b0035bf   Rename cv_scores(_) back to grid_scores(_) to keep the name free for a future refactoring
       new  2df651b   Merge pull request #2299 from ogrisel/grid-scores
       new  9bffc8f   MISC: fix plot_multilabel example
       new  17ecff2   FIX make BaseSGDClassifier an ABC
       new  1f78b89   Merge pull request #2295 from larsmans/fast-sigmoid
       new  06a68ee   BUG: type->dtype in a cross-platform way
       new  76e1e86   compute offset in a cross-platform way
       new  a15d5cc   BUG: don't subtract offset in binary_tree
       new  fbc21a8   add explicit types to neighbors cython code
       new  5824bc2   WIP: explicitly mark all base classes as ABC with abstractmethod inits
       new  3be787e   Add concrete __init__ for LinearSVM
       new  44646da   Add concrete implementation for SGDClassifier
       new  742b416   DOC Remove deprecated reference + acknowledge @larsman
       new  4e56ff0   BUG: implement concrete __init__ for SGDRegressor
       new  d99bd2b   BUG: tests were raising the DataConversionWarning
       new  0b04d3c   Merge branch 'pr_2304'
       new  92b544f   Update typedefs.pxd with correct ITYPECODE
       new  8533d7c   MAINT: recompile Cython files
       new  9c1bc1b   DOC: add whats_new on the news
       new  203ee3d   FIX for MSVC inline fmin, fmax and log2
       new  e4a5444   FIX for MSVC inline fmax in dist_metrics
       new  11d1810   rearrange permutation_score parameters to match previous ones.
       new  c4e1f56   TST: adjust test relying on change order
       new  b5c1490   Add LibSVM random state to changelog
       new  e3a764d   FIX backward compatibility was broken
       new  89011e3   MISC: deprecate balance_weights (it's internal)
       new  fc0b766   REL: 0.14a1 Release candidate for 0.14
       new  e24af1e   What's new: put major 0.14 additions in first
       new  401668d   Ignore the generated MANIFEST file
       new  8fb2e7c   Fixed a typo in a contributor's name
       new  8ea2f26   Fixed a typo in a contributor's name
       new  a906f52   Also clean the dist folder when calling make
       new  937b06a   Add missing credit to @smoitra87 for the Python 3 support
       new  0bc1737   Re-align the what's new file with the new ordering of items from master
       new  a186bc4   DOC fix comment on svm probability param
       new  529655e   partial_fit for naive Bayes was done for 0.14-rc, not 0.11...
       new  1ae93a2   partial_fit for naive Bayes was done for 0.14-rc, not 0.11...
       new  3441e8f   Ignore the generated MANIFEST file
       new  ad48e84   Also clean the dist folder when calling make
       new  ec47860   DOC fix comment on svm probability param
       new  7048837   DOC Update what's new
       new  855a753   Merge pull request #2324 from arjoly/missing-contributions
       new  5cc1660   DOC Update what's new
       new  b1f9f11   DOC added link from banner to example.
       new  e91b23f   DOC added link from banner to example.
       new  bc785f8   DBSCAN BallTree implementation
       new  c30a062   fix and squash broken branch, remove all but section 3 and redo some docstring breaks
       new  3dd6a4e   DOC now building to html/stable
       new  e39759a   DOC home always points to stable
       new  8229823   fix broken directory name
       new  e917a39   Merge pull request #2329 from NelleV/website
       new  baf7b77   DOC credit to @ephes and myself for log loss in metrics
       new  728e61a   MISC: update whats_new
       new  8f493d9   MISC: fix reference to example
       new  5ebf551   DOC: DBSCAN misc doc formatting
       new  1467fb2   DBSCAN BallTree implementation
       new  9512d32   DOC credit to @ephes and myself for log loss in metrics
       new  9091c93   MISC: update whats_new
       new  96a2a66   MISC: fix reference to example
       new  b6c1555   DOC: DBSCAN misc doc formatting
       new  98b95ca   DOC fix example comments
       new  5b178e8   FIX: remove dupplicate entry in What's new
       new  1cb9f2c   FIX: remove dupplicate entry in What's new
       new  750d88d   ENH/FIX: improve logcosh function + tests
       new  b65ab86   FIX/ENH: revert changes, improve doc
       new  55a92e7   FIX: 80 characters
       new  9a4080e   COSMITS
       new  b4555cb   COSMITS
       new  421f2be   COSMITS
       new  953bebf   ENH: address discussion
       new  5bb8da1   ENH: use empty, not zeros
       new  96dcfd2   COSMITS
       new  2b60c81   Merge pull request #2290 from dengemann/more_ica_improvements
       new  1f20ec2   DOC: few small doc fixes to layout bugs on new website
       new  54da506   comments added to the changes
       new  b236030   FIX : backport tanh out param in old numpy
       new  5cadbe6   DOC typos involving Nyström
       new  a6f91b3   revert the tanh fix
       new  2119c75   Merge pull request #2331 from jaquesgrobler/master
       new  12a7545   DOC: few small doc fixes to layout bugs on new website
       new  c4e5165   comments added to the changes
       new  7def601   fix sidebar button size staying constant and leaving a mountain of white-space due to expanding/collapsing toctree
       new  3fa71e9   fix cut-off ML map
       new  444a536   DOC now building to html/stable
       new  0aad47c   DOC home always points to stable
       new  c23bb69   DOC: also point installation menu to stable
       new  e24a307   DOC: also point installation menu to stable
       new  5eab030   DOC copyedit SGDClassifier docstring
       new  c9ba2c3   DOC copyedit SGDClassifier docstring
       new  4644209   BUG: FIX Crammer-Singer formulation in the binary case.
       new  d71b013   Merge pull request #2336 from jaquesgrobler/DOC_fix_whitespace_on_userguide
       new  bd64ec3   Better test for auc_score.
       new  053723c   Optim for Crammer-Singer formulation in binary case.
       new  80b0bfe   Completely avoid for loop in _auc.
       new  a390a07   commit added for hack-warning
       new  4da05d8   Merge pull request #2337 from jaquesgrobler/DOC_fix_cut_ML_flowchart
       new  3c257fd   ENH : simplify ProabilisticPCA covariance_ computation + misc in pca.py
       new  ed936e5   ENH : avoid extra alloc in _infer_dimension_ in pca.py
       new  b1098b5   FIX : self.component in PCA was changed by fit when passed mle or explained variance
       new  238b330   ENH : avoid extra allocation in FactorAnalysis
       new  ab25660   ENH : use math.log and not np.log for scalars
       new  801e770   doc fix
       new  88dc9f6   misc
       new  20a8072   FIX : broken fit_transform
       new  fd60787   Merge pull request #2338 from agramfort/pca_cleanup
       new  4b658aa   DOC add link to fetch_covertype to covertype narrative docs
       new  2c48bad   DOC improve docstrings in sklearn.base
       new  0872592   Merge commit '0.14a1-20-gc9ba2c3' into releases
       new  9a44eb3   Merge commit '0.14a1-239-g0872592' into dfsg
       new  1e6e59a   Merge branch 'dfsg' into debian
       new  72a8de1   changelog entry
       new  3898ece   Converted Markdown style link to restructured text
       new  d34120b   Merge pull request #2345 from rgbkrk/patch-1
       new  07c15de   Typo.
       new  e387e7c   Converted Markdown style link to restructured text
       new  bec85af   MAINT: remove sklearn.test()
       new  987e2a0   MISC: deprecation notice
       new  276af2a   first carousel version added
       new  49a8344   firefox fix and more images added, auto-cycling disabled
       new  d5590d0   arrows switched for dots
       new  ab88461   have images link to relevant examples
       new  5074c04   slight layout adjust
       new  e0f5c66   debian/control - python-imaing to build-depends (for documentation) and removed not needed XS-DM-Upload-Allowed
       new  1962149   small layout changes for firefox, images taken from generated images now
       new  32b6dc1   indentation fixes
       new  910aa98   add more examples and cropping to first image
       new  8c07e8b   disable carousel for small displays, small tweaks
       new  7c6e398   Let's upload to experimental for testing
       new  5709b79   DOC: reduce the number of examples
       new  49a416a   first carousel version added
       new  21b51fa   firefox fix and more images added, auto-cycling disabled
       new  4876f68   arrows switched for dots
       new  9c6e849   have images link to relevant examples
       new  5bb0097   slight layout adjust
       new  c5c7901   small layout changes for firefox, images taken from generated images now
       new  af6c2ad   indentation fixes
       new  12e07ea   add more examples and cropping to first image
       new  d06abca   disable carousel for small displays, small tweaks
       new  ae055c0   DOC: reduce the number of examples
       new  6b1fbba   MAINT: remove sklearn.test()
       new  9c1a577   MISC: deprecation notice
       new  d4c3940   Typo.
       new  7078f3f   DOC add link to fetch_covertype to covertype narrative docs
       new  47e390e   MISC: document sklearn.test deprecation
       new  b4e22b6   ENH: custom distutils clean command
       new  b317121   MISC: document sklearn.test deprecation
       new  855bc9e   ENH: custom distutils clean command
       new  f72d803   ENH added an orange cite us button on the front page
       new  113be0c   FIX cite us buttong made blue bar span too much
       new  4a01681   DOC added testimonials
       new  7073c67   FIX forgot evernote's logo
       new  ad51c5a   ENH added telecom to the testimonials
       new  c7cd047   DOC updated evernote's testimonials
       new  2f7c944   ENH added AWeber's testimonial
       new  5320633   ENH added carousel back on front page for testimonials
       new  b2af93e   ENH better spacing on the first page
       new  8a920fc   ENH testimonials img are now centered.
       new  ec0f575   FIX typo in testimonials
       new  e8cbe20   Polishing on "Who's using scikit-learn"
       new  bfa656b   DOC: layout tweaks
       new  ad0ae75   ENH added an orange cite us button on the front page
       new  17707b2   FIX cite us buttong made blue bar span too much
       new  73e20ae   DOC added testimonials
       new  6c79db7   FIX forgot evernote's logo
       new  8395f70   ENH added telecom to the testimonials
       new  3cad446   DOC updated evernote's testimonials
       new  afa7345   ENH added AWeber's testimonial
       new  3a5248d   ENH added carousel back on front page for testimonials
       new  c1d91d7   ENH better spacing on the first page
       new  f98e54f   ENH testimonials img are now centered.
       new  002fda7   FIX typo in testimonials
       new  222fee3   Polishing on "Who's using scikit-learn"
       new  e143f68   DOC: layout tweaks
       new  10fcbdc   DOC: bigger menu fonts
       new  dc94662   DOC: bigger menu fonts
       new  a25e436   merge conflict fix
       new  9b5dff1   redo old deletions
       new  cf73a9e   ENH more explicit name for auc + consistency for scorer, fix #2096
       new  06be306   DOC put the narrative documentation of roc_curve and roc_auc_score in one place
       new  47a6624   FIX search and replace misstake
       new  19a3316   Website: bottom buttons
       new  0d3413c   DOC: button layout tweak
       new  b6be7bd   Website: bottom buttons
       new  c69201a   DOC: button layout tweak
       new  4cb0024   Merge pull request #2348 from arjoly/deprecate-auc_score
       new  1f0815b   ENH more explicit name for auc + consistency for scorer, fix #2096
       new  ef3da88   DOC put the narrative documentation of roc_curve and roc_auc_score in one place
       new  5902a9e   FIX search and replace misstake
       new  e0f288f   contributors logos added to index screen footer
       new  5dda942   changed hyperlink to funding anchor - added same for the images
       new  c2f828c   old footer removed for index page
       new  f953723   removed little underscores appearing between images - caused by link added to images
       new  5637fb2   TST: avoid a crash in Windows + Anaconda Py3.3
       new  1748d0e   TST: avoid a crash in Windows + Anaconda Py3.3
       new  d518171   FIX integer types in Ward clustering
       new  c2b8b5d   FIX integer types in Ward clustering
       new  372f59e   MISC: fix wrong timing in example
       new  2943786   TST: avoid nose running sklearn.test as a test
       new  c3bcbbd   MISC: fix wrong timing in example
       new  6bb4cd7   TST: avoid nose running sklearn.test as a test
       new  f67b21c   MAINT: randn on float is deprecated
       new  a6439d8   MAINT: randn on float is deprecated
       new  33cf98c   MISC: deprection is in 2 releases
       new  921b29f   MISC: deprection is in 2 releases
       new  d8bd3d3   DOC: update documentation for release
       new  5e95e40   DOC: fix CSS bug
       new  56fa23c   DOC: fix CSS bug
       new  6f6a3fd   MAINT Update mailmap
       new  95e433c   MAINT Update mailmap
       new  d13928c   REL: 0.14 release: update whats_new and version
       new  4323c11   REL: update whats_new and version
       new  af5e297   MAINT/DOC: bump docs and rev numbers to 0.15-git
       new  3d68c33   DOC: link to documentation, not main page
       new  934108b   Merge tag '0.14' into releases
       new  608cd63   Merge branch 'releases' into dfsg
       new  631b7c8   Merge branch 'dfsg' into debian
       new  ff2608a   fresh changelog
       new  c8cbd9c   DOC: link to documentation, not main page
       new  9202d52   refreshed joblib patch
       new  475aaeb   debian/rules - overload dh_auto_clean to call "clean" without -a (see https://github.com/scikit-learn/scikit-learn/issues/2351)
       new  5cb9f5a   debian/control - require sphinx 1.1.3 due to problems with unicode in earlier present on older Debian/Ubuntu's versions
       new  9285800   multilined Build-Depends
       new  2722755   dropping swig from build-depends
       new  314f296   ENH: Move sphinx&Co to build-depends-indep
       new  36cde78   boost policy to 3.9.4
       new  43005e7   dropping use of python-support
       new  70ff5e8   Update my URL.
       new  40b42e1   Added unit tests for parameters in isotonic regression
       new  4dbc33c   Fixed style issues as indicated by pep8
       new  c71685e   Fixed passing y_max and y_min to IsotonicRegression in test_isotonic
       new  508f82b   MISC : pep8 and pyflakes on test_isotonic.py
       new  60da310   MAINT deprecated ENet param used in doc and benchmark
       new  a82b315   MAINT remove deprecated parameters (the easy cases)
       new  fcc6c54   FIX rm spectral_embedding import from sklearn.cluster
       new  4ee4fe6   MISC: switch line returns back to unix
       new  8546776   BUG: restore setup.py clean functionality
       new  34c4908   REL: Release 0.14.1
       new  34b8592   RF: simply do not build sphinx documentation on systems with older sphinx
       new  f6c0bd6   Merge tag '0.14.1' into releases
       new  cd6d7d2   Merge branch 'releases' into dfsg
       new  18579d5   Merge branch 'dfsg' into debian
       new  264de0d   The freshiest upstream release - addressed 'clean -a' issue, thus removing the dh_auto_clean override
       new  980a8a6   if available cython >= 0.19 -- re-build Cythonized sources while monitoring for changes using dh_autoreconf
       new  00871e3   DOC typo
       new  af549f2   FIX + COSMIT Reuters out-of-core example
       new  4e8ac10   FIX py2.6 compat in biclustering example
       new  31ad968   MAINT remove Counter from fixes; no longer used
       new  90988fd   [DOC] Remove that RBM's are not implemented yet
       new  500f30f   Merge pull request #2361 from rolisz/patch-1
       new  1627a92   COSMIT refactor Hungarian algorithm
       new  40683d5   DOC while we're at it, link to our own RBM docs
       new  69f49b1   FIX: release reference to X in _tree.splitter
       new  68478ef   Merge pull request #2363 from glouppe/tree-memory
       new  2efa979   DOC: fix metrics documentation format
       new  c2cf21d   FIX+TST non-consecutive or duplicate vocabulary indices
       new  c041d69   Better docstring for KMeans.predict.
       new  211944a   Move estimate_bandwidth test to its own function.
       new  130729b   Add predict method to MeanShift.
       new  ce1fdf2   Add predict method to AffinityPropagation.
       new  c5dd039   Add missing docstrings to test functions.
       new  0551acb   Add what's new item.
       new  e3596a0   Remove warning in AffinityPropagation.
       new  1ac7c58   Fix grid search test.
       new  789acd6   More user-friendly error message.
       new  057db3b   COSMIT: change variable name.
       new  2aa3e58   DOC adding description for handling of the auto parameter in neighbors
       new  73de584   Added cosine distance metric for sparse matrices
       new  1428f6d   added missing assert in unit test
       new  fc91c33   Fixed doc string; compute cosine distance without copying matrix.
       new  41d0fb8   DOC make DictVectorizer docstring refer to FeatureHasher
       new  b46ea71   FIX remove warnings from univariate FS
       new  81c8557   DOC: Simpler cross-validation iterator doc
       new  dcf827e   Merge pull request #2368 from emsrc/cosine_distance
       new  9033baf   Merge pull request #2369 from larsmans/no-warnings-in-fs
       new  36ba287   TST: fix test on scipy dev version
       new  006a6cb   ENH add VarianceThreshold feature selection method
       new  23d2b98   Add .bmp sklearn logo of correct width for Win setup
       new  88a9e41   2-fold => 2-fold cross-validation
       new  a43e309   approximately the same percentage of samples
       new  23e9e1e   Remove spurious print statements in sample snippets to make the doc easier to follow
       new  7426ef3   Lowercase y and more consistent blank line usage
       new  d334176   MISC: switch line returns back to unix
       new  e7a9331   BUG: restore setup.py clean functionality
       new  1175352   MISC: fix rst
       new  2ddea88   Updated with more explanatory text. Also changed to use train_test_split function
       new  716bece   Corrected some typos, and added more explanation for the precision recall example.
       new  3af3bb2   Clarified a few wordy sentences
       new  392866f   Attempted a clearer explanation of precision-recall score
       new  5e3bc74   Adjusted to fit 80 character columns, and corrected definition of precision
       new  bf946fd   Removed random_seed argument to svm.SVC
       new  c7d0112   Added contributions to whats_new
       new  4168576   Further expansion of the precision_recall curve explanation, spelling corrections, and general cleanup.
       new  56f0b86   Added random_state argument to svm.SVC
       new  8b2d9eb   Added links to cross_val_score, auc_score, and recall_score
       new  d811298   Added link to precision_score
       new  f066c52   Updated whats_new and fixed a typo in plot_precision_recall.py
       new  d085b85   WEBSITE add yhat testimonial
       new  8ecd509   DOC: missing documentation of splitter parameter in tree.py
       new  257e100   Merge pull request #2385 from ndawe/tree
       new  aa251cd   DOC: removed duplicated Python 3 section
       new  12c44b7   DOC: docstring for predict_log_proba was wrong
       new  8c21817   Typo
       new  b952cdb   GBRT checks if ``loss`` is in self.supported_loss
       new  fd3a816   renamed supported_loss to _SUPPORTED_LOSS (constants) add test for deprecated warning
       new  f2f67e2   Merge pull request #2308 from pprett/gbrt-check-supported-loss
       new  3d01cf5   COSMIT: fix docstrings in affinity_propagation
       new  2240e44   ENH: add fast_dot function
       new  19e0401   employ DataConversionWarning
       new  3f3fb83   FIX: avoid zero-dvision warnings
       new  eeaa6e1   FIX: fix test warning
       new  bf16551   better handling of warning in test
       new  67ceee1   COSMITS
       new  e98d1cc   FIX: workaround for missing BLAS
       new  5a3a432   ENH: return np.dot if ndim does not match
       new  bc15f30   CLEANUP
       new  69fc297   FIX: remove spurious test
       new  dbc751f   ENH: add fast_dot shape check
       new  dcd6292   ENH: cover another cornercase + fix messages
       new  b86f3d5   ENH: add new Warning class, improve tests, update docs
       new  f36cba5   FIX: superfluous check
       new  6c30677   ENH/DOC: add performance tips on fast_dot
       new  005d210   COSMITS
       new  f6a1c5a   update what's new
       new  7808e09   FIX: proper doctest
       new  1322da9   ENH/FIX: address @ogrisel comments
       new  c9ff080   FIX: move inline comment to its right place
       new  83f48e4   FIX: performance.rst doctest
       new  f91fba2   ENH: restrict fast_dot to  np < 1.8
       new  d5c3933   ENH: add single_linkage
       new  d393f48   ENH: single linkage with Cython
       new  9149b1c   ENH: speed up single linkage
       new  4cd9c49   ENH: start of complete linkage
       new  6bae11f   ENH: first sketch of weighted linkage
       new  df8ccf7   BUG: fix bug introduced in Ward
       new  8674fd4   ENH: add average linkage
       new  30bedca   BUG: remove debug print
       new  3e54eab   WIP: MST for single linkage
       new  35bc01a   Cosmit
       new  49954dc   MISC: refactor Ward and linkage in same object
       new  5a5d944   ENH: hierarchical clustering example
       new  4fdd3ee   DOC: nice hierarchical clustering example
       new  4333ec7   API: rename to agglomerative clustering
       new  8daa93a   Add cython code
       new  6ff6b34   MISC: manual merge of master
       new  965a742   MISC: Take in account @NelleV's comments
       new  e2f0314   FIX spelling mistakes and whitespace nitpick
       new  b8b78d5   PEP8 fixes on hierarchical.py
       new  912ea55   ENH improved the documentation of the fix_connectivity function
       new  1e61f4d   FIX deprecated the copy arguments in hierarchical clustering
       new  0e9e579   FIX a convert on list was applied twice on the inertia matrix
       new  aa216b3   DOC added the docstring to linkage_tree and AgglomerationClustering
       new  cf397fb   ENH Removed copy option and deprecation on new functions and classes
       new  108271f   FIX spelling mistake
       new  54897c9   MAINT deprecated ward class
       new  73591a1   TEST test that ran with ward linkage now also run with average and complete linkage
       new  1b42853   TEST: increase test coverage
       new  732003e   MISC: move fast_dict to utils
       new  0e799b7   DOC small fixes on the hierarchical clustering
       new  5924270   DOC improve narrative docs on hierarchical clustering
       new  48a5231   TEST FeatureAgglomeration does not behave like normal clustering
       new  4109e34   ENH AgglomerativeClutering now supports different metrics
       new  d3eb06d   DOC/TEST improved failing errors and docstrings on metric
       new  eb54dc3   ENH now used  instead of euclidean distance to gain speed
       new  ad40483   DOC/TEST improved doc and tests on the paired distances
       new  795576f   MISC: new formulation in message
       new  b36dcf4   COSMIT: remove unused import
       new  a8ee6f9   ENH: add single_linkage
       new  511748f   ENH: single linkage with Cython
       new  8cf4be9   ENH: speed up single linkage
       new  58d2558   ENH: start of complete linkage
       new  3f24dc3   ENH: first sketch of weighted linkage
       new  0369c2b   BUG: fix bug introduced in Ward
       new  88ae7a4   ENH: add average linkage
       new  7cd77d3   BUG: remove debug print
       new  2eb7e55   WIP: MST for single linkage
       new  a1c9583   Cosmit
       new  3aa9158   MISC: refactor Ward and linkage in same object
       new  2077779   ENH: hierarchical clustering example
       new  7aa627b   DOC: nice hierarchical clustering example
       new  f95502d   API: rename to agglomerative clustering
       new  22eda02   Add cython code
       new  75a34a2   MISC: manual merge of master
       new  017de2e   MISC: Take in account @NelleV's comments
       new  638d590   FIX spelling mistakes and whitespace nitpick
       new  481f7fe   PEP8 fixes on hierarchical.py
       new  e937f4e   ENH improved the documentation of the fix_connectivity function
       new  057bc27   FIX deprecated the copy arguments in hierarchical clustering
       new  ed8bc97   FIX a convert on list was applied twice on the inertia matrix
       new  be8edeb   DOC added the docstring to linkage_tree and AgglomerationClustering
       new  ec08281   ENH Removed copy option and deprecation on new functions and classes
       new  7e0aff4   FIX spelling mistake
       new  00a4397   MAINT deprecated ward class
       new  4352d23   TEST test that ran with ward linkage now also run with average and complete linkage
       new  cf5b4ba   TEST: increase test coverage
       new  d5ac03c   MISC: move fast_dict to utils
       new  0f53ede   DOC small fixes on the hierarchical clustering
       new  a45d23f   DOC improve narrative docs on hierarchical clustering
       new  5688ef6   TEST FeatureAgglomeration does not behave like normal clustering
       new  61e1fdc   ENH AgglomerativeClutering now supports different metrics
       new  2ee7315   DOC/TEST improved failing errors and docstrings on metric
       new  805fd1c   ENH now used  instead of euclidean distance to gain speed
       new  3ce669f   DOC/TEST improved doc and tests on the paired distances
       new  7a9844b   MISC: new formulation in message
       new  57cc2f1   Merge branch 'hc_linkage' of github.com:GaelVaroquaux/scikit-learn into hc_linkage
       new  1eade9a   ENH: better error message
       new  599ba2d   FIX: ghost diff
       new  fbd6d9c   ENH: update version checking
       new  c70a63d   ENH: add the cosine distance to paired distances
       new  baa8ba4   Merge pull request #2288 from dengemann/fast_dot
       new  592e330   ENH: put fast dot to PCA, ICA and FA
       new  1a4a167   ENH: different metrics in hierarchical cluster
       new  ffbead9   FIX: test doc np.allclose
       new  9657bd6   Merge pull request #2396 from dengemann/insert_fast_dot
       new  b62b0d3   tree export_graphviz: remove unused close parameter and close the file if out_file is a string
       new  e478f27   Merge pull request #2394 from ndawe/export_graphviz
       new  ef513e3   DOC clarify distances in KMeans' _labels_inertia
       new  ed846b1   BUG: Remove UTF8 character
       new  6dcfe69   MAINT: remove utf-8 headers
       new  e3583da   COSMIT skip some repeated computations in k-means
       new  2eddb1a   BUG fix whitespace in error messages
       new  44003f8   fix: typo - y_pred instead of y_true
       new  8811030   Added euclidean_distances_argmin
       new  a39ea4d   Chunking on both arrays in euclidean_distances_argmin
       new  b8d29de   Improved tests for euclidean_distances_argmin
       new  148d4a4   Wrote docstring for euclidean_distances_argmin
       new  e0011e4   [BUG] safe_asarray() converts sparse matrices dtype
       new  65a7c79   Used safe_asrray() in check_pairwise_arrays()
       new  34bc6b2   Change signature of euclidean_distances_argmin()
       new  fe76818   Added pairwise_distances_argmin()
       new  9c6aab4   Optimized euclidean case in pairwise_distances_argmin
       new  5a16ed7   Removed euclidean_distances_argmin
       new  85b241c   Minor code cleanup
       new  1cd73c6   Added gen_batches() to utils
       new  8a65ce2   Changed pairwise_distances_argmin API
       new  14124ba   Code cleanup in pairwise_distances_argmin
       new  5626b7f   [DOC] Fixed pairwise_distances_argmin docstring
       new  711bfd2   [API] Renamed pairwise_distances_argmin
       new  220e9d0   [DOC] updated pairwise_distances_argmin_min doc
       new  31f0270   Performance enhancement for non-euclidean metrics
       new  7fac203   Small fixes
       new  a54358d   WIP
       new  e8bad3d   DOC ASCII only in docstrings
       new  41b3ddd   ENH speed up NMF (about 30% off topic extraction runtime)
       new  dedc822   ENH export randomized_svd publicly
       new  6a885cc   CSR matrix support in pairwise_distances_argmin_min
       new  357bad6   Added argument metric_kwargs
       new  2d9e41a   Merge pull request #2210 from pgervais/distances_argmin
       new  54c6bec   Revert "CSR matrix support in pairwise_distances_argmin_min"
       new  cb89679   Use pairwise_distances_argmin_min function.
       new  70695ca   Use pairwise_distances_argmin_min in examples.
       new  3d772e9   Add pairwise_distances_argmin.
       new  5ce1ae5   Add tests for pairwise_distances_argmin.
       new  959485a   fixed visual indentation errors
       new  bcfcc45   fix indentation errors
       new  0f1893d   use spaces in indentation
       new  34f23a0   fix indentation
       new  b392afe   fix E501: line too long
       new  9f60f76   fixed indentation and visual indentation errors
       new  93bc883   fixed visual indentation errors
       new  bd28db2   fixed indentation and visual indentation errors
       new  2fefd23   fix for 'line too long' warning
       new  200a355   pep8 fixes
       new  69d4148   fix indentation and convert tabs to spaces
       new  eec3a9a   fix indentation
       new  3e9971d   fix too many blank lines
       new  ca6b42c   fix visual indent
       new  616dda3   add expected blank line
       new  33507a1   fix line endings and visual indents
       new  fcce303   fix visual indentation
       new  b0de2e3   fix visual indentation
       new  b38366e   removed unused import from example
       new  c06fcff   Merge pull request #2410 from Balu-Varanasi/pep8_fixes
       new  b6f4393   Merge pull request #2411 from Balu-Varanasi/remove_unused_import
       new  c64163f   wip
       new  841fb88   ENH prettier output from NMF example
       new  d862bc0   [DOC] Typo fixes in documentation for Novelty and Outlier Detection
       new  d38b303   Merge pull request #2412 from rolisz/patch-2
       new  6ef08cf   FIX failing test in NMF due to negative zeros
       new  b9ca036   BUG: remove remaining utf-8 characters in docstrings.
       new  656b506   skip perfect checkerboard test
       new  76f64df   DOC broken link in HMM narrative
       new  fd1df8e   ENH micro-optimize pairwise_distances_argmin_min
       new  2fb55da   BUG duplicate finity check in input validation
       new  d473766   MAINT dedup Brandyn A. White in mailmap
       new  ceffed1   MAINT remove useless deprecation in sklearn.utils
       new  7b14020   BUG rename n_iterations to n_iter in TruncatedSVD
       new  efa74b4   FIX: renable test_k_means_plus_plus_init_2_jobs on OSX 10.8
       new  a81d534   ENH: add randomized_svd option to FactorAnalysis
       new  467e570   FIX: misc
       new  6d41219   misc
       new  1e117b2   COMSITS+DOC
       new  8651986   ENH: use `algorithm` keyword + expose failing test
       new  5a15e0d   FIX : fix randomized SVD in FA
       new  495e751   FIX: make SVD working again + add tests + update whats new
       new  09d925a   COSMITS
       new  a9250af   ENH/FIX: misc + address discussion
       new  8a97407   COSMITS
       new  37724de   ENH: add strong comparisons
       new  823d702   COSMITS: address discussion
       new  69a7421   COSMITS2
       new  33e0653   ENH: add check on init + appropriate test; COSMITS
       new  1cdbb25   ENH: address discussion + actuall make randomized default
       new  110d6d4   ENH: exposed n_iter; add missing fast_dot
       new  7ed5d86   ENH: improve wording + recommendation
       new  9cfb44f   COSMITS3
       new  48db715   ENH: add warning instead of print
       new  09dbdfa   ENH: deprecate `verbose` parameter
       new  6b719cf   ENH: optimze euclidean norms
       new  34bfaf6   Merge pull request #2406 from dengemann/tinker_fa
       new  9ad70ea   Edit typo
       new  76572e8   Merge pull request #2418 from rsivapr/typo-docs
       new  bef1b19   Merge pull request #2415 from kemaleren/skip_checkerboard
       new  0a8a6a1   ENH : add score to PCA from ProbabilisticPCA
       new  f9478b1   API : deprecating ProbabilisticPCA
       new  cb766fa   misc
       new  c09267e   typo
       new  0fdd7ab   ENH : make PCA scoring work with n_features is big. Avoid covariance computation in PCA.fit
       new  691c9e6   DOC : remove reference to ProbabilisticPCA in doc
       new  a3c6e61   update what's new
       new  615d2da   FIX : do not store precision_ attribute in score method
       new  68a4def   ENH : add get_precision method with matrix inversion lemma  to PCA for faster scoring
       new  daca399   ENH : add tests for get_precision in PCA + some fixes for corner cases
       new  1be9335   ENH : add get_precision method with matrix inverse lemma  to FactorAnalysis + use precision in score
       new  8145f52   pep8
       new  d024326   ENH : add PCA+FA model selection example
       new  c1ad0c7   pep8 + misc
       new  36ad48e   ENH : add score samples to PCA
       new  7d83cf3   API : add score_samples in FactorAnalysis + update example
       new  6e0eef0   better wording
       new  f62f1b7   add ref to online pdf
       new  bf1cf77   FIX : fix ProbabilisticPCA.score with homoscedastic=True
       new  4289f62   for loop in tests of get_covariance + get_precision
       new  4514bbc   typo
       new  2fb4979   typo
       new  0a3e585   wording
       new  84fa0ba   simplify FA tests after rebase
       new  ab015af   pimp example
       new  ce4b041   show example in narrative doc
       new  915a58b   avoid zip + topic in doc
       new  4971546   Merge pull request #2404 from agramfort/fix_probabilitic_pca
       new  4e0d186   DOC: add Rangespan testimonial
       new  775dac1   Fix broken link
       new  6c3e176   Remove execute pemission
       new  11a6b33   Remove shebang in library code
       new  5e375b5   Move __future__ import after license text
       new  36327a0   DOC: add new testimonial
       new  4cf42c8   ENH: Added comparison of other classifiers using partial fit methods
       new  5c7050c   Fixed misc style changes suggested by Ogrisel
       new  6460fec   first draft of scaling_strategies.rst - still wip
       new  4015b3c   improved scaling_strategies.rst, inc. working links
       new  3d862ab   added Federico as author
       new  9f325a6   cosmit pep8/257
       new  fad62b9   added link to user guide + small cosmit
       new  3765db6   fix re: ogrisel comments
       new  1c206d7   added note on learning rate evolution
       new  9ed0a6e   readded a paragraph that got lost in the rebase
       new  572b30d   COSMIT pep8/257
       new  5b8b616   COSMIT typos
       new  727ffd4   COSMIT typos
       new  d83f663   added links to plots in narrative docs
       new  b8e1a6c   ENH SGD Cython improvements
       new  f77bb9a   MAINT: Add COPYING and README to MANIFEST
       new  3072c66   Merge pull request #2424 from sergiopasra/no-shebang
       new  b154499   changing private _fit to public partial_fit
       new  825e2e2   added bestofmedia logo
       new  8b69152   added bestofmedia testimonial
       new  d49b347   cosmits
       new  bf04557   cosmits (2)
       new  a333542   DOC: fix bestofmedia testimonial
       new  1f3a03b   Cleanup
       new  6b45d13   FIX: free self.estimators_
       new  0f5f54e   Merge pull request #2430 from glouppe/forest-2414
       new  e18d11a   COSMIT more typos/precisions
       new  79dc288   DOC add missing Returns description
       new  cd950d1   DOC: Update pyamg site and fix a typo.
       new  0c84624   Merge pull request #2432 from jwkvam/docs_cluster
       new  f128072   Merge pull request #2370 from ogrisel/doc-cross-validation
       new  d0c981f   MAINT deprecate indices=False for cross validation generators
       new  0f5abaa   DOC boolean masks in CV generators are deprecated
       new  6955322   DOC what's new: fast_dot is internal, so don't mention it
       new  86ea13d   COSMIT rewrite precision_recall_fscore_support
       new  68e9bcf   FIX warning interaction problem, DOC clarify parameters
       new  fd8acb8   FIX ignore corrcoef warning
       new  eb806cd   Use label_binarize instead of LabelBinarizer
       new  a3f97ea   Use specialised warning class
       new  d66d537   TST messages for UndefinedMetricWarning from prf
       new  474d11f   TST use ignore_warnings helper
       new  5f5e96a   TST classification_report example that does not trigger warning
       new  513b0f4   Merge pull request #2278 from jnothman/prf_rewrite3
       new  4861ce6   FIX use fixes.bincount for numpy 1.3.0 support
       new  c760a2f   FIX Py2.6-compatible use of str.format
       new  74d3952   ENH bagging meta-estimator
       new  cc74076   ENH: move _partition_estimators to ensemble.base
       new  cebb7da   TST: test_base in sklearn.ensemble
       new  94bd0e9   BUG: do not force base estimators to inherit from sklearn in AdaBoost
       new  ccb7a6e   TST: fix random_state in test_weight_boosting
       new  5183828   DOC: base_estimator_ + estimators_ in ensemble.base
       new  9a90e13   COSMIT: tidy up forest code (inspired by bagging)
       new  524daee   Merge branch 'bagging'
       new  9c245f2   What's new: bagging
       new  1872aaf   DOC copyedit DBSCAN implementation notes
       new  8d17166   DOC reference for k-means++ in clustering narrative
       new  f8bf44a   FIX: logaddexp(-inf, -inf) == -inf and not NaN
       new  d5d7718   Merge branch 'master' of github.com:scikit-learn/scikit-learn into logaddexp-fix
       new  6454b5f   DOC: add docstring to _logaddexp
       new  7ecae94   DOC: bootstrap=True by default
       new  7551a82   ENH reduce memory overhead of storing tree ensemble
       new  4331b47   Merge pull request #2438 from arjoly/tree-mem-overhead2
       new  0cda1dc   Update what's new
       new  bdf2048   MAINT Update six to version 1.4.1
       new  581d535   DOC output types should never be "array_like"
       new  9a9b9e8   DOC+COSMIT: typos, lots of them
       new  5ae399f   ENH: don't call astype when a copy is not needed
       new  60e91f5   BUG oops, *safe_* asarray
       new  cab6517   Add what's new entry for pairwise_distances_argmin_min.
       new  0cc3291   DOC improve encoding docs
       new  72334f1   removing NWS from list of stock prices
       new  680234d   MAINT remove deprecated code from CD
       new  255f57e   FIX finish deprecation of indices in CV
       new  67b143b   Partially revert "MAINT remove deprecated code from CD"
       new  8c987ef   FIX issue #1993: passing a multilabel indicator is no more noop
       new  92917e1   TST speed up biclustering tests
       new  4fcb24c   MAINT remove dead code from LibSVM
       new  faddb45   FIX : get rid of coef_init in Lasso in dict learning
       new  41dc21a   add about us to community block
       new  8765d95   Merge pull request #2350 from jaquesgrobler/DOC_inria_google_logos
       new  9cb3b96   Merge pull request #2222 from FedericoV/Out_of_core_example
       new  24216b5   Issue #2455: Make RBM work with sparse input when verbose=True.
       new  b3d9a30   Change RBM sparse format from CSC to CSR.
       new  4a2acd3   Cosmetic changes.
       new  eaa8a5d   Add issparse import.
       new  0ac17a5   Cosmetic changes.
       new  ad15405   ENH 10x speedup in dump_svmlight_format
       new  91ae7cd   DOC: NMF narrative: describe optimization problem
       new  9e85797   FIX #2372: StratifiedKFold less impact on the original order of samples.
       new  4c95f5e   Fix accidental doctest breakage.
       new  cf9788c   Instead of linking to NB, explain the problem inside the test itself.
       new  bbaa9d9   Avoid list, preallocate a numpy array for indices instead.
       new  17f62e6   Update comment with numbers for when we run with 800 samples.
       new  6c7e751   Add entry for #2372 to whats_new.rst
       new  fee660c   More detailed entry for StratifiedKFold fix in whats_new.rst
       new  d611c32   Revert "More detailed entry for StratifiedKFold fix in whats_new.rst"
       new  3aede7f   Revert "Add entry for #2372 to whats_new.rst"
       new  c60a4fd   Revert "Update comment with numbers for when we run with 800 samples."
       new  f7d1b6b   Revert "Avoid list, preallocate a numpy array for indices instead."
       new  38f703c   Revert "Instead of linking to NB, explain the problem inside the test itself."
       new  0ab5278   Revert "Fix accidental doctest breakage."
       new  1a79640   Revert "FIX #2372: StratifiedKFold less impact on the original order of samples."
       new  fc6456d   ENH: address @larsman comments on fast_dict.pyx
       new  0c79334   TST "confusion_matrix" was a duplicated key in CLASSIFICATION_METRICS and NOT_SYMMETRIC_METRICS dicts
       new  3a2173d   TST style fix: old_error_settings should be outside try-finally block to be safely used in finally statement
       new  f8c7ca1   FIX classification_report shouldn't fail on unicode labels in Python 2.x
       new  cb7b423   FIX: more accurate description of eta0 in SGDClassifier
       new  aa2d045   Merge pull request #2426 from larsmans/sgd-improvements
       new  7a5d4b4   DOC typo in SVM narrative
       new  b460ffb   ENH: added PresortBestSplitter
       new  1c0dc55   ENH: remove assert
       new  151725d   MISC: improve tests of fast_dict
       new  3b8c0a1   Merge pull request #2462 from kmike/metrics-unicode
       new  978737c   BUG: fix typing
       new  501e32b   BUG: fix types
       new  ba45736   Clean: Separate hierarchical code from fast_dict
       new  0a2518a   COMIST: shorten some long lines
       new  190c787   DOC: added entry to What's New
       new  8416b10   Merge pull request #2469 from glouppe/tree-presort
       new  19688d9   FIX: don't bias random feature selection
       new  dc36bf3   FIX: random number generator
       new  9c07ca3   FIX: module RAND_R_MAX
       new  cdf495f   TEST: improve test_distribution
       new  924bf5c   TEST: improved test_distribution
       new  2d16aee   FIX: use XorShift random number generator
       new  d8158f5   DOC: remove rand_r license
       new  f92c734   TEST: add tests for splitter="presort-best"
       new  6e95e99   Merge pull request #2474 from glouppe/tree-tests
       new  d7b6a2b   fix broken logo links on homepage and remove unused duplicate images
       new  f7026b0   Expanded docstring of the verbose parameter.
       new  605d34f   Check that verbose output is sound.
       new  7d9984b   Merge pull request #2442 from glouppe/tree-shuffle
       new  130601e   FIX: unsigned int -> UINT32_t
       new  e3002c5   FIX #2372: non-shuffling StratifiedKFold implementation and updated tests
       new  53788b0   Merge pull request #2463 from ogrisel/stratified-kfold
       new  3084d7b   FIX: skip numpy.dot + multiprocessing test that segfaults under recently updated OSX 10.8
       new  be71b38   FIX: broken doctest impacted by stratified CV and tree RNG changes combined...
       new  c98a649   Merge pull request #2457 from untom/rbm_csr_format
       new  c63bb62   FIX: use next(iterator) built-in instead of iterator.next()
       new  f343262   MAINT zap unused import
       new  e123352   DOC optimized PNG images with OptiPNG
       new  88f38c5   DOC remove qda and faces from website carousel
       new  a8c7ed5   DOC replace jquery by minified version
       new  984a9a4   Merge pull request #2479 from larsmans/website-speed
       new  6dfc416   FIX : cleanup coef_init in coordinate descent
       new  2ca4832   DOC typo: polynominial
       new  179c073   FIX : warm_start in coordinate_descent (was ignored)
       new  d6186e8   FIX : set deprecation to 0.16 in coordinate descent
       new  10b0480   API : set return_models to False by default in path functions in coordinate descent
       new  98ebde6   FIX Pipeline should raise ValueError for duplicate name
       new  98ecfb6   MAINT list @paulgb's full name
       new  e472e5c   FIX : get rid of convergence warning in test_coordinate_descent
       new  f244a28   ENH : Gaussian process for arbitrary-dimensional output spaces by @JohntheBear
       new  af4b940   ENH : simplify input checking in GP
       new  56c0631   update what's new
       new  0d2c8c1   DOC improve biclustering docstrings
       new  81950ba   ENH sparse matrix support in pairwise + optimizations
       new  a2a4257   ENH optimize NMF inner loop
       new  94d66a0   FIX max_features in CountVectorizer
       new  baad8de   COSMIT pyflakes feature_extraction.text tests
       new  1d0a3fd   DOC error in feature_extraction.text docstrings
       new  bb2d007   another speedup attempt for index.html: carousel images to use thumbnails instead, 1 blocking js removed for page (sidebar.js)
       new  e95533d   cleanup changes
       new  8160b8a   inital checkin
       new  67010c7   first plot working
       new  7f6e398   remove paren
       new  d0c3395   ENH: small optimization in PresortBestSplitter
       new  eade1b7   add lossless image compression for docbuilding machine
       new  09d1ce7   switch species_kde to thumbnail
       new  2cae5d6   logo lossless compressions
       new  c326be1   Fix typo in the Gaussian PDF
       new  6f3a805   Merge pull request #2489 from samuela/patch-1
       new  36d500d   working boxplot
       new  26dcdfa   add coverage test for travis
       new  8f9f658   coverall added
       new  015113a   remove whitespace, fix pip install
       new  add591c   FIX scipy API
       new  c1df89e   ... add the fix to import check
       new  a3995ec   Merge pull request #2497 from dengemann/extend_fast_dot
       new  2988a23   AdaBoostRegressor: fix redundant recalculation of error_vect.max()
       new  2c05daf   Merge pull request #2500 from ndawe/r2
       new  097874c   Merge pull request #2495 from jaquesgrobler/ENH_coverage_travis
       new  29916ae   ENH: Call plot twice per class label rather than for every point.
       new  94642a4   ENH: remove offset calculations
       new  d2305b6   ENH: use memset and memcpy when possible
       new  29fb370   ENH: make sort an inline nogil function
       new  b9384c9   ENH: make node_reset/split/value gil-free
       new  6a5a65d   ENH: use npy_int32 type
       new  d61ffe8   Merge pull request #2487 from jaquesgrobler/DOC_speed_up_frontpage
       new  e3211e1   Cosmit: move log_loss.
       new  3bf4474   pep8/257 fixes
       new  a540309   more pep8/257 fixes
       new  0974fb6   ENH: Call plot twice per class label rather than for every point.
       new  5df97f6   Merge branch 'plot_dbscan' of https://github.com/ericjster/scikit-learn into plot_dbscan
       new  cb58700   bug fixes in graph_lasso and moving non-custom log-likelihood definition to specific definition of cost
       new  33ae502   corrected sign error in _objective
       new  29de0c8   pep8 compliance
       new  f01e215   Merge pull request #2502 from ericjster/plot_dbscan
       new  ef108b0   MAINT add authors to validation.py and pairwise.py
       new  c117799   MAINT: use python setup.py clean in Makefile
       new  809378f   FIX parallel FeatureUnion.fit and fit_transform
       new  edf747e   DOC docstring for extmath.norm
       new  24af95a   ENH refactor squared-norms computation to extmath
       new  b7f482b   ENH use row_norms in KDE code
       new  55a2f29   FIX fit followed by partial_fit in multiclass SGD
       new  a6b8913   FIX: clean-build target does no longer exist
       new  a2b13bd   MAINT use subprocess.call, not os.system
       new  761dde6   COSMIT remove dead code in k-means
       new  a2d95d3   Merge branch 'refactor-squared-norms'
       new  551b6c2   ENH honor Y_norm_squared when X=Y in euclidean_distances
       new  9e10ec0   COSMIT use norm function in feature selection
       new  551625f   MAINT simplify f_oneway in feature selection
       new  cad1287   adds minutes + seconds to examples
       new  dde0ad9   cosmetic - shorten long line
       new  8a02b4b   add copybutton to code examples
       new  4710045   Making the results NaiveBayes example more explicit
       new  616e026   remove writing of minutes/seconds
       new  5c2a60f   remove duplicate line
       new  88185eb   cleaner toctree collapsing
       new  83ebcf9   remove unused images
       new  22ff18b   mention python.org in js file
       new  67eb1b8   Merge pull request #2515 from jaquesgrobler/add_minutes_examples
       new  3a770a2   Adding doc for bin_seeding parameter in cluster.MeanShift
       new  5539b87   Update linear_model.rst
       new  44cc36f   Merge pull request #2521 from dmedri/master
       new  a2580d6   Merge pull request #2516 from jaquesgrobler/add-copybutton-examples
       new  913e471   FIX #2481: add warning for bug in old numpy with unicode
       new  d581f97   added test cases to test_graph_lasso for alpha=0
       new  b98b71b   Fix bug identified in #1817, comment 17340049
       new  3294ec5   Fixes to HMM docstrings
       new  687452d   Update permutation_test_score docstring
       new  cdc81a7   Merge pull request #2527 from shoyer/permutation_test_score-docstring
       new  8ffee3f   Wrapping line in naive_bayes.rst and specifying bin_seeding argument as optional
       new  aedef90   deleted the .orig file in this commit
       new  73b1b82   Remove leftover print statement
       new  679a49a   Merge pull request #2519 from ankit-maverick/minor_docfix
       new  a207c70   Use a RuntimeError instead of a warning to avoid raising a ValueError randomly later
       new  530c2e4   Merge pull request #2504 from glouppe/tree-tweaks
       new  b807036   Merge pull request #2523 from ogrisel/skip-numpy-unicode-bug
       new  7443e49   deleted *.py.orig from branch
       new  f859b7d   Release the GIL at tree building time
       new  e3cc3dc   cosmetic remarks incorporated
       new  34a3d47   Merge branch 'glbugfixes'
       new  f66e30a   cosmetic remarks, pep8 compliance against trailing white spaces, spelling errors
       new  b303b37   clean up
       new  5b817a1   Merge pull request #2505 from rphlypo/master
       new  1e3cadf   Put more code under non-GIL block
       new  a99ed56   Clarification of the sequence length, per @ogrisel
       new  c9a10b7   Merge pull request #2525 from rmcgibbo/hmmfix2
       new  02e0267   Merge pull request #2517 from jaquesgrobler/simpler-collapsible-toctrees
       new  46c260b   Update dict_learning.py
       new  47db7d2   Merge pull request #2533 from samuelstjean/patch-1
       new  b66288e   DOC norm optimizations in what's new
       new  12c8aa3   MAINT remove mlcomp document classif. example
       new  c96c6ae   ENH optimize one more sq. distance in k-means
       new  668c197   Merge pull request #2528 from ogrisel/tree-nogil
       new  daca763   Add implementation of RANSAC algorithm
       new  d2c5f98   Add ransac function to __init__.py of utils package
       new  63137a4   Fix bug in residual determination
       new  a971649   Add step-by-step description of RANSAC algorithm
       new  26818ff   Add parameter description to RANSAC and modify API
       new  23d3430   Add example plot script for RANSAC
       new  d0d1074   Add simple RANSAC unit test
       new  87ad0da   Add unit tests for RANSAC is_model_valid and is_data_valid
       new  623e8fe   Add unit tests for RANSAC max_trials option
       new  354fe56   Return number of trials in RANSAC
       new  d7fdb14   Fix n_trials return value and add additional test
       new  1f3e7fb   Add n_trials return value to doc string
       new  cddbf55   Add test for stop_n_inliers parameter
       new  cec4ed7   Add test for stop_score parameter
       new  99d2624   Move ransac.py to _ransac.py to avoid nosetest namespace conflict
       new  5c5953f   Set numpy random seed for all test functions
       new  755d90c   Remove unused variable
       new  5ce892f   Add missing information predict method of estimator
       new  5b46b59   Skip iteration for empty inlier sample set
       new  fabf192   Move algorithm description to notes section
       new  223edc3   Add some reference papers
       new  44d98b3   Fix estimator parameter description
       new  f6b2608   Add empty line between parameter and return value description
       new  c74b1f8   Move RANSAC implementation to linea_model subpackage
       new  0b6c470   Implement RANSAC as estimator class
       new  c1cda27   Fix indentation
       new  0321cb4   Add score method
       new  0cd1374   Add RANSAC to __init__ of linear_model sub-package
       new  8e1e4df   Add author and license info in source file header
       new  598b0d8   Define source file encoding
       new  04c19e2   Fix deprecated naming of estimator object
       new  8c37a2b   Update unit tests with new estimator class interface
       new  940c117   Update RANSAC example script for new estimator class interface
       new  c36b6a5   Add doc string for predict method
       new  1a007e7   Add doc string for score method
       new  d25d3ab   Set default initialization parameters
       new  64999f6   Fix default parameters of RANSAC estimator
       new  eadebe8   Add missing trailing underscore for estimator attributes
       new  34a9a1a   Remove deprecated ransac implementation
       new  c7bb3cf   Add description for min_n_samples parameter
       new  77b466e   Move estimator_, n_trials_, inlier_mask_ initialization to fit method and use LinearRegression and Perceptron as default base estimators.
       new  bb81e85   Add extra line between doc string sections
       new  8ebe2be   Add support for absolute and relative min_n_samples
       new  18a59a0   Add default values for min_n_samples, residual_threshold
       new  8413efa   Include random_state utility
       new  bdabe89   Use random_state in test cases
       new  3dd7628   Explain behavior when no base estimator is specified
       new  b15a8fc   Fix test case function name
       new  7d4d9f8   Add specific test case for score method
       new  ae6ee97   Add specific test for predict method
       new  36390cc   Mark all parameters as optional
       new  fe201f8   Add missing reference papers
       new  eb1adfc   Set random state in fit rather than __init__
       new  5cf7ac5   Add specific exception description for ValueError
       new  2c72bf8   Change indentation of multiline if-statement
       new  a12eff1   Remove unused variable
       new  a702773   Use term callable instead of function in doc string
       new  76f3d79   Add more precise description of is_*_valid parameters
       new  30dff8c   Combine nested if-statements
       new  dfc48d4   Remove default perceptron base estimator for integer data type
       new  28f24e4   Add support for sparse feature vectors
       new  a9392e6   Fix indentation of multi-line if-statement
       new  42b0fe7   Remove trailing empty line after doc strings
       new  68cf5f9   Add test case for sparse feature matrix
       new  a8db5b1   Fix flake warnings
       new  5e786c7   Add test case for case without specified estimator
       new  4760978   Add test case for min_n_samples parameter
       new  f4991e6   Fix description of default base estimator behavior and add corresponding test case
       new  5bb9e8b   Add description of RANSAC to linear-model docs
       new  dc0f897   Move ransac example to linear-model folder
       new  1d501c3   Add meta-data title, author and year to references
       new  011b1ba   Fix indentation
       new  748f855   Change ransac docs section title
       new  0151c73   Remove academic detail description
       new  8ca857b   Remove duplicate detailed description from doc string
       new  54475d4   Change default parameters, so example works without specific parameters
       new  f529fd4   Add support for multi-dimensional target-values
       new  b301464   Add reference to narrative documentation for detailed description of RANSAC
       new  218c905   Remove empty line
       new  d4cd61d   Add more descriptive explanation for raised ValueError
       new  5116637   Fix bug in is_data_valid test
       new  4d17686   Add more specific test for is_data_valid and is_model_valid functions
       new  2c8f27b   Stylistic multi-line statement change
       new  ada4778   Improve ValueError description
       new  0b278f2   Remove email adress
       new  608c284   Fix typo
       new  58d5654   Set random state of base estimator as well
       new  3120a3c   Use MAD as default residual threshold
       new  64eb3c4   Use np.logical_not rather than tilde
       new  0378f8c   Extend RANSAC example with comparison to BaggingRegressor
       new  c9199ac   PEP8 indentation fix
       new  722a467   Remove double-space
       new  3342f40   Add more detailed description for score method
       new  3423f16   Improve description of base_estimator parameter
       new  160414a   Add tests for more types of sparse matrices
       new  3166e22   Derive Ransac also from MetaEstimatorMixin
       new  cf1b2b8   Document use cases of is_*_valid functions
       new  b20f13c   Improve and document n_trials_ attribute
       new  018a947   Fix parameter description of X and y
       new  69113d1   Change naming scheme of variables for consistency
       new  5be0d71   Fix return value description of predict
       new  e5e04b7   Fix y paramter description of score
       new  0e0db46   Change linear_model access and call explicitly with kwargs
       new  db75461   Remove bagged regressor
       new  bc45750   Add missing parameter of score method
       new  3753696   Unwrap lines
       new  338b9f6   Make sure min_n_samples is not larger than number of samples
       new  6ee0197   Remove _n_ from min_samples
       new  6ec1281   Sample subset without replacement
       new  ed19096   Add reference to max_trials
       new  5d26ad7   Rename min_samples
       new  a9afc73   Improve explanation of degenerate case
       new  8c079e9   Fix typo
       new  28569ec   Improve RANSAC example script
       new  0a0b5d9   Make sure dimension is correct for all estimators
       new  a01b871   Add note about computational cost of is_*_valid functions
       new  c2a2bfe   Rename RANSAC to RANSACRegressor
       new  a555437   Rename base_estimator to estimator
       new  ae1bea9   Use LinearRegressor as default estimator for all data types
       new  6e2fb82   Add RegressorMixin base class
       new  555bcc8   Raise ValueError for non-integer absolute min_samples value
       new  a5aba07   Rename base_estimator to estimator in docs
       new  8cae900   Clarify default estimator by explicitly instantiating the class
       new  4d5ce35   Remove unused outliers parameter
       new  20e45bf   Rename estimator to base_estimator
       new  9aeb896   Add additional floating point test for absolute min_samples value
       new  250c6a3   Use assert_equal, assert_less rather than plain assert statement
       new  32ca06b   Add RANSACRegressor to whats-new doc section
       new  af30cf8   Add test for default value of min_samples
       new  4dfbca1   Add test for invalid value of min_samples
       new  3e38738   Add test for custom residual_metric
       new  ecc7c4a   Add test for default residual threshold
       new  b859c96   Remove reference to web page
       new  b1aea57   FIX: make RANSACRegressor pass test_common
       new  a4db8e2   FIX: np.ones_like does not support dtype on old numpy
       new  c14a619   TST fix broken fast_dot test
       new  d208c13   COSMIT friendlier output from faster NMF benchmark
       new  02d9726   ENH micro-optimize NMF inner loop
       new  311172e   TST fix still broken test_fast_dot
       new  286d375   Add a estimator introspection check to test_common
       new  e42f211   ENH make sure all "SkipTest" calls have an error message.
       new  7e7dc78   don't import tests
       new  e24e391   FIX don't raise file-level SkipTests
       new  ccad81c   Merge pull request #2538 from ogrisel/test-test-common
       new  d82cf06   Merge pull request #2536 from larsmans/rm-mlcomp-doc-class
       new  35f5b4b   LOO is bad doc
       new  d4447e7   Update cross_validation.rst
       new  1d8f696   Update feature_selection.rst
       new  d43a767   Merge pull request #2546 from StevenMaude/patch-1
       new  48e0f8e   updated for typos and a few wording changes
       new  4e5c637   one last comment fix
       new  7430eba   Add plot titles and newstyle plt import in OLS vs Ridge example
       new  3290453   OPTIM: remove memcopy for X_argsorted in GBRT models
       new  29fa8a7   fixes mis-alignment of Documentation link on front page
       new  03259da   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  fc3fcb8   Add inline comment
       new  55efb67   Merge pull request #2524 from rmcgibbo/hmmfix
       new  84094f4   Merge pull request #2543 from johncollins/dev
       new  b01f960   Minor typo fixes in grid_search_text_feature_extraction.py
       new  0e218c6   Merge pull request #2550 from StevenMaude/minor-typo-fixes
       new  4f3722c   ENH: improve FA vs PCA example
       new  4516bf5   COSMITS
       new  101f0ee   ENH : better shrinkage range for ShrunkCovariance
       new  8922501   DOC sparse matrix support in BernoulliRBM
       new  13a8315   FIX sparse matrix indexing in BernoulliRBM
       new  8a0b317   TST clean up after ourselves in SVMlight test
       new  64de651   1st draft of merformance docs
       new  9593fa7   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  1543480   Merge master in hc_linkage
       new  9a48821   DOC Clarified the doc of the hierarchical clustering
       new  48c2c15   Improve the GMM PDF example.
       new  e628aad   FIX precomputed distances on the hierarchical clustering
       new  7210aac   ENH callable metrics now work for the hierarchical clustering
       new  ba80756   some substance to the perf docs
       new  30eb78d   FIX #1622: OPTIM: remove obsolete random_state instance in the Tree class
       new  1564765   BUG: Compare strings for equality, not identity.
       new  03926cc   Merge pull request #2558 from Jorge-C/patch-1
       new  01e110c   Merge branch 'pr/2556'
       new  b2c5f4d   Better title
       new  abe3d05   COSMIT micro-optimize norm computation in NMF
       new  2f998ca   DOC fix cross-decomposition docstrings
       new  8ad2f5e   ENH: Make forest prediction code more robust to mutations of the estimators list
       new  0bb6871   added influence of n_features; added throughput explanation/graph
       new  87180a8   adding change.org testimonial, logo
       new  e7a40dc   Constant output dummy classifier
       new  5e9a015   Minor doctest change
       new  f2ceb4f   Merge pull request #2565 from change/change_org_testimonial
       new  098f630   GBRT enhancements:   Moved verbose output code to VerboseReporter class.   Cosmit: logical structured code blocks   Refactored: moved init state to method   Added partial_fit to GradientBoosting   Fix: is_classification not available; n_features not n_features_   Tests   Add ``partial_fit`` to docs
       new  5dc3754   added EPSILON_FLT and EPSILON_DBL for almost equal impurity and fx value comparisons
       new  0412d4d   fiddled with EPSILON_DBL
       new  ecd606b   added monitor callback w/ early stopping support
       new  7eb94f6   fix: VerboseReporter remaining time in the case of partial_fit
       new  e7bf1aa   added ZeroEstimator
       new  d2c2daea  use StackRecord as elements of stack not 5 consecutive entries
       new  1192fe6   Tree: compute partition specific impurity and total impurity in criterion.children_impurity. Pass partition impurity to stack to avoid re-computation (saves some runtime). New tree parameter: complete; specifies whether complete binary trees are grows or if a greedy branch of max_depth is grown with at most max_depth + 1 leafs.
       new  cdef5d6   set EPSILON_DBL to old val of 1e-7
       new  d7e5802   add tests for complete
       new  3914a22   fix: use regressor in regression test
       new  7eb679d   Add stack class
       new  2bf8318   Handled exceptions in fit; Added tests
       new  e9cea77   PriorityQueue for best-first tree growing max_leaf_nodes instead of complete parameter
       new  173f6e4   fix: test renamed and check only on nr of leafs.
       new  5f09b5a   tree code now supports both c-style and fortran inputs
       new  a9c2f89   tree code now supports both c-style and fortran inputs
       new  79a67d0   updated after tree.pxd change
       new  766e11f   Tree ensemble classes don't enforce c-style inputs.
       new  bef7d40   use assert_equal if possible
       new  3f0c3f0   check value of max_leaf_nodes
       new  68d63b7   TST Tests for string labels, DOC Minor doc changes
       new  8fc0622   adding test to rbm
       new  3a2ef79   changing partial fit to be seperate from _fit
       new  edb4e8f   FIX website: css hickup
       new  139cc68   Added contributing subsection and fixed link to web doc
       new  9b47992   SGD and SVC duplicate example sorted
       new  ea7ad29   sort out plot_iris vs plot_svm_iris
       new  ed82f9b   some cosmetic commits and delete duplicate files
       new  1a86056   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  6bc4012   No whitespaces at the end of file
       new  aafb4c5   wip sparsify()
       new  bd2cd7b   common test for sparsify + fix for SGDRegressor
       new  0546d31   DOC Minor doc changes
       new  501230d   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  23331a9   Removed unnecessary attributes
       new  c9b6d14   sparsify test: add multi-class test too
       new  1bdb7b2   completed sparsity sections; added benchmark
       new  e86a580   cosmit link
       new  6241ac7   cosmit double backticks
       new  4052f64   precisions from @ogrisel
       new  f39c759   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  5f8f7a0   Merge pull request #2571 from edran/readme
       new  84f86c2   Only run make test when coverage is disabled
       new  d6e5462   Merge pull request #2574 from ogrisel/travis-speed
       new  7a6d653   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  edb4a2d   support 'zero' init, more tests for ZeroEstimator
       new  bc54b7b   @ogrisel precision on sparse + elasticnet tradeoff
       new  c6bae4c   link to benchmark in github
       new  28997f6   Returning an array mask using np.ma.getmaskarray
       new  c7c6999   fix for np.count_nonzero not present in Numpy < 1.6
       new  7b3e7c7   MAINT: make gen_rst more robust (minor)
       new  613cf8e   s/n_features/n_samples in doc/datasets/index.rst
       new  6f8186f   Adding a test for verifying the shape of imputed matrices
       new  067ace1   Using assert_equal routine from nose
       new  3fff76d   whats_new.rst
       new  f719d09   Merge pull request #2576 from ankit-maverick/issue2560
       new  d0cdcde   Test correctness of average_precision_score.
       new  63f7dd1   Merge branch 'master' of github.com:scikit-learn/scikit-learn
       new  b7f98ac   Remove some warnings.
       new  63aa0c4   FIX the option affinity wasn't used in the hierarchical clustering
       new  610679c   DOC metrics and hierarchical clustering
       new  6d09f05   Add docstrings to _auc and _average_precision.
       new  179d758   s/speed/latency/g
       new  4ae976b   added section on model complexity
       new  c5ce49b   fix: Huber loss function in gradient_boosting fails if negative_gradient is not called before __call__; now computes on-demand.
       new  623dfef   add huber loss bug fix to whats new
       new  bf033bf   Merge pull request #2566 from Manoj-Kumar-S/constant-output
       new  a150413   What's new: DummyClassifier now supports constant output
       new  87a01e1   now compares linear vs svr vs rf
       new  e62d196   Merge branch 'master' into gbrt-enh-stackrec-greedy
       new  8143e40   BUG: ValueError was assigned as local variable!
       new  c0e686b   COSMIT: no need for parentheses.
       new  9162dd5   More robust unit tests for fast_dot.
       new  be2894a   Remove warnings in ProbabilisticPCA tests.
       new  d6e722a   One less warning in QDA tests.
       new  112befd   Simplify fast_dot.
       new  dd15031   DOC explanatory sentence for svm grid search example
       new  24a06c9   Merge pull request #2583 from amueller/doc_rbf_parameters
       new  5944e1e   Remove print statement.
       new  281017c   Add lobpcg in possible solvers.
       new  31d747b   Fix fragile symmetry check.
       new  2a23fdc   Fix fragile test.
       new  215ca71   Fix typo in test.
       new  317e97a   DOC remove confusing comment from TruncatedSVD
       new  568c43d   ENH: add PolynomialFeatures preprocessor
       new  a8089b0   FIX: errors in kernel parameters for OneClassSVM
       new  d071ef1   address PR2585 comments
       new  eb22538   import pylab -> import matplotlib.pyplot
       new  217834b   TST: fix PolynomialFeatures test error
       new  2983ab9   TST: fix doctests & pep8
       new  f1618a5   BUG: ward does not take an affinity
       new  ecdedd5   DOC: adjust notes in PolynomialFeatures
       new  bced673   changing test
       new  dd84e10   added example and fullfilled section on influence of model complexity
       new  0371831   removed CV code to tune hyperparameters (better for CI)
       new  dc763a3   mentionned training throughput
       new  8e2d6a2   added ElasticNet to the narrative doc
       new  4427c7a   more descriptive legend for ElasticNet
       new  ae4d0ac   FIX limit warnings for recall_score, precision_score, f1_score, fbeta_score
       new  4c00042   reorder polynomial features
       new  f08d2c6   merge doc/whats_new.rst
       new  f372e12   fix ordering & docs of PolynomialFeatures
       new  6715860   TST: remove stray print statement
       new  c19dd19   DOC: example metric + hierarchical clustering
       new  af8904c   ENH: ElasticNetCV and LassoCV raise ValueError with multitarget outputs
       new  e804575   DOC: multiclass: make the warning more prominent
       new  e2b6cc7   DOC improve feature selection docs
       new  af43c88   COSMIT remove useless "if False" in kernel approximations
       new  da4f2c6   FIX regression in CountVectorizer handling of float min_df/max_df
       new  21f8fd0   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into prf_derivative_warnings
       new  a23662d   COSMIT use assert_no_warnings where appropriate
       new  e5f1cec   BF: use hasattr with providing attr name (Thanks to Timo Schulz)
       new  5393bd3   WIP example l1 metrics hierarchical cluster
       new  e5bed51   BUG: fix partial_fit of MiniBatchDictLearning
       new  86954e2   DOC: update whats_new
       new  13e43d2   Implement GBM's best-first heuristic tree growing procedure .
       new  6bd380e   remove asserts
       new  4b45668   Best-first instead of branch heuristic - identical score as GBM on covertype
       new  0808cfc   cleanup: remove BranchBuilder, added more comments
       new  ab14a08   remove check for condition that might not hold
       new  b3aff11   DOC explain ARPACK algorithm in TruncatedSVD docstring
       new  3a91193   cosmit
       new  24b6fb0   rm MSE import
       new  5847be4   removed Impurity struct, Criterion.child_impurity now returns impurity_left and impurity_right (not total), fix: moved pos >= end check to if branch
       new  635505e   moved data structures (Stack + Heap) to _utils use impurity_improvement for choosing best split
       new  9d485d7   fix: correct impurity_improvement formula (weight by n_left and n_right)
       new  cde390f   remove split_impurity from node_split signature
       new  9f23eb7   update test (slight changes in tree output because of removal of EPSILON insensitive checks)
       new  6aec00e   use max_leaf_nodes in regularization example
       new  e7a2c2e   narrative documentation on controlling tree size - might be relevant for weighted boosting and CART as well.
       new  cfa7775   DOC: typo in kernel formulas
       new  7ca3906   Merge pull request #2608 from jakevdp/master
       new  f63e5fd   Merge pull request #2585 from jakevdp/polynomial
       new  dd73b35   Disable reporting to google analytics -- we haven't got any permission from the users allowing to report back to 3rd party services (thanks Ambrose Andrews, Closes: #730259)
       new  96cd50a   FIX verbosity of the mds
       new  99e3245   Merge branch 'master' into DOC_olivier_tut
       new  36d68b3   MAINT: unreachable code removed from BernoulliNB
       new  119cc24   BUG SVMlight loader should check whether n_features is big enough
       new  642cbb2   MAINT: use stable blas getter API
       new  47f58d4   rename GBM_MSE to FriedmanMSE add friedman_mse to regression criterions in tree.py
       new  669e2b4   BUG lower space complexity of estimate_bandwidth to linear
       new  bbf3b57   Merge branch 'master' into DOC_olivier_tut
       new  dea7bbb   BUG: typo fixes in sklearn.mixture.gmm
       new  f642aee   Merge pull request #2614 from zyv/patch-1
       new  b5e197f   MAINT: fix broken links to numfocus.org on donations page.
       new  a7d19e3   Merge pull request #2615 from rgommers/fix-numfocus-link
       new  35daf32   prefer np.as(fortran|continuous)array over np.asarray(order=order) pre-offset X pointer (X_fx); doesn't help performance-wise
       new  c02dee5   fix: pass warm_start to BaseSGDRegressor fix: make Regressor test case extend CommonTest (move two clf specifc tests to clf test case)
       new  e0cb201   DOC be more explicit on mean-shift scalability
       new  80b96bc   Merge pull request #2617 from pprett/sgd-regressor-fix
       new  e85ac0e   some cleanup, conflict fixes, squashes, re-dos
       new  c224111   redo typo fixes
       new  e8b3d39   small css tweak
       new  9e1edd1   FIX stop forcing deprecation warnings for external packages
       new  61c3b6a   FIX use of str.format for Py2.6
       new  2373961   speedup for docbuild using joblib
       new  293c020   change cache directory
       new  d4fd210   Merge pull request #2620 from jaquesgrobler/master
       new  fda6211   pass impurity to ``impurity_improvement`` - now it is correct so that arbitrary Criteria can be used in best-first search.
       new  b583310   typo fix
       new  fbab2cd   COSMIT + FIX in _utils.pyx
       new  8524c64   changes to headings and sections
       new  c0e25bb   Merge commit 'fbab2cd' into gbm
       new  209c93a   remove face-recognition exercises, skeletons, etc and merge artifacts
       new  f0a7062   further changes to headings and subsections
       new  89c8625   re-order some Documentation categories in documentation.html
       new  feb4312   fix broken link for stable version from dev version
       new  4d01a77   FIX: unintended assignment to ValueError
       new  5055836   STY: tuple for Gael
       new  d2f3a8a   ENH: refactor warnings 1
       new  e1bdd99   ENH: add assert_warn_message
       new  f9cd599   ENH: revem tests_metrics.py
       new  3a6c034   ENH: more of this
       new  9b1207b   ENH: final fix: clear all regisitries
       new  a0f0a3d   FIX: PPCA related warnings
       new  9fd2be4   ENH: context manager + decorator
       new  a4e67d6   FIX: doc example
       new  163e659   ENH: address discussion
       new  4c10c82   ENH: address discussion
       new  c300814   rebase
       new  e141a98   cosmits
       new  686fc3a   FIX
       new  26a5dd7   FIX typo 2
       new  39fb5c5   address discussion
       new  0b62e57   remove import
       new  d8ad680   Improve doc
       new  ce0d52e   COSMITS + FIX to impurity_improvement
       new  da1c71e   FIX: impurity_improvement
       new  722afb2   TEST: disable test_32bit_equality
       new  12290f2   Update max_depth in DepthFirstSplitter + Release splitter in BestFirstSplitter
       new  7ec6531   Merge pull request #7 from glouppe/gbm
       new  65634f6   fix tutorials index, update dropdown menu in navbar, update stat_learning tutorial link on documentation.html page
       new  acc897d   typo fix
       new  8e458d4   Merge branch 'master' into DOC_olivier_tut
       new  4d175e9   X need not be continous anymore (since we adhere to col and row strides)
       new  0451b24   remove extra headings from tuts page
       new  8e41ba2   cosmit: doc undocumented functions
       new  ae1ea1c   update what`s new
       new  dc8a03e   BUG: fix convergence check in OMP
       new  8b037ad   Merge pull request #2541 from dengemann/fix_dot
       new  1eacd31   py3k syntax
       new  391c913   Merge pull request #2624 from dengemann/fix_dot
       new  2ff06a2   cosmit: remove commented out code - add comment
       new  c281adf   Add max_leaf_nodes to forest + DOC
       new  f172759   Merge pull request #8 from glouppe/gbm
       new  09d7e97   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  c025984   ENH: color scheme + add PCA
       new  93f2a9a   COSMIT
       new  2b1863b   address duscssion
       new  d990ad9   address duscssion 2 + expose modern API
       new  05188ef   ICA vs PCA barrier free + fixed
       new  efe5294   COSMIT
       new  5821143   TST add generic test for averaging
       new  9e56092   ENH test that most metrics work with one sample
       new  e90317d   ENH auc averaging for multilabel-indicator format
       new  5b9c369   ENH add multilabel-indicator support for average_precision
       new  5ddea8c   TST perform testing on average precision + simplify invariance test system
       new  3df6e8a   ENH add scorer support for multi-label threshold scorer
       new  afc82c5   TST scorer works with multi-output decision function
       new  7ee45a5   flake8
       new  205a422   TST more clean up test_metrics
       new  348a342   DOC narrative doc for averaging support of average_precision_score
       new  e9b4fcf   FIX decision function scorer multilabel
       new  e05cb16   DOC narrative doc for roc_auc_score multilabel-indicator
       new  873a58c   DOC not all functions support multilabel-sequence format
       new  c53e977   DOC wording
       new  87b2785   FIX doctests
       new  fbb62f4   DOC extend roc example
       new  8832bde   DOC update example of precision-recall curve
       new  eb56b9e   TST properly raise ValueError...
       new  e778c36   DOC nicer plot
       new  7bafbca   TST add roc_curve and precision_recall_curve on toydata
       new  a9aa616   DOC TST explain how to use common tests in test_metrics
       new  59fd981   Typo
       new  85e1db0   TST clean copy paste mistake
       new  303a465   TST remove print
       new  da1b56a   TST full coverage for _average_binary_score
       new  3d0211c   DOC TST typo
       new  19bf47e   WIP TST more sample invariance test
       new  1aa0f07   WIP TST adapt test for sample based metrics
       new  5cf711a   TST finish to add invariance test for sample based metrics
       new  41ecb9b   DOC typo
       new  3c4dc02   DOC wording and typo
       new  909d45d   TST add test for garbage averaging input string
       new  050f7ec   TST typo
       new  49186bb   DOC explain why roc auc score is useful in multilabel classification
       new  8c2c1c5   typo
       new  f3c09df   Improve wording + COSMITS
       new  925aa1e   Improve wording + COSMITS II
       new  2843ee1   manual subplots adjust
       new  527d5f7   Merge pull request #2628 from dengemann/improve_ica_example
       new  a82c8ac   Merge pull request #2460 from arjoly/auc-multilabel
       new  2e07f4e   FIX: broken sparse matrix check under scipy 0.13.0
       new  04977db   FIX: Python 3 dict keys cannot be concatenated with a list
       new  3e7f445   found a good elasticnet example
       new  7a01a5c   TST more corner case tests
       new  848eab3   TST ensure that exceptions are properly raised on np 1.3
       new  06a1eaf   Merge pull request #2629 from arjoly/fix-auc
       new  1b8ea52   max_leaf_nodes has precedence over max_depth if the former is not None.
       new  7c9d618   fixed link & desc in WN.rst
       new  2723270   added model complexity in latency/throuhputs example plots
       new  f4d32b2   added prediction time plot to Reuters example
       new  3b7bc1f   typos in narrative doc
       new  d101ef4   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  87343d7   Update what's new
       new  3b254d5   Joel's feedbacks
       new  6ec80b5   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  fc15658   FIX: random_start feature in GaussianProcess
       new  2912384   FIX: gp_diabetes_dataset examples (theta_ attribute)
       new  30578da   fix: l1_ratio is incorrect; its (1.0 - rho); added test case to make sure elastic net penalty with l1_ratio near to 1 and 0 matches L1 and L2 penalties.
       new  11d256c   added change log entry
       new  6a25717   set l1_ratio to (1.0 - l1_ratio)
       new  3721988   revert: reverted l1_ratio change - not SGDRegressor - sorry
       new  d27cbac   Merge pull request #2631 from pprett/fix-sgd-l1-ratio
       new  a1aa449   Fixing small typos in the docs.
       new  7abe96d   Merge pull request #2633 from josericardo/FixingTypos
       new  699079e   ENH: more explicit traceback + message in case of common failure on sparse input
       new  54b03b3   Merge remote-tracking branch 'upstream/master' into pred_latency
       new  cdf70f4   added model re-shaping + compressed verbose §
       new  4674552   Merge pull request #2592 from jnothman/prf_derivative_warnings
       new  a3a4b53   emph. reshaping benefits on I/O
       new  de7bee4   ENH support multiclass targets of string objects
       new  d025750   MAINT: use $(MAKE) for recursive make
       new  7643f44   DOC add sklearn.base to generated docs
       new  c94ea6a   ENH use fast row_norms helper in preprocessing.normalize
       new  21f0479   ENH use fast row_norms in dictionary learning
       new  66a5a4a   Merge pull request #2488 from oddskool/pred_latency
       new  b0485c4   BUG don't densify sparse matrix in BernoulliRBM.score_samples
       new  4706435   ENH speed up progress reporting in RBM
       new  fa5acb2   TST Add benchmarking script for multilabel metrics
       new  d02a15b   ENH better error message when CountVectorizer prunes away all terms
       new  be744a0   TST disable non-doctests in Comp. Perf. docs
       new  dcb2834   DOC {min,max}_df look at document freq, not term freq
       new  393a894   Merge pull request #2626 from jnothman/string_objects
       new  3aeae7b   DOC Add what's new entry for #2626
       new  8b6b770f9 DOC fix BernoulliRBM._fit docstring
       new  f2b91dc   Merge pull request #2642 from larsmans/rbm-speedup
       new  03352bf   ENH: faster cartesian product in make_classification and document complexity
       new  4b85423   Revert "ENH: faster cartesian product in make_classification and document complexity"
       new  912d550   Fix exception message when cloning a estimator that does not implement 'get_params' method.
       new  baac91c   Merge pull request #2647 from trein/patch-1
       new  84ec4f9   COSMIT in response to feedback
       new  5f57f85   Merge pull request #2643 from jnothman/bench_mutilabel_metrics
       new  9739224   Merge branch 'master' into gbrt-enh-stackrec-greedy
       new  88bff86   DOC: PolynomialFeatures narrative doc
       new  c66d3df   DOC: fix auto-links
       new  6cb1e3d   DOC: fix Polynomial Regression docs
       new  0bd0552   ENH: Raise explicitly on non-unique vocab.
       new  79c1c38   remove EllipticEnvelop deprecation
       new  b308a53   Merge pull request #2653 from jaquesgrobler/remove_ellipticenvelope_deprecation
       new  a64d5aa   Remove deprecated zero_one and zero_one_score
       new  34005e9   Merge pull request #2649 from jakevdp/polynomial_doc
       new  c80116c   flake8
       new  ded381b   Merge pull request #2654 from arjoly/metrics-0.15
       new  ffde690   DOC: fix test failure in linear model
       new  25ed3f5   Website: add spotify testimonials
       new  833a1f3   Merge pull request #2651 from amueller/spotify_testimonial
       new  20694a8   FIX silence numpy 1.8 warning for using non integer
       new  f86ed77   Merge pull request #2656 from arjoly/np1.8-warnings
       new  dba9ade   SPD: reuse variable in GaussianNB for speedup
       new  aa8139b   Merge pull request #2659 from jakevdp/gaussiannb_speedup
       new  9671b3c   ENH make minibatch k-means use batches in prediction and computing labels
       new  71d2392   FIX : explained_variance_ratio_ in RandomizedPCA
       new  4ce2eb7   update tests
       new  9b37285   better sparse support
       new  08469a6   pep8
       new  2b8d06b   update what's new
       new  0553bf1   add test with sparse data
       new  0413b55   saving the random state
       new  fe1c284   address @ogrisel's comments
       new  22322f1   fixed bug in mds.py
       new  8e58fcb   added not
       new  9d92982   Merge pull request #2667 from kowalski87/master
       new  8024ccf   Cleanup criteria
       new  547c9d6   Nitpicks
       new  8dabff1   Merge pull request #2664 from agramfort/fix_randomized_pca
       new  5b300e5   Merge pull request #2660 from amueller/mbkmeans_compute_labels_fix_new
       new  a030c5d   Normalize variance
       new  05cce23   DOC: add phimeca testimonial
       new  aac7e67   Merge pull request #2669 from GaelVaroquaux/phimeca
       new  44270dd   fix f_oneway with ints
       new  f31d6d7   pyflakes
       new  d4c5b18   Merge pull request #2671 from agramfort/fix_f_oneway_int
       new  263f754   Merge pull request #9 from glouppe/gbm
       new  6996bde   whatsnew conflict fix
       new  4f57a94   MAINT remove some dead code from the LibSVM wrapper
       new  3ee7b72   BUG: OneVsOneClassifier was broken with string labels
       new  81336ae   MAINT use TruncatedSVD in pipeline tests
       new  fedff32   COSMIT: simplify code
       new  6ec2c8b   Merge pull request #2674 from GaelVaroquaux/bug_ovo_string_y
       new  0d3e15e   FIX: Typo
       new  06984c2   FIX: Typo
       new  2f0c3bb   Merge pull request #2683 from blagarde/master
       new  a90e8a5   FIX: Typos
       new  8d7cfa0   Merge pull request #2684 from blagarde/master
       new  0a8538e   FIX: test failure under windows caused by temp files handling
       new  745e9fc   Fix OOB score calculation for non-contiguous targets
       new  f65b289   Merge pull request #2685 from bicycle1885/fix-oob-score
       new  c0b7882   DOC optimize Phimeca logo, size halved
       new  f02cd21   MAINT some more pointer safety in LibSVM
       new  48c9d65   MAINT import LibSVM 310
       new  23f2328   add a comment for sigmoid_predict in svm.cpp
       new  45a8ed3   MAINT import LibSVM patch from upstream
       new  7490e1a   Merge pull request #2673 from larsmans/libsvm-310
       new  cf7cc70   ENH filter out zeros early in FeatureHasher
       new  b32e16e   DOC: typo in binary_tree docs
       new  e2c0306   FIX PCA.score_samples didn't do input validation
       new  8be7553   DOC missing parts of docstrings
       new  674e0ea   FIX Do one run with MiniBatchKMeans and explicit centers
       new  c12d717   FIX MBKMeans w/ explicit centers and n_init>1, part 2
       new  49dc6b3   ENH show top term per cluster in doc. k-means example
       new  e4b2094   FIX n_init bug in k-means
       new  7a22b77   ENH make_pipeline and make_union utility functions
       new  5f3f794   MAINT refactor fast_dot
       new  8615684   ENH Upgrade to joblib 0.8.0a2
       new  c681e5d   ENH: reduce memory usage and IPC overhead when fitting forests by using the new threading backend
       new  43cb5a5   ENH Release gil in _tree.pyx:predict
       new  6a1d900   ENH Use backend="threading" for forest regression as well
       new  78326ab   DOC Update whats_new.rst to document threading backed for forests
       new  67c5532   FIX: broken sed command under OSX in joblib sync script
       new  72b5e65   ENH: parallize the BaseForest.apply method with the threading backend
       new  40024f6   ENH use nostests yield construct for better error reporting in common tests.
       new  e8848dd   enh give yielded tests nicer names
       new  9775be9   Merge pull request #2700 from amueller/refactor_common_tests
       new  a600205   FIX suspected bug for non-hypercube make_classification
       new  ba1a684   FIX Low-memory cluster sampling in make_classification
       new  f8a4a63   FIX don't leave clusters in uint8
       new  bd58cf9   Remove debug print statement
       new  f4af1ec   FIX update data-dependent test assertion
       new  33b8430   Use internal sample_with_replacement rather than Python's random.sample
       new  224a4d5   TST/COSMIT cleaner make_classification and fix tests
       new  08debe2   FIX: Ensure expected behaviour when copy=True|False in Imputer
       new  c931854   FIX: Don't copy X at imputation in the sparse case
       new  42ea5ec   COSMIT: remove useless import
       new  cd4f5af   Update docstring + Use np.diff
       new  fd2226d   Update dpgmm.py
       new  4508600   Update dp-derivation.rst
       new  b0abd2f   DOC Remove target_names from boston dataset object description
       new  da78718   Merge pull request #2710 from eshilts/DOC-load_boston
       new  e9e9480   DOC add missing comment for C parameter
       new  13ad2f9   DOC remove duplicate reference to example
       new  abc9301   Merge pull request #2702 from glouppe/imputer-copy
       new  4791c48   Merge pull request #2708 from likang7/patch-3
       new  f72ecb0   Merge pull request #2706 from likang7/patch-2
       new  60a0c2f   FIX: don't force finite in imputation
       new  2079732   FIX: use _assert_all_finite
       new  b57b3a3   Merge pull request #2712 from glouppe/imputer-copy
       new  a5210cf   Merge branch 'master' into gbrt-enh-stackrec-greedy
       new  f512bd8   renamed partial_fit -> fit_more for GBRT
       new  181c515   moved fit_more to warm_start=True
       new  6c033bd   fix: resize oob improvements, train score when early stopping
       new  0d412d2   Updating copyright year to 2014
       new  600be82   Update AUTHORS.rst
       new  9e9a656   Merge pull request #2714 from amormachine/master
       new  f9104f6   warm_start semantics now fit exactly n_estimators rather than self.estimators_ + self.n_estimators
       new  bae41e7   fix: use estimators_.shape[0] rather than n_estimators to make predictions; what if user just changes n_estimators of an already fitted obj.
       new  2dde9a5   fix: warm_start demo in narrative documentation
       new  0000583   updated whats new
       new  bb743dd   ENH speed up SVMlight loader using Cython's array support
       new  e0c635f   cosmit
       new  e9e1ea5   FIX remove ``.description`` from test generators as this is not thread-safe.
       new  2dbc996   ENH Define fused types so only two functions are generated.
       new  dd680d6   malloc / realloc checks correct rc values
       new  51763c8   Merge pull request #2719 from amueller/remove_description_in_common_tests
       new  dce1d49   cosmit: better errors
       new  a8bdbff   do not alter n_estimators after early stopping
       new  904a526   Merge pull request #2720 from amueller/chi2_fused_types
       new  36ab692   Pre-initialize all trees before dispatching
       new  b5b653e   Merge pull request #2724 from glouppe/forest-2721
       new  ca55b40   FIX Use Py2.6-compatible string formatting
       new  a19bbe3   ENH: make forests' test_parallel_train run faster
       new  4cf9865   Merge branch 'master' into DOC_olivier_tut
       new  b278b52   WEBSITE FIX spotify logo!
       new  8082309   Typo in examples
       new  e4c27d5   FIX: np.searchsorted numpy bug on unicode objects also impacts 1.6.1
       new  73c7de3   cosmit: better exception remove troublesome 32bit test
       new  9a3c22a   cosmit: better exceptions
       new  aaa20ca   recommit joel comments
       new  4e4a8a9   fix: wrong placeholder in format string
       new  792a09c   better test coverage in tree module
       new  46c4bbb   where to from here section
       new  0ee004d   whitespace removals
       new  05d88c7   GBRT subsection for warm_start
       new  b8bea8e   Merge pull request #2570 from pprett/gbrt-enh-stackrec-greedy
       new  99158e1   olivier suggestions
       new  8b9b1ee   fix: remove print from gbrt test
       new  ae3fe45   fix: use six.string_types instead of basestring
       new  834b375   uses raises instead of with assert_raises - apparently doesn't work on python3 nose version.
       new  ef6af87   changed PCA to truncated SVD
       new  60ba427   Testing log_loss and hinge_loss under THRESHOLDED_METRICS
       new  10e384c   FIX: Removed redundant code
       new  86b7051   TST: Test class variance and string input
       new  c9714b5   FIX: Changed str vs float invariance test
       new  8f2d8b9   Merge pull request #2717 from Manoj-Kumar-S/test_log_loss
       new  a876682   Add learning curve
       new  7b4b0e7   MISC remove accidentally commited file. Whoops!
       new  7de3d96   DOC add learning_curve to what's new and API reference
       new  1bceee0   FIX explicit type for offset variable
       new  8f71e46   Handle n_features < n_informative case in make_regression.
       new  b3a9da9   Regenerate C files
       new  9329577   Wrong format specifier used when formatting exception message.
       new  d32d7ed   Python 3 compat
       new  65460d7   Python 3 fixes in exercises
       new  67d3bb2   Python 3 fixes for the second exercise
       new  5847006   COSMIT get rid of deprecation warning in tree tests
       new  c77255f   MAINT deprecate HMMs
       new  f309023   MAINT ignore sklearn/tree/_utils.c in diff
       new  b83e74d   COSMIT tree: unused variable warnings and use for/range
       new  7fc52ff   BUG fix unchecked mallocs in trees
       new  9d5d2da   Merge pull request #2715 from larsmans/tree-malloc
       new  e81f2d5   Merge pull request #2734 from eloj/cv-broken-format-string
       new  7001128   MAINT remove last trail of the ArrayBuilder
       new  0b20feb   Add option to restrict LassoCV to positive-only coefficients
       new  d6e7e51   Add option to restrict ElasticNet to positive-only coefficients; add test for this case, and the same for LassoCV
       new  a64ec51   TST : speed up tests + cosmit
       new  7013e24   FIX : n_jobs was missing from LassoCV
       new  d4f66a0   update what's new
       new  921dead   Changed assert_array_equal() in Line 45 and 46 to assert_array_almost_equal(,,decimal=5). This has fixed the AssertionError, which occurs during the installation test.
       new  0ccdc3a   Merge pull request #2742 from ai8rahim/issue-2741
       new  8f76f48   DOC clarify order of output in NB predict_proba
       new  63f830b   Small documentation fix: from CMS to VCS
       new  260bdb9   TST Gini is equivalent to mse in binary classification
       new  82502e5   Merge pull request #2743 from eltermann/doc_fix
       new  bc6ad64   Merge pull request #2744 from arjoly/test-impurity
       new  cdebc18   FIX don't put data in source dir in bench_covtype
       new  ffd1e9a   FIX remove confusing BaseEstimator.__str__
       new  41ae4c9   Test primal-dual relationship in Ridge.
       new  33c6635   More robust test for sample_weight in Ridge.
       new  bac878f   fix: remaining time not currectly computed
       new  1c6c36b   Merge pull request #2753 from pprett/gbrt-verbose-remaining-time
       new  2a69266   Fixed documentation typo
       new  0f70379   OPTIM MSE criterion
       new  a2cb257   OPTIM more optimisation of mse
       new  55af57d   OPTIM more optimisation of mse
       new  04c7597   ENH use memset instead of for loop
       new  6b40580   Typo
       new  98f29cc   Update what's new
       new  0b7c79b   Merge pull request #2745 from arjoly/mse-criterion
       new  d4519d2   Added multiple linear Theil-Sen regression
       new  fdad6d5   Merge pull request #2754 from eltermann/doc-fix-tfidf
       new  18aaa26   Add mean_absolute_error to scorers.
       new  62b0180   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  45f61ae   Fix doctest.
       new  1d7a2d4   COSMIT: remove unused import
       new  1dcab45   conflict fix
       new  11ec836   small fixes
       new  91ae206   Merge pull request #1971 from jaquesgrobler/DOC_olivier_tut
       new  abe888d   sphinx colon error fix
       new  1357a19   FIX #1565: fix race condition in parallel pre-dispatch by upgrading joblib
       new  9446e33   Update whats_new.rst for race condition fix
       new  a0b9695   Merge pull request #2756 from ogrisel/joblib-0.8.0a3
       new  6bc5756   FIX #2645: fix 20 newsgroups downloader under Python 3
       new  e7e7133   ENH/FIX Change Tree underlying data structure
       new  5d8b429   Refactor cv code
       new  1d4d3f8   Clean up
       new  37e7c0a   Refactor RFE and add _check_scorable
       new  2b45b7c   FIX typo in docstring
       new  2bc3d49   Merge `fit_grid_point` into `_cross_val_score`
       new  50459fa   Return time
       new  712d3d0   Move set_params back to fit_grid_point
       new  8d3667d   Log score and time in 'cross_val_score'
       new  526d1dd   check_scorable returns scorer
       new  f95219f   Clean up
       new  0a000e0   Replace '_fit_estimator' by '_cross_val_score'
       new  7c60251   Fix PEP8, style and documentation
       new  64dbd45   Remove wrong variable names
       new  0f591e7   Remove helper function '_fit'
       new  b1dfa07   Remove 'fit_grid_point' from 'BaseSearchCV'
       new  046e20c   Check substrings of error messages
       new  aaa4f81   Rename '_split' to '_split_with_kernel'
       new  b4839bf   _passthrough_scorer is a function
       new  5010f39   Remove '_deprecate_loss_and_score_funcs'
       new  4e968a2   Check error message
       new  d818519   Rename _split_with_kernel to _safe_split.
       new  de2be61   Rename fit_and_score to _fit_and_score.
       new  2074965   Merge pull request #2732 from jnothman/two-array-tree
       new  0bb04e1   COSMIT: PEP8 + coding style
       new  9d7e30c   COSMIT: PEP8 in test_gradient_boosting.py
       new  3e6402f   COSMIT: Fix some PEP8 in gradient_boosting
       new  9e432a8   COSMIT: PEP8 in test_imputation
       new  1a887dd   COSMIT: Various PEP8
       new  51dab8d   DOC note #2732 in what's new
       new  64b65f6   Add thanks for infrastructure supporters
       new  3654c2d   Add a paragraph about installing a binary release in Fedora
       new  7c2d243   Merge pull request #2762 from sergiopasra/fedora-package
       new  e54c54a   Merge pull request #2761 from ogrisel/thanks
       new  57916b7   Use assert_raises_regexp to check error messages
       new  23fb798   BUG: dual_coef must be defined inside the try block.
       new  bf1635d   ENH: made example/svm/plot_iris.py clearer
       new  b3e8a64   Added an example and documentatin for Theil-Sen
       new  adea283   Merge pull request #2760 from AlexanderFabisch/check_error_patterns
       new  4b53258   TST: add Py2.6 on travis
       new  7c3a1c6   BUG: assert_raises_regexp for Py2.6 compatibility
       new  60e2053   TST: get travis to find the system packages
       new  fea0ceb   TST: try to get travis to work with Py2.6
       new  9e2f7ed   TST: get travis working on Py26 and Py27
       new  2ec6ed1   TST: Add Python 3.3 to travis
       new  877d892   Merge pull request #2767 from GaelVaroquaux/my_master
       new  93c9868   Update to mean shit clustering narrative documentation.
       new  65381d2   Update to docstring of meanshift
       new  683f5f1   docstring of module
       new  34a3331   Fix typos found by Alexandre
       new  a36c72a   DOC fix documentation to match default value
       new  4ec4b2f   ENH remove unnecessary CSR->CSC transform in text feature extractors
       new  8f1dd32   python 3 compatibility fix
       new  725923f   added howaboutwe testimonial
       new  cc41d83   Merge pull request #2772 from jnothman/tfidf_no_copy
       new  a917a02   ENH: MultiTaskElasticNet (and Lasso) CV
       new  8c3efd4   DOC make indifferent to compressed sparse orientation
       new  c75f0a6   Fix some typos in whats_new.rst, using ispell
       new  a9c55b4   DOC add link to module reference from class/function pages
       new  080887e   Merge pull request #2779 from jnothman/ref_sec_link
       new  8b7175c   Update plot_digits_pipe.py
       new  aed3ddd   Merge pull request #2780 from ugurthemaster/patch-1
       new  fb43369   DOC: some underlines were too short
       new  e98f4fd   FIX: Typo
       new  af1c98b   FIX: Typo
       new  e3d7347   Merge pull request #2781 from blagarde/patch-1
       new  4801f20   Merge pull request #2782 from blagarde/patch-2
       new  ab52527   Merge pull request #2776 from DanielWeitzenfeld/master
       new  c15054a   Merge pull request #2775 from stefan-w/python-3-fix
       new  06d7536   WIP: fix StratifiedShuffleSplit
       new  41b312c   BUG: StratifiedShuffleSplit not obeying n_train
       new  58da238   BUG: avoid same indices in test and train
       new  f8e7cf1   COSMIT
       new  8e162a0   Update plot_tree_regression_multioutput.py
       new  e532639   FIX remove reference cycles from Tree
       new  86fbebd   FIX dtype refcount; block resizing after building
       new  065fca0   Merge pull request #2789 from ugurthemaster/patch-2
       new  2111e3e   svm fit numpy array indexing deprecation warning fix.
       new  f203953   Merge pull request #2791 from adrinjalali/master
       new  87194f7   Remove 'locked', add comments
       new  e8fdfa6   Merge pull request #2790 from jnothman/tree_without_cycles
       new  271b054   DOC + FIX mean_ in PCA
       new  7c33502   ENH faster heapsort in trees
       new  31491f9   ENH introsort in tree learner
       new  a681c9b   ENH Make PresortBestSplitter cache friendly + cosmetics
       new  c83c450   fix numpy deprecationWarning('using a non-integer number ...') in svm
       new  6a46c8a   Merge pull request #2794 from cli248/master
       new  5ea9aad   Proper centering of alpha_grid for sparse matrices
       new  6859851   FIX: Normalize=True
       new  b92d19f   update what's new
       new  9f6dbc5   Merge pull request #2747 from larsmans/tree-sort
       new  7dea60d   Added subpopulation parameter to Theil-Sen
       new  c9e049c   clarify what's new
       new  ffec52b   fix converting an array with ndim > 0 to an index DeprecationWarning
       new  f657d9c   using floor division in python3
       new  1666c4f   Added docstrings for model attributes in LabelPropagation and LabelSpreading
       new  5874e74   Merge pull request #2799 from charlescearl/charles-label-prop-doc-updates
       new  60e3289   Fixed documentation on sklearn/tree
       new  04bf96a   Merge pull request #2800 from eltermann/doc-fix-tree
       new  0832d14   handle the case when length of sign changed coefficients is greater than 1 in LARS algorithm
       new  8ccb14a   Merge remote-tracking branch 'upstream/master' into fix-DeprecationWarning
       new  b92b7b4   handle the case where length of idx in least_angle greater than one
       new  1b73d8b   DOC logistic regression attribute docs + authorship
       new  6ebcf4d   COSMIT fix PEP8 errors
       new  5e59f4e   DOC up the sales pitch for SGD
       new  b75e509   MAINT: Fix PEP8 warnings in sklearn/covariance
       new  b895586   DOC fix roc_curve docstring.
       new  c997cde   COSMIT pep8 + full stop police
       new  4ae78ca   Merge pull request #2797 from cli248/fix-DeprecationWarning
       new  b247349   DOC: more context in embedded code
       new  baa3c7e   DOC: more robust compilation and CSS
       new  e952a58   Add prototype for validation curve
       new  b574858   Add documentation and tests
       new  fc1008d   Improve test coverage
       new  4764245   Improve interface and documentation
       new  ca60a0b   Mocks inherit from BaseEstimator
       new  7c63d08   Correct docstring
       new  849d159   Add narrative documentation
       new  ab41996   Improve documentation
       new  1128dc7   Fix link to validation curve
       new  0852a54   Add example with polynomial regression
       new  5544b3e   Correct and improve documentation
       new  4cf2229   Improve phrasing
       new  b68bd36   Rephrase sentence
       new  fe014d9   Simplify first part of the documentation
       new  4809168   Merge pull request #2768 from robertlayton/meanshiftdoc
       new  dbc26ce   COSMIT remove unused imports and variables
       new  587deb1   Minor doc example fix
       new  fedc3a3   Merge pull request #2820 from ankit-maverick/issue2819
       new  39949ed   Merge pull request #2818 from jnothman/flakes
       new  d83d652   Fixed FS1995 citation.
       new  16e7c82   Merge pull request #2825 from bryan-lunt/master
       new  d6eb004   MISC plot_rfe_with_cross_validation: better comments
       new  17d77ee   Fix typo
       new  1ada65f   Merge branch 'master' into hc_linkage
       new  c5af7cb   BUG: different distance names in scipy
       new  4a4bfc5   Return scores of all folds
       new  bb6ad77   BUG: fix minor floating-point precision detail
       new  e3f1c50   DOC: grammar and spelling fixes
       new  74d3fcf   Merge pull request #2826 from perimosocordiae/patch-1
       new  27e37a0   DOC make loss function in SGD consistent with subgradient. Comment by Martin Jaggi :)
       new  6610b3d   Fix documentation
       new  38579e7   ENH: Py3 support
       new  42d789d   Merge pull request #2827 from amueller/fix_subgradient_equation
       new  73246b1   Cosmit in SGD doc.
       new  5029361   BUG: fix tests under Python 3
       new  5afd11f   DOC: improve the clustering metric example
       new  b53b573   Replaced abbreviated 'w.r.t' to 'with regards to'
       new  82a5ba8   Fixed small typo
       new  92101da   Doc fix - Compiled .pyx files with Cython 0.20
       new  5b1f20a   Merge pull request #2831 from eltermann/doc-typo
       new  2fb8bc7   Improve test coverage
       new  5314720   Matching colors
       new  5319994   Merge pull request #2765 from AlexanderFabisch/validation_curves
       new  b928cdb   making adjustments to variable names and account for sparse matrices
       new  016469b   changing xrange to range
       new  cfa55d3   DOC: AgglomerationClustering doc and examples
       new  4d80664   BUG: _alpha_grid undefined symbol
       new  a938b20   DOC: Agglomeration clustering docs and examples
       new  fa3edaa   DOC: connectivity in agglomerative clustering
       new  5ed6b26   DOC: finish touches to hierarchical docs
       new  4ebf883   FIX: Weight impurity decreases with weighted_n_samples
       new  31c842f   TEST: add non-regression test
       new  8833d59   ENH: add weighted_n_node_samples property
       new  125fe3d   Cleanup
       new  5cc89b2   FIX: Recompile _gradient_boosting.c
       new  961694d   DOC: fix rest syntax
       new  de38475   MISC: typos in docs
       new  75705e2   MISC: remove reference to deprecated Ward
       new  e757a67   MISC: replace 'assert' with actual ValueError
       new  f963317   MISC: increase test coverage and cosmetics
       new  150fee8   TEST: check with scaled sample_weight
       new  8974b2d   Merge pull request #2836 from glouppe/tree-importances
       new  76c5802   DOC: Entry for #2835 + Put together entries wrt. trees and ensemble
       new  8e0b4be   BF: Correct degrees of freedom in f_regression + test.
       new  4d18fd8   BF in f_regression: variable naming + use assert_*array*_almost_equal.
       new  4c129e5   f_regression and degress of freedom: update whats_new + minor.
       new  71fa268   FIX error message with sparse precomputed kernels
       new  67e21b1   Revert "FIX error message with sparse precomputed kernels"
       new  f54ffa0   FIX error message with sparse precomputed kernels (second try)
       new  8e5055f   MAINT optimize Spotify logo (- a few hundred bytes)
       new  6331623   FIX restore _pairwise on SVMs
       new  541f1ca   Fixed incredibly minor spelling mistake
       new  8c7a233   Merge pull request #2840 from earino/develop
       new  c72805a   Merge pull request #2830 from eltermann/wrt-abbreviation
       new  74861d0   Merge pull request #2750 from jnothman/no-str
       new  b87a2f2   What's new: group entries by topic
       new  ed928ad   Remove superfluous AdaBoostRegressor call.
       new  998a57a   Merge pull request #2852 from jwkvam/adaboost_test
       new  722f062   MISC: better error messages in BaseLibLinear
       new  64efbd6   MISC+DOC: documentation for FeatureAgglomeration
       new  26c51b7   Uniformize max_features semantics
       new  e368d9a   TST add tests with constant features
       new  5d02521   Update what's news
       new  d7c535f   Remove duplicate what's new
       new  9e4e56a   /s/2013/2014: Updated project copyright date
       new  8518502   Merge pull request #2857 from eltermann/copyright-updated
       new  d63af22   Update naive_bayes.rst
       new  5ffe7c9   Merge pull request #2858 from ameasure/patch-1
       new  a3e5c99   s/svn/svm
       new  e20aebe   Merge pull request #2859 from eltermann/fixed-typo
       new  3c543bf   Merge pull request #2829 from arjoly/maxfeatures-seamantics
       new  ee85da8   ENH sparsify and densify methods for CD models
       new  1a3fd94   FIX "ValueError: startprob must sum to 1.0" in HMMs
       new  e1f8bfb   MAINT Rename over/underfitting example
       new  8f94ae6   PY3 fixed examples
       new  03f33ba   MAINT comment out unused import in example
       new  94d5013   MAINT full name for @h10r
       new  5f027f3   FIX error message from trees for large inputs
       new  7906086   DOC: minor improvement in comments of an example
       new  a3878f6   Revert "ENH sparsify and densify methods for CD models"
       new  3b20a02   Merge pull request #2864 from sciunto/doc
       new  4a29b98   MAINT missing import_array() in isotonic r. Cython code
       new  884889a   MAINT get rid of compiler warnings from Liblinear
       new  76206f1   COSMIT: many cosmetic comments
       new  0cd5251   FIX: fix failing doctest
       new  445eee2   DOC: link FeatureAgglomeration better to dim red
       new  3ac3455   MISC: documentation and cosmits
       new  7c621fc   DOC no more "arithmetical order" for classes
       new  4c370a7   BF: load_boston() return 13 features, but 14 feature names
       new  8826e68   BUG: work aroundd Cython bug to build with clang
       new  aa66938   Fixes Issue 2751
       new  6beae0a   Merge pull request #2865 from hanke/boston_13features
       new  a23b868   Merge branch 'master' into hc_linkage
       new  c7aae5c   ENH Cache features value for extra trees
       new  c6410d1   FIX: Removed coef and improved initialization
       new  4403d33   Merge pull request #2866 from Manoj-Kumar-S/Iss2751
       new  7324e8f   FIX integer dtype for labels in DBSCAN
       new  0c22cbb   FIX numerical stability issue in BernoulliRBM
       new  0fa9008   Fix typo in fast sgd classifier implementation comments
       new  ba89ed0   Merge pull request #2869 from jperla/master
       new  6740de0   FIX: sample_weight='auto' for RidgeClassifier
       new  122cb23   Fix citation in TfidfTransformer
       new  80ccf7b   Merge pull request #2877 from pmandera/tfidf-citation
       new  9e8caa5   ENH: small optimization of the classification criterion
       new  69b7162   ENH: small optimization to the regression criterion
       new  0ba1e28   ENH speed up RBM training with scipy.special.expit
       new  fd54575   ENH re-instate extmath.logistic_sigmoid
       new  bd0cf75   FIX DictVectorizer handling of empty inputs
       new  89d94ca   TST older nosetests compat in DictVectorizer test
       new  24dd341   putting cluster  on new line to solve sphinx documentation formatting issue
       new  c52476a   Merge pull request #2882 from larsmans/expit
       new  b96f354   Merge pull request #2884 from dsullivan7/clusterdoc
       new  5e0ed70   [TYPO] s/migh/might/
       new  3ec5632   Update supervised.py
       new  a12f46d   Update supervised.py
       new  7d1ba65   Merge pull request #2895 from kaushik94/master
       new  4640608   Update plot_forest_importances.py
       new  8cdd2be   Merge pull request #2897 from ugurthemaster/patch-1
       new  b499743   DOC: update examples, use matplotlib.pyplot instead of pylab
       new  b0f3f1a   peerindex testimonial
       new  9a0ed69   Merge pull request #2909 from ogrisel/peerindex-testimonial
       new  0f5685b   FIX predict_proba status on SGD and SVC when disabled
       new  aebcb16   ENH use hasattr "predict_proba" in bagging
       new  97aed28   FIX one more unchecked malloc in the tree code
       new  dd335cb   DOC SVC.predict_{,log_}proba does not return X
       new  aa53f80   DOC what we call poly features are called interaction features in stats
       new  0a488ff   FIX OneHotEncoder: check value max when n_values is integral
       new  f73ce69   Merge pull request #2910 from larsmans/hasattr-predict-proba
       new  8d48d45   Added parallelization support to TheilSen Estimator
       new  566237d   FIX: Label encoding done in compute_class_weight
       new  88ecfa0   Merge pull request #2876 from Manoj-Kumar-S/fix_auto
       new  276e260   ENH add sparse parameter to OneHotEncoder
       new  57d9ede   DOC: note about numerical precision in euclidean_distances
       new  caf0035   TST fix failing doctest for OneHotEncoder
       new  b3a8759   changing formatting of documentation
       new  a83fddf   Merge remote-tracking branch 'upstream/master' into rbmbranch
       new  b80b0ea   STY Simplify and avoided numpy warnings in imputation
       new  eb9bc96   FIX GBRT missing from covertype benchmark usage
       new  dabf48c   DOC: Update dead link in cross_decomposition.rst
       new  551ea22   Merge pull request #2917 from ariddell/patch-1
       new  b97fb3c   MAINT drop support for NumPy < 1.6.1
       new  23b3cc2   MAINT drop support for SciPy < 0.9
       new  20b87f9   TST: test utils.extmath more robust
       new  86a7461   Improved parallelization for Theilsen estimator
       new  601b82b   Merge branch 'master' into theilsen
       new  baae60b   MAINT remove useless import
       new  c9e4cd0   ENH more optimizations for RBM
       new  596879c   Fix a minor typo: 'They requires' should be 'They require'.
       new  295c5ed   Merge pull request #2920 from yoni/master
       new  5ce5e81   [Cross Validation] Use itertools.izip consistently across Python 2/3 when computing folds in StratifiedKFold.
       new  e721508   Merge pull request #2925 from ajtulloch/izip-cross-validation
       new  61998ac   FIX #2924: make lobpcg test pass with reference and ATLAS impl of LAPACK
       new  8eeda19   [tests] Use Python 3 zip/Python 2 zip consistently in tests
       new  43da4ad   ENH/TST remove unnecessary sorts and complete testing for sparse median
       new  dc27992   MAINT: remove our solve_triangular
       new  f6d3e31   MAINT: commit generated code
       new  33a4538   FIX + TST stability problems with scipy.special.expit
       new  1b67fe3   MAINT: scipy 0.10 and 0.11
       new  93e5f32   Add test for _safe_split with pre-computed kernel.
       new  10bfa55   Fix test failure on Python3.
       new  13c477f   More stable test problem for eigen_solver='lobpcg'
       new  2a4402b   DOC: rmk different feature scaling in agglomeration
       new  39f3a76   Merge pull request #2199 from GaelVaroquaux/hc_linkage
       new  0d96a48   Merge pull request #2929 from ajtulloch/izip-tests
       new  2bac9eb   Merge pull request #2931 from GaelVaroquaux/rm_solve_triangular
       new  c8a372b   Update plot_svm_regression.py
       new  092b50e   BUG: avoid NaNs in arrays passed to scipy.linalg
       new  6114260   FIX input validation in Nystroem
       new  034808d   FIX Nystroem input validation, again
       new  232fad1   DOC: docstring formating
       new  4437d56   FIX error handling in SVM
       new  c19d967   Cleanups and corrections for Theil-Sen regression.
       new  b4ca1d6   COSMIT Remove unnecessary assertion
       new  e54af63   BUG: avoid NaNs throwing off class probabilities
       new  ac0f46f   fixing warning ignore type
       new  ad22a3b   adding test case
       new  bd1788e   DOC: fixed docstring formatting
       new  cbc005f   Added reference to function
       new  c523d1b   Merge pull request #2928 from ogrisel/fix-linalgerror
       new  25af09d   Merge pull request #2943 from jyu-rmn/patch-1
       new  392630d   conform to pep8
       new  4b32c8f   Added :func: reference
       new  1ebb228   Added :func: reference
       new  27cde31   Merge pull request #2945 from maheshakya/olivetti
       new  79a8012   FIX for numpy compatibility, code clarity
       new  0d8c2d7   ENH: update installation instructions for Ubuntu / Debian
       new  64fd085   FIX decouple spectral embedding from TransformerMixin
       new  8653c20   Removed subpopulation=None option in TheilSen
       new  960c45b   xrange fix for Python3 in TheilSen
       new  0eda5bf   FIX Theil-Sen unittest for older Scipy Version
       new  d133712   FIX that some functions in Theil-Sen were public
       new  d1d221b   FIX usage of linalg from Scipy in Theil-Sen
       new  61a5195   FIX: Let Theil-Sen handle n_samples < n_features case
       new  13d9212   FIX: Python 2.6 format syntax in Theil-Sen
       new  3e8ca8c   Vectorization of theilsen._modweiszfeld_step
       new  1d75896   FIX: Parallel unittests for Theil-Sen estimator
       new  3753563   ENH - Improve performance of isotonic regression
       new  6091d7d   DOC fix rendering of fetch_mldata example
       new  28c8d0b   MAINT warn of future behaviour change proposed in #2610
       new  0afa596   changing computation of batch slices to np.array_split
       new  abaf100   WIP add n_constant_features argument
       new  3e123a1   WIP add constant_features array
       new  57f93fa   ENH avoid splitting on constant features for best splitter
       new  25e2c19   FIX bug with BestSPlitter + ENH avoid searching for a split on constant features
       new  e509724   ENH avoid trying to split on constant features for presort best splitter
       new  69d4857   COSMIT
       new  b815e90   COSMIT
       new  2f1fa9b   FIX deallocate memory
       new  44fe814   FIX proper inititalization
       new  d9d1b71   ENH use memcopy instead of while loop
       new  42d7d08   COSMIT
       new  e3415ea   FIX bug with invariant
       new  4ab12f4   Simpler swap
       new  88eb3cb   ENH: add explanatory coments
       new  7b4dcb8   pep8
       new  6cac2ed   ENH remove unused constant
       new  a9f21a9   ENH document and rename EPSILON_FLOAT
       new  7419965   DOC improve documentation for the splitting algorihtm
       new  c49294c   DOC verbose documentation of features and constant_features
       new  d41fb76   Implemented median and constant strategies in DummyRegressor
       new  9879c95   Depreciated y_mean_ and y_median. Code is formatted according to pep8
       new  5503494   Removed extra lines. Added deprecation warnings for y_mean_
       new  66e2fa4   Updated document to reflect changes in the DummyRegressor
       new  9799a3d   Fixed value error. Scalar values are allowed for constant when not multioutput
       new  815af68   Done refactoring and necessary changes for multiouput constant strategy. Added tests and updated the documentation.
       new  c20c16a   Changed value error messages to be more informative in constant strategy.
       new  3755d5c   Edited the documentation-unordered list
       new  b047518   Implemented additional test cases. Removed the check for lists and numpy arrays in constant strategy.
       new  ee59a0c   Refactored the code and added quotes
       new  a37aa8f   Recreated test cases with random state
       new  5158913   Updated tests. Fixed flake8 errors
       new  5348177   fixed pep8 errors in dummy.py
       new  8320331   Changed to random_state
       new  f1d95f2   Changed documentation in DummyRegressor
       new  18d62d7   Rename EPSILON_FLOAT -> CONSTANT_FEATURE_THRESHOLD + DOC n_drawn_constants & n_found_constant
       new  57f73f1   DOC more comments about features and constant_features
       new  3412f29   Removed one of 'is a regressor' part in the DummyRegressor documentation
       new  d576ec8   pep8 corrections
       new  9b92c45   DOC explain invariant with constant features + DOC clarify comments
       new  e6c8a50   DOC more comments of the splitting algorithm
       new  302a3c0   Merge pull request #2886 from maheshakya/dummy_regressor
       new  721efea   Update what's new
       new  ac8cbb1   MAINT remove leftovers from solve_triangular
       new  d9d02aa   COSMIT get rid of warning from expit import
       new  b374e7e   FIX: TheilSen supports old Numpy versions
       new  3324a41   COSMIT hinge_loss: better input validation
       new  55ef5a4   apply the fix closer to the source of the issue
       new  acd2316   Merge pull request #2941 from perimosocordiae/patch-1
       new  b37958b   Merge pull request #2875 from arjoly/et-skip-invalid-
       new  9a995db   Update what's new
       new  eb04ccb   Implemented a check for ndim exceeding two in the utils.check_arrays function
       new  26d3485   ENH: Speed up sparse_coordinate descent
       new  6f157bf   Fix silhouette score n_labels
       new  e84f4d0   Added number of labels for debug and regexp in test
       new  2fe838c   FIX stability in HMMs
       new  a48c51b   Removed unnecessary variable n_dim in utils.check_arrays, removed unnecessary parens in same place
       new  b911a5e   Replaced cython calls for dot operations with BLAS calls
       new  c587d4f   Trivial change in utils.check_arrays from > 2 to >=3 in attempt to rebuild Travis CI
       new  14b435c   Merge pull request #2938 from jnothman/clean_impute
       new  c530f5a   FIX for early numpy where astype(..., copy=...) unavailable
       new  5a92c1d   Merge pull request #2958 from hamsal/check_arrays_n_dim
       new  0abe21e   MAINT remove some deprecated stuff
       new  4c6e026   MAINT final occurrence of "Scaler"
       new  eef9fd5   Fix typo in linear_model documentation
       new  524f4cd   Merge pull request #2960 from ethanwhite/docs-typo
       new  878cd8b   Speed up using typed memory views
       new  720b77f   Merge pull request #2951 from Manoj-Kumar-S/speed_sparse
       new  1d0da8a   staged_predict predicts classes, not probabilites
       new  e1edcb1   Merge pull request #2964 from abhishekkrthakur/patch-1
       new  ffb89eb   Added _LearntSelectorMixin in BaseGradientBoosting
       new  08b82f8   TEST: fix test_forest:test_boston (#2965)
       new  71a50af   Minor grammar fix.
       new  b8f9835   Removed duplicate word and corrected grammar in decomposition documentation.
       new  71cba01   typo
       new  69a39a3   Merge pull request #2971 from jyu-rmn/master
       new  71ce1d5   DOC: the search for a split does not until one valid partition is found
       new  adb44cd   add datarobot testimonial
       new  b981a85   Merge pull request #2973 from pprett/datarobot-testimonial
       new  bfb4b7e   DOC/MAINT: clarify cblas/README.txt
       new  b0e11f5   Added extra check to feature importances
       new  99f2860   Merge pull request #2966 from maheshakya/feature_selector
       new  9e2071b   DOC clarify Imputer constructor: arbitrary strings not accepted
       new  b90654a   Merge pull request #2959 from Oscarlsson/silhouette_score_label_number
       new  751049e   FIX: np.abs might not work on scipy.sparse matrices
       new  6af42a7   DOC Remove redundant parameter docs
       new  096070d   MAINT remove deprecated functionality from SGD
       new  85922d3   ENH use threads instead of multiprocessing in SGD
       new  f434c8c   Clean tree builder interface
       new  73f0271   ENH inline sme function
       new  e22045a   ENH avoid serializing the splitter & criterion + avoid storing tree parameter in the tree structure
       new  0334d76   ENH avoid serialising the random_state in the tree
       new  07d6e8d   ENH clean public tree strucure
       new  8f6b753   ENH for the tree structure max_depth is part of the inner structure
       new  39a2a76   ENH stricter separation between tree builder, splitter and criterion
       new  203a598   ENH add a dedicated constant for min_impurity_split
       new  8534989   Document maximal depth of the tree
       new  24db12c   ENH stop splitting for best first tree builder if pure node + change strict to unstrict inequality
       new  9d728f3   ENH stricter separation between splitter and criterion
       new  2c82c8a   FIX tons of flake8
       new  5839dda   DOC document max_depth of TreeBuilder
       new  84c3742   ENH re-generate c file
       new  bb5f643   MAINT use CBLAS instead of Fortran API in Liblinear
       new  08a7f67   fix issue #2901
       new  d413229   Merge pull request #2981 from matrixorz/master
       new  61d9c5e   TST skip `test_k_means_plus_plus_init_2_jobs` on Mac OS X 10.9. See GH-636
       new  88a85e0   TST fix sklearn.ensemble.tests.test_bagging.test_parallel
       new  20c45ae   TST move Mac OS checking utility to sklearn.utils.testing
       new  e87653d   TST skip SparsePCA n_jobs!=1 test on Mac OS X because it hangs
       new  5c120f5   adding test for csr data for partial fit
       new  eb037dd   pep8-ing
       new  7776d67   PY3 file handling
       new  db0b223   BUG restore joblib logging behavior
       new  5dbb039   BUG joblib writes to wrong dir
       new  cc7f7d6   DOC trees/ensembles: class labels need not be integers
       new  3a6b834   Add space to be consistent with _tree.pyx
       new  a4680b1   DOC document internal tree class
       new  2e5b210   ENH rename CONSTANT_FEATURE_THRESHOLD to FEATURE_THRESHOLD + revert some pep8 fix for readability
       new  d774953   DOC missing words
       new  d9bd47d   Merge pull request #2977 from arjoly/tree-cleanup
       new  874235c   removing copy() and adding tests for multiple forms of sparse matrices
       new  696cf55   removing tests for multiple sparse matrix formats
       new  0b10f49   DOC: Comparison of Theil-Sen and RANSAC
       new  d40eeb0   DOC: wordings
       new  ca9a275   DOC: Fixed typo in Theil-Sen example.
       new  8ff19bd   Preserve CSR storage format when input is CSR in sparse_center_data
       new  675babb   FIX: Some coding style fixes in TheilSen unittest.
       new  61e5966   Merge pull request #2991 from Manoj-Kumar-S/csr_sparse_center
       new  279dc90   FIX: Reduced the runtime of the TheilSen unittest.
       new  39efe30   DOC: Small corrections in the docs of Theil-Sen
       new  b40e58e   Cosmit
       new  0043b62   MAINT: point out HMMLearn
       new  689bf29   DOC: improve Thiel-Sen vs RANSAC example
       new  b2366b2   ENH: speed up theilsen
       new  4de5e6c   Merge pull request #2955 from larsmans/less-blas
       new  261da69   TST remove n_jobs=-1 usages in tests
       new  3bc3d9f   ENH pairwise L1 distances for sparse matrices
       new  faba875   DOC example rendering
       new  39b859b   FIX TfidfVectorizer to no longer ignore binary param
       new  3f2547e   DOC: add OkCupid testimonial
       new  32fefb7   Merge pull request #2997 from GaelVaroquaux/okcupid_testimonial
       new  c49723d   Merge pull request #2982 from kmike/fix-macos-hangs
       new  43d76ac   DOC: Explanation when TheilSen outperforms RANSAC.
       new  7ed792f   adding test for csc and csr sparse matrix
       new  b238b11   DOC use hyperlink colour when in tt
       new  b4625c7   Merge pull request #3003 from jnothman/hyperlinked_tt_colour
       new  9627853   adding change to whats_new
       new  90ce739   FIX: more robust skip of implicit constructor
       new  3c753a3   Merge conflict in whats_new.rst
       new  e77fc16   Merge pull request #3006 from ogrisel/fix-implicit-init-introspection
       new  26d6f8c   MAINT don't use Perl in Makefile when sed suffices
       new  785e3ef   ENH: Refactoring and optimisation of sparsfuncs.pyx
       new  8cf30f1   ENH: Optimise sparsefuncs a] Removed numpy overhead in finding mean b] Removed enumerate c] Added tests to test_sparsefuncs d] Replaced multiple looping with a single loop
       new  bbb8f1d   Moved sparsefuncs to sparsefuncs_fast
       new  22c412d   TST: Added tests for non-CSR/CSR format
       new  c1003ec   DOC heapsort is not stable at all
       new  aeb1dcd   FIX cross_val_score to take y as a list
       new  6b38d3b   FIX cross_val_score to take y as an *optional* list
       new  d0985e8   Merge pull request #3007 from Manoj-Kumar-S/refactor_sparsefuncs
       new  2b2c5b2   FIX: np.random.randint expects signed 32 bit integers under Windows
       new  b7aa549   FIX sed usage in Makefile
       new  eb10c4c   FIX: numpy mtrand does not accept Python long instances under Windows
       new  f1e2e1c   Fixed SelectKBest corner case: k=0
       new  a3f81d9   Merge pull request #3010 from griffinmyers/fix-select-k-best
       new  4c4cb7a   FIX: make python setup.py clean also delete __pycache__ folders
       new  5857720   FIX: remove casting warning under Python 3
       new  86c07b5   FIX #3014: use a different folder for covtype data under Python 3
       new  033ae09   FIX: fix build under Python 2.6
       new  eb39ff2   BUG FeatureAgglomeration meaningless for no samples
       new  070f01d   FIX Makefile to use Perl again
       new  32c843b   COSMIT: pep8, trailing spaces
       new  4c10d9c   Fix ValueError: Buffer dtype mismatch, expected 'INTP' but got 'long' on win-amd64
       new  7c979f9   Merge pull request #3017 from cgohlke/patch-2
       new  c95fcd0   FIX MSVC compile error C2036: 'void *' : unknown size
       new  627c564   TST: Fix ValueError: Buffer dtype mismatch, expected 'npy_intp' but got 'long' on win-amd64
       new  eb7e86a   TST don't run fast_dot tests on numpy>=1.7.2 + pep8
       new  a786e12   TST/COSMIT tests/comments for [sparse]center_data
       new  fbe974b   DOC: tfidf is actually tf*(idf+1) = tf + tf*idf
       new  3445813   COSMIT: More readable drop for good test
       new  fec2867   FIX: checked in bad assertion in the last commit by mistake
       new  2495327   Remove redundant yet unstable test_spectral_lobpcg_mode
       new  58ffba9   TST: more robust test of dropping in Lars
       new  9c80d11   TST: relax warning class checked in lars
       new  1c73c69   TST: lars tests: more brutal ill-conditionning
       new  c50cbbd   ENH: make CD Lasso raise ConvergenceWarning
       new  13f82ac   Merge pull request #3030 from ogrisel/fix-lars-drop-for-good-test-failure
       new  86f8577   DOC: clear up the big elastic net confusion (I hope)
       new  42663e4   Merge pull request #3025 from ogrisel/remove-unstable-lobpcg-test
       new  0834a55   ENH: stable test + catch warning + pep8
       new  8d04380   ENH: more robust test_toy_bayesian_ridge_object
       new  0dc5f37   ENH: configure travis to also test old numpy & scipy
       new  80ef9eb   DOC: add some header doc to the travis scripts
       new  e538a85   More informative numerical error message in SGD
       new  1f71224   cosmit
       new  140488c   DOC: Fixed roc_curve docstring
       new  4c4bda3   Merge pull request #3035 from ogrisel/test-sgd-stability-error
       new  7554c36   Merge pull request #3022 from ogrisel/travis-old-numpy-scipy
       new  9c51bc9   MAINT remove _label attribute from SVC.
       new  269afc1   FIX numerical stability issues on 32 bit platform
       new  719e3f0   MAINT lazily import scipy.cluster
       new  91a45af   ENH micro-optimize a few tests
       new  e6003bc   ENH micro-optimize fast MCD
       new  05807fd   FIX random_state validation on c_step
       new  912823f   DOC mention "shape" in AP docs
       new  c0657ce   ENH less copying in validation for neighbors
       new  deff716   FIX safe_asarray to handle LIL, DOK formats
       new  c7519ee   TST safe_asarray for dok_matrix and lil_matrix
       new  8e5b679   DOC: website: feature selection on front page
       new  e1d4df4   FIX loss function example
       new  c7bd7a3   DOC: clustering: merge discussions of k-means and inertia
       new  070d7a6   MAINT remove unsupported documentation formats
       new  b08d193   DOC: move mean_shift docs to MeanShift
       new  d954846   Merge pull request #2952 from jnothman/future_warn_two_labels
       new  fb795eb   ENH factor out squared norm helper
       new  50e35b5   Now capable of treating sample_weights in feature space
       new  da95ad8   FIX: Set correct impurity values in BestFirstBuilder
       new  2633774   PEP8
       new  7b46c13   Merge pull request #3049 from glouppe/tree-bestfirst
       new  60b3120   target is response, not explanatory
       new  224d769   typo: "fot" -> "for"
       new  157a4d9   Merge pull request #3055 from ajschumacher/patch-3
       new  f617253   Merge pull request #3053 from ajschumacher/patch-1
       new  61dd4b4   FIX: weight impurity improvement by the size of the node
       new  698eafa   DOC: better doc for impurity_improvement
       new  dcc6cc1   ENH: declare weighted_n_samples
       new  57a743a   ENH: rename Node attributes to avoid confusion
       new  45c0f04   ENH: rename normalizer -> weighted_n_samples
       new  9f0828a   ENH: renamed things too fast...
       new  7ce38f4   ENH: better API for passing weighted_n_samples
       new  e6e7acf   typo: "requiered" -> "required"
       new  68f013a   Merge pull request #3057 from ajschumacher/patch-4
       new  adc24fd   DOC no input validation in constructors
       new  1d2baf8   ENH: Add dynamic maximum trial determination to RANSACRegressor
       new  b906d53   DOC: typo: 1e-2 is 0.01 not 100
       new  b6eedcc   DOC remove import scikits.learn from tutorial
       new  3f9dff9   FIX: Add allow_nans option to check_arrays
       new  efd610d   ENH: remove unused variables
       new  c3ab3ba   Merge pull request #3056 from glouppe/tree-bestfirst
       new  e841d9d   TST: Switch from python 3.3 to 3.4 in travis
       new  6b91f6d   Merge pull request #3059 from mdbecker/update_from_33_to_34
       new  c6fb578   updated AMI documentation for issue #2686 .. _clustering:
       new  f54ad81   [DOC] Updated install instructions for Arch Linux
       new  0d15cb2   Improved plot digits classification example.
       new  f1c8dcb   update ami doc for feedback on PR
       new  8d78111   Merge pull request #3061 from jess010/ami-docs
       new  2e03618   <MRG> Doc generation works in Python 2 and 3
       new  f4d4d84   FIX Move pooling_func to constructor.
       new  2c17db1   [MRG] Addressed comments
       new  8307eee   Ridge regression now can use sample_weights in feature space. Summary commit over around 20 commits to avoid failing tests
       new  4a3c44e   FIX: AdditiveChi2Sample can be initialized with sample_interval, #3068
       new  4dd36b3   Fix merge conflict and missing URL ref in whats_new.rst
       new  c039705   [MRG] Addressed comment, fixed open-ended except
       new  24f0a09   DOC Fix ``copy_X `` default in documentation
       new  b88cec5   Merge pull request #3070 from luispedro/fix-enet-doc
       new  a94b82c   Merge pull request #3062 from abatula/archlinux-install-doc
       new  3c1db12   More Py3 fixes for sphinx build
       new  94b800d   Merge branch 'python-3-sphinx-fixes'
       new  6f6de86   Merge pull request #3063 from chalmerlowe/improved-digits-example
       new  6982977   added tests for sparse matrix inputs to BaggingClassifier and BaggingRegressor
       new  ba352bf   open iris files with  statement to avoid ResourceWarning
       new  a32eb88   FIX: One more PY3 fix in the documentation generator
       new  f014b69   Add tests for the sample_interval in #3069
       new  65bbe9e   plot_adaboost_multiclass.py: handle case where boosting terminated early. Add missing author on other boosting examples.
       new  f260259   xrange -> range
       new  64f3026   Merge pull request #3078 from ndawe/examples
       new  6384f2d   Merge pull request #3074 from msalahi/iris-file-warnings
       new  9f07963   FIX: typo in gen_rst.py
       new  b70a481   FIX: More robust test_extmath.py, compatible with older numpy
       new  5ebba54   Fix the pep8 violations
       new  00bb7a4   TruncatedSVD: Calculate explained variance.
       new  a30b2fd   DOC: Remove dead link
       new  6441caa   ENH: Compute lower number of points by default
       new  e4f9b5b   Improved the learning curve example.
       new  a63da37   Fixed style errors detected by pep8.
       new  6b2caf0   Merge pull request #3069 from ssaeger/issue_3068
       new  4009769   updated authors
       new  b965452   DOC: Change URL-based link to sphinx :ref:
       new  2e85d0a   Merge pull request #3085 from eickenberg/update_ridge_authors
       new  4605439   Merge pull request #3082 from ElDeveloper/install-link
       new  47080ec   Merge pull request #3067 from mdbecker/truncated_svd_calculate_explained_variance
       new  d31dee1   BaggingClassifer/BaggingRegressor tests for sparse input
       new  9931e34   Authors: Update based on #3067
       new  56a251c   Merge pull request #3083 from sdenton4/enh_learning_curves
       new  4205d1f   Merge pull request #3076 from msalahi/sparse-bagging-tests
       new  ea91673   Merge pull request #3086 from mdbecker/update_authors
       new  6c0d41a   FIX: remove deprecation warnings in learning curves under Python 3
       new  a03abd6   Merge pull request #3090 from ogrisel/learning-curves-warnings
       new  9792624   COSMIT: use plural, there are 2 learning curves
       new  f8aeecb   Update testimonials.rst
       new  4c41b51   adding lovely logo
       new  f2fa887   Merge pull request #3066 from denisgarci/agglomeration-transform-constructor
       new  e9d660d   add sample_weight to base score and weight_boosting staged_score
       new  2cdf283   weight_boosting: unneeded np.copy
       new  3965205   weight_boosting: include sample_weight in test_staged_predict
       new  c5a0ee4   metrics: add sample_weight support
       new  9cac887   rm (default=None)
       new  93f2b07   require sample_weight support for binary_metric
       new  4c13257   newline
       new  b9371fa   atleast_2d.reshape -> reshape
       new  57c18e4   weighted metrics: fix sample_weight handling for average=samples
       new  9762d59   format
       new  4657705   metrics tests
       new  b4df4cd   doc: fix default
       new  56afa3e   weighted metrics tests fixes
       new  43e225b   np.sum(np.multiply( -> np.dot(
       new  08669f4   update whats_new.rst
       new  9df5626   add test_base.test_score_sample_weight
       new  9cc38dd   sample_weight metrics tests: add missing micro and macro averaging for precision, recall, and f-score
       new  3c9e458   TST define only one list of name for sample weight invariance testing
       new  57f906f   Typo: thepath -> the path
       new  e247a5b   Merge pull request #3096 from ash211/patch-1
       new  7e341ab   FIX missing comma in METRICS_WITHOUT_SAMPLE_WEIGHT
       new  069bfac   Add artwork with reference to logos in about.rst
       new  feb6eb1   Merge pull request #3101 from fabianp/logos
       new  4d9a67f   DOC Comment on thresholds[0] for roc_curve
       new  463cea5   BUG: test_lasso_lars_vs_lasso_cd_ill_conditioned
       new  5119d68   ENH: Swap rows in sparsefuncs
       new  aac2f4e   PY3: fix map usage in ensemble.partial_dependence
       new  7c140a1   COSMIT six.{map,range} usage in partial_dependence
       new  924db9b   Made the following changes a] Replaced numpy slicing with concatanate b] Added swap_sparse_column
       new  ffc0fa6   DOC advertize GenericUnivariateSelect in the narrative documentation and api
       new  3a4edd0   Merge pull request #2936 from ugurthemaster/master
       new  31d438e   COSMIT: Replaced ptr1/2 with start/stop
       new  4d5aebf   typo
       new  3a881c7   wrap lovely testimonial's paragraph
       new  b3affdb   Merge branch 'pr/3091'
       new  1e5e7e1   failing test for wrong solver exception
       new  a8e3d99   Added raise statement in ridge_regression solver check
       new  9a91431   FIX raise memory error if constant_feature can't be realloc
       new  6945d5b   Merge pull request #3116 from eickenberg/ridge_wrong_solver_exception
       new  02b59de   COSMIT: rng.permutation rather than rng.shuffle
       new  fc634f9   Made the following changes 1. Minor changes to docs 2. Replaced swap with inplace_swap
       new  b842d4d   Merge pull request #3104 from MechCoder/swap_sparse
       new  48e2b13   Update numpy and scipy requirement
       new  1efb594   Change the order of test_score and train_score in the _fit_and_score docstring to reflect what the code actually does
       new  29e9ed1   Merge pull request #3127 from rmcgibbo/_fit_and_score-docstring
       new  0dbbf19   DOC random_state is an arg to LogisticRegression
       new  9d1a850   pep8 remove unused variable
       new  1f9c9ae   ENH: use setuptools for bdist_wheel command
       new  17bca23   DOC a note on shuffling for cross-validation
       new  b67ba03   ENH shuffle option for StratifiedKFold
       new  3ab286a   DOC shuffled StratifiedKFold now in whats_new
       new  ba922d1   BUG: fix doctests
       new  c008b0b   DOC: fix restructured text
       new  65bb024   FIX MemoryError raising in trees (+test)
       new  db51f7f   DOC mention that train_test_split returns a random split
       new  992899f   Merge pull request #3133 from kmike/patch-2
       new  0651d73   Using the 'is' token compares identity; this should not be used for string value comparison.
       new  5aa5357   Merge pull request #3134 from jrouly/string_values
       new  880bf96   Update working_with_text_data.rst
       new  1928f6f   Merge pull request #3135 from ugurthemaster/patch-1
       new  4464a56   Add FAQ to docs.
       new  c7d912b   removed sklearn, removed statsmodel, capitalization in SciPy
       new  9926867   Add GPU support (there is none and will probably be none) to faq.
       new  9c69e2b   DOC faq formulation
       new  5e5ed7a   typo
       new  477b47e   DOC: add PyPy support to FAQ
       new  528c75b   DOC typo
       new  e58405d   MAINT remove redundant class hierarchy + fix api perform parameter check in fit
       new  9a14d37   DOC fix example in univariate selection (fix issue #3132)
       new  defa6c4   FIX testing datasets have too few feature for default of SelectKBest
       new  db4230a   TST improve parameter checking + use nose assert_ functions
       new  185beb3   DOC remove comment thanks to @jnothman
       new  494a91b   Merge branch 'simplify-fselection'
       new  ac9bfdb   MAINT make export_graphviz more exception-safe
       new  5d47cfc   Improved formatting of unsupported sparse matrices in swapping
       new  1094149   Merge pull request #3140 from MechCoder/small_fixes
       new  2a40a4e   Revert "COSMIT skip some repeated computations in k-means"
       new  9d18b3b   DOC grid_search_digits.py does no do nested CV
       new  f097dd1   MAINT: ignore coveralls failures
       new  0d1bdea   Merge pull request #3145 from ogrisel/travis-coveralls
       new  6430e19   TST MiniBatchKMeans with many reassignments
       new  d62971d   DOC L1 distance works for sparse matrices
       new  3e4382c   added patch changeset_36ba287b29f3d039b8ecb5194dae1c47cf24641a.diff to "cherry-pick" 36ba287b29f3d039b8ecb5194dae1c47cf24641a (TST: fix test on scipy dev version)
       new  b6b136b   changelog
       new  2244098   Updated examples to use pyplot for plotting instead of pylab
       new  5e20b8c   Fixed pep8 violations. Some 'line too long' errors still remain.
       new  7d10d98   Fixed some PEP8 violations even present in master branch, in examples/
       new  248cf69   MAINT remove default values from private funcs in GMM
       new  3716e88   ENH faster sampling in GMMs
       new  96e9275   Merge branch 'pr/3120'
       new  c6b7e3b   FIX numerical stability in GMM with eigh sampling
       new  c2a3296   Merge pull request #3151 from rajatkhanduja/fixing_examples
       new  99f14e2   adding conditional import to lfw.py for python3 support
       new  cca6fc7   Some more examples updated to use matplotlib.pyplot
       new  7263d58   Modified example 'plot_stock_market.py' to use matplotlib.pyplot
       new  c944113   Merge pull request #3152 from jdowns/lfw-conditional-import
       new  6937fce   Merge pull request #3154 from rajatkhanduja/fixing_examples
       new  8dc8673   fix wrong confidence interval
       new  325e998   Extend utils.sparsefuncs: inplace scale and axis min/max
       new  974fb95   Remove HMM examples
       new  6469483   Adding string={pearson, spearman} option to increasing argument in IsotonicRegression.
       new  031e5e7   Adding increasing and decreasing tests for both Pearson and Spearman increasing argument options
       new  8523766   Merge pull request #3156 from IvicaJovic/fixes
       new  795a616   FIX: use clip(0) instead of abs()
       new  4a04b3f   DOC : fix alpha docstring in dict learning
       new  42920cd   Refactoring increasing_bool set into _check_increasing method
       new  fb9357e   PEP8ing isotonic tests
       new  5572775   PEPing isotonic regression
       new  645863e   DOC : really fix alpha docstring in dict learning
       new  d565386   Docstring style changes
       new  1f00663   Revert "DOC : really fix alpha docstring in dict learning"
       new  7c9a353   FIX unused pos_label parameter in metrics.precision_recall_curve
       new  3d04309   Change arguments to increasing={'auto', True, False} and default to 'auto'; implement Fisher transform and warning on 0 \in CI
       new  1b2a194   Adding test for CI check and removing Spearman/Pearson-specific tests.
       new  6eb98d9   Minor docstring cleanup
       new  e730a6d   Minor docstring cleanup
       new  f63e42e   Docstring fix
       new  d8099d6   Replacing the non-test .todense() methods with .toarray()
       new  724b63a   Replacing the test .todense() methods with .toarray()
       new  a4fd3a5   Improving tests to affirm no CI warnings are thrown when appropriate
       new  3b79250   Matching CI calculation and docstring with reference
       new  8186e33   Fixing docstring
       new  50d40fb   FIX : omp default param could fail
       new  4550aae   ENH : use local seed + avoid warning
       new  a015c4a   FIX: ENetCV and LassoCV now accept np.float32 input
       new  d9cc662   Merge pull request #3178 from MechCoder/Iss3174
       new  09e6e35   Refactoring check_increasing, ensuring that increasing_ is only set on fit/fit_transform, and fixing rho \in {-1, +1}
       new  e2d4c67   Reorganizing tests to isolate check_increasing from increasing='auto'
       new  5cb9b76   TST Refactor test of sklearn/ensemble/tests/test_forest.py
       new  c7a1db6   Fixing redundant np.asarray(X.toarray()) in metrics
       new  b9dc77d   TST refactor oob score testing
       new  4c3e7a9   Improving tests based on feedback from @GaelVaroquaux
       new  61a89ba   Improving tests based on feedback from @GaelVaroquaux
       new  01d4bd2   Fixing matrix vs. vector notation for X
       new  40048fa   Merge pull request #3182 from arjoly/test-forest
       new  216b4e5   CI: make travis run the doctests
       new  8e24535   Merge pull request #3189 from ogrisel/travis-doctest
       new  95619a9   DOC multiclass: OvO needs predict_proba or decision_function
       new  62e4975   FIX: more Python 3 fixes for doc/gen_rst.py
       new  92add1d   FIX TfidfVectorizer exports idf_ attribute
       new  d0fc60d   Merge pull request #3184 from YS-L/tfidfvectorizer_idf
       new  1a05e6d   FIX: euclidean divide in plot_image_denoising.py for Py3 support
       new  bbd7d91   PY3: more gen_rst.py fixes for Python 3 compat
       new  f2766ec   COSMIT: remove unused import
       new  edc5484   MISC: use precomputed variable
       new  d428606   MISC: update comment in source
       new  337d9e6   Typo.
       new  327abfe   DOC: better docstrings for PCA models
       new  0cb8377   Merge pull request #3192 from mmaker/master
       new  8725373   DOC fix typo
       new  73afdf1   Fixing space in docstring for default value
       new  16bf2f4   Adding check_increasing to classes.rst
       new  3fe9641   Adding no-warning assertions to IR auto tests
       new  27a7d36   Additional .todense() -> .toarray() fixes
       new  f4506d0   Fixing small formatting issue in doctest
       new  6c86a0e   FIX: windows test failures in test_ransac
       new  4a4d8e7   Fixing classes.rst
       new  40597f1   Merge pull request #3169 from mjbommar/issue-3167-eradicate-todense
       new  123aa8a   Adding Notes section to img_to_graph and grid_to_graph re: np.matrix->np.ndarray
       new  ec38713   Adding what's new item for sklearn.feature_extraction.image np.ndarray changes
       new  110dd5c   Merge pull request #3193 from mjbommar/issue-3167-eradicate-todense
       new  c5fc383   Workaround for andrewclegg/snake-charmer#12
       new  c217d10   Fixed typo
       new  b3cda92   Merge pull request #3196 from andrewclegg/master
       new  9094078   Switching from np to math for scalar float ops
       new  4baaaf3   Merge remote-tracking branch 'upstream/master' into isotonic-increasing-auto
       new  5cef947   Merge pull request #3157 from mjbommar/isotonic-increasing-auto
       new  8f2e208   Explicitly invoke the Python interpreter.
       new  46d27f5   Use double backticks.
       new  bf5798e   Remove extraneous period.
       new  eb86ff8   Explain the -1 Python syntax.
       new  bca3a4c   Additional docstring fixes for issue #3167
       new  6a7786a   Merge pull request #3197 from mjbommar/issue-3167-eradicate-todense
       new  5710edf   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  0e0da77   Adding tests to handle out_of_bounds parameter.
       new  2629fc7   Adding out_of_bounds parameter to handle values outside training domain
       new  b929656   Update plot_adaboost_regression.py
       new  c5fa168   Merge pull request #3200 from ugurthemaster/patch-2
       new  c1e5f94   ENH: PEP257 style + small code simplication
       new  376a9ac   Adding test for out_of_bounds argument validation
       new  eddad02   ENH: more explicit failure messages in test_common
       new  4884482   Reworking .transform conditionals for out_of_bounds argument
       new  5148b50   PEP8 fix
       new  d0f6052   Merge pull request #3199 from mjbommar/isotonic-out-of-bounds-2
       new  daa1dba   Merge pull request #3198 from apw/patch-1
       new  b8073a4   Improved documentation of the "estimator_params" argument for RFE and RFECV.
       new  83d56f2   DOC: DBSCAN: there's no calculate_distance function
       new  e75b2c4   Merge pull request #3208 from staubda/improve_RFE_doc
       new  85bdeba   DOC: improve feature_extraction.text docstrings
       new  181ffb8   MAINT: joblib 0.8.0
       new  1e64b7b   Merge pull request #3212 from ogrisel/joblib-0.8.0
       new  da43e4c   Add t-SNE
       new  7a1f84e   Generalize with average of nearest neighbors
       new  b88c7e2   Shorten documentation of perplexity
       new  4f2154a   Compare t-SNE with other manifold learners
       new  53e015c   Simplify optimization
       new  a2d209c   Refactoring
       new  67ad82f   Add first test
       new  56087e8   Test gradient descent
       new  6551b5d   Test binary search
       new  8c969cb   Add t-SNE to other examples
       new  58717d3   Add more tests
       new  63b5177   Use regular expressions
       new  405bf17   100% test coverage
       new  5de4c7d   Remove example with digits dataset
       new  006087f   Rename dist to affinities
       new  43b41f1   Modify learning schedule
       new  02f7d7b   Fix tests
       new  1b668d4   Explain example
       new  d7e915a   Rename attributes
       new  a2510de   Document attributes
       new  0567782   PEP8
       new  8d616fd   Update documentation
       new  b73ca8d   Do not stop too early
       new  6a0ba6c   Describe how to set the learning rate
       new  8d0fada   Remove generalization
       new  5cb4fa2   Add section in narrative documentation
       new  f628081   Optim: small optimizations
       new  25a8c5f   Adress Gael's comments
       new  cd5bf2c   Do not use global random number generator
       new  270d4d6   Replace squareform(pdist(*, "sqeuclidean"))
       new  58bcbaf   Add optional PCA init to t-SNE
       new  f378fc5   Adress Olivier's comments
       new  b59b844   Mention Barnes-Hut-SNE and fast optimization
       new  91e95df   Integrate PCA initialization
       new  08c17d2   Use PCA initialization in examples
       new  6487885   Fix docstring
       new  39fe650   Use euclidean_distances in original space
       new  e718603   Mention TruncatedSVD and clean up (PEP8, Pyflakes)
       new  222250d   Affinity must be 'precomputed' or 'euclidean'
       new  cda03fa   Rename arguments
       new  6b353c4   Allow sparse data
       new  f59e08f   Correct examples
       new  48c8ee6   DOC: Replace GT/LT with angle brackets for inner product
       new  b041039   Merge pull request #3217 from bwignall/innerprod
       new  92c4308   ENH Dense pipeline support for RandomTreesEmbedding via sparse_output param
       new  46aebf1   ENH: use partial sort for kneighbors selection
       new  c76a7a1   CLN: Fix typo
       new  92c187a   Merge pull request #3225 from bwignall/quickfix-typo
       new  bb0689f   ENH: make trees work with very small max_features.
       new  fed4692   TST skip flaky label propagation test
       new  34fc4d4   CLN: Capitalize "Gaussian" in example docstrings
       new  a611867   Merge pull request #3176 from agramfort/omp_default
       new  45e2a44   Merge pull request #3230 from bwignall/quickfix-gaussian
       new  9bee5b4   CLN: Capitalize "Dirichlet" and "Mexican" in example docstrings
       new  083b5f5   Merge pull request #3231 from bwignall/quickfix-cap
       new  dab017d   MAINT: run tests on files with the exec bit
       new  2729a74   Merge pull request #3234 from ogrisel/nosetests-exe
       new  04b73a4   Included sprse testing functions test_sparse_classification() and test_sparse_regression() in test_weighted_boosting.py, based on the corresponding testing functions found in test_bagging.py
       new  bcdc26d   Removed densification calls on X  '.asarry()' from all predict methods, removed 'dense' parameter from check arrays call, removed contiguous cast of the input array X, detected X as list or array to find the number of examples accordinly using len() or shape[0]
       new  ea8f194   In functions def test_sparse_classification() and test_sparse_regression() added tests for coo, lil, and dok sparse matricies. Removed parameter set feautre since there are not parameters that alter sparse/dense data usage. Revised data set to use dataset.make_{regression/classification} for improved test time.
       new  63934fb   Revised recuring import statement to respect conventions
       new  bffadaa   Updated documention to reflect sparse support on input data
       new  9e7351d   Removed prints statements used for debugging
       new  fef730a   Revied test_sparse_{regression,classification} to use make_multilabel_classification, revised import statements to adhere to one name per import conventions
       new  fd1e316   Enforced dtype for X, and clarified documentation by indcating the sparse formats supported and how they are treated
       new  372c0ed   Re inserted the check for X.ndim == 2 in initialization, updated a doc detail for X: coo is converted to csr only in the initialization not in the predict functions
       new  d55959a   Inserted base classifier check for BaseDesicionTree and and BaseForest before enforcing DTYPE in wieghted boosting fit
       new  df31dbe   Resolved python3 forest import error in weighted_boosting.py
       new  7d664ca   Updated documenation for weighted_boosting fit() to indicate behavior of forced dtype
       new  ecba04f   Included testing of the all of the predict/score functions of AdaBoost
       new  e7eb17c   Assert sample weight negative sum value error, add authors entry
       new  483796c   Fix score function parameter mistake in test_weighted_boosting
       new  bb752bd   Remove exectuable permision bit
       new  0234a57   Fixed typo toCSR -> tocsr
       new  1ddc225   Trivial doc alteration, travis timeout
       new  0a61119   DOC: kmeans runs inits in parallel, not distance computations
       new  e74c5ae   MAINT: bump up to scipy 0.14.0 in travis CI config
       new  795f377   Merge pull request #3237 from ogrisel/travis-scipy-0.14.0
       new  32f0069   Fixed paramaters to make_multilabel_classification and make_regression to keep tests under 1 second, reformated some blank lines
       new  96436df   Merge pull request #3161 from hamsal/ada-sparse
       new  fe6b9ea   fix defaultdict call
       new  c3ed6c2   Update what's new with adaboost sparse input support by @hamsal
       new  c18c2c1   DOC we work with Python, not against it :)
       new  0a4ba72   Merge pull request #2822 from AlexanderFabisch/tsne
       new  e47c68f   MAINT: joblib 0.8.1
       new  d298a37   Merge pull request #3242 from ogrisel/joblib-0.8.1
       new  18c399e   MAINT deprecate sequences of sequences support
       new  742505d   FIX remove duplicates in MultiLabelBinarizer
       new  37730f4   COSMIT use MultiLabelBinarizer in LabelBinarizer during deprecation
       new  bbc3e7e   COSMIT use mlb for MultiLabelBinarizer instances
       new  8ce9077   ENH do not allocate memory for temporary array of 1s
       new  255895d   DOC remove mention of sequence of sequences in Parameters sections
       new  f18a07a   FIX don't use dict comprehension for Py 2.6
       new  4dfde8b   FIX No set construction shorthand in Py2.6
       new  b2fe367   DOC comment on _transform interface
       new  5b4a3c8   TST Validate input MultiLabelBinarizer.inverse_transform
       new  dcb2417   No set construction shorthand in Py2.6
       new  fc0150d   DOC/FIX Address @arjoly's comments
       new  9cfa1f3   TST stronger test for non-integers in MultiLabelBinarizer
       new  28eb1eb   Assert or ignore all sequence of sequences deprecation warnings
       new  efede63   TST avoid more warnings related to sequence of sequences
       new  238def1   TST fix testing for sequence of sequences warning in metrics
       new  c013e26   ENH use the np.int dtype to encode integer classes
       new  37c65e5   FIX: Use coordinate_descent_gram when precompute is True | auto
       new  ae7a4ad   Remove unused param precompute from MultiTask models
       new  7b75ebb   FIX: Raise ValueError for invalid precompute
       new  c9048fb   FIX workaround doctest failure with old numpy
       new  5c46c0c   Merge pull request #3248 from MechCoder/remove_precompute_multi
       new  e4b2d96   Merge pull request #3246 from ogrisel/rebased-pr-2657
       new  dcd5da3   DOC add whats new for precompute fix
       new  ff8b12d   Merge branch 'pr/3247'
       new  d7c27d3   Refactoring, small improvements, and cleaning notation of IsotonicRegression
       new  fffdfaa   Merge pull request #3240 from JelleZijlstra/fix-defaultdict
       new  875138a   DOC more complete coverage of API linking in examples
       new  aca4925   TST add test for name identification in RST generation
       new  6f6a537   TST avoid nose running setup() in gen_rst.py
       new  b20f1db   TST add sphinxext to testing
       new  2728104   Changing docs from "vector" to "1d array"
       new  b117e97   FIX use NPY_INFINITY instead of C99 INFINITY for MSVC
       new  9f468f3   Merge pull request #3251 from ogrisel/fix-msvc-INFINITY
       new  f425134   ENH/DOC fix poly features complexity
       new  de42d69   ENH interaction_only in PolynomialFeatures
       new  7accbfa   Merge pull request #3239 from larsmans/faster-poly-features
       new  1c7e37e   COSMIT remove duplicate key 'hamming' in METRIC_MAPPING
       new  f309126   FIX PCA error handling for invalid n_components
       new  5b26438   Fix Machine Learning for NeuroImaging in Python link.
       new  f571834   Merge pull request #3254 from laurentluce/presentations-links-fix
       new  a30fafa   TST remove dependencies for testing sphinxext
       new  b16ffbb   DOC: fix 2 covariance examples rst math markup.
       new  bc8666f   Branching 0.15.X to prepare the 0.15.0b1 release
       new  15590e5   Added Windows Powershell download script
       new  8683c75   Merge tag '0.15.0b1' into releases
       new  7d1cc45   Merge branch 'releases' into dfsg
       new  0a249be   Merge branch 'dfsg' into debian
       new  72b6488   changelog, refreshed patches
       new  ba1914c   Build-depend on libatlas3-base
       new  60e8aea   TST test interaction features outside doctest
       new  67b8601   Minor fixes in the docs
       new  d12d313   Fix some fun column span alignment errors.
       new  bd7d423   Reverting array X to vector x rename
       new  6ce6d33   DOC fix deprecation in CD
       new  b54a9af   DOC missing verb in README
       new  bcaee04   Merge pull request #3260 from larsmans/cd-deprecation
       new  2df32d0   DOC : clarify error message in _binary_roc_auc_score
       new  7656498   TST silence spectral test/test warning message
       new  cc2eb23   FIX wrong parameter name in deprecation warning
       new  4a8a597   Docstring addition and x->X
       new  ad3156d   TST reduce max_depth for numerical stability
       new  f677630   Merge pull request #3261 from larsmans/test-interaction-features
       new  c7db70a   removed unused variables and imports, add used imports where missing.
       new  55ecfa2   COSMIT TYPO
       new  926e651   Stop using some deprecated interfaces.
       new  e2857b9   DOC typo in FastICA docs
       new  5b85451   Merge branch 'pr/3263': minor fixes from the 0.15.X branch
       new  42c3143   MAINT: restore -git version in the master branch after the last release merge
       new  ac0a6e7   Incorporating @ogrisel's comments for simplification
       new  5622cfb   More simplifications
       new  2ae8e2a   Fixing simplification for clip
       new  099df41   Adding test coverage for weight deprecation in IsotonicRegression and isotonic_regression
       new  e586256   DOC: remove broken example link
       new  3f5194e   MAINT removed shadowed broken test
       new  dcf48a5   DOC add Data Publica testimonial
       new  3636210   Improving exception and warning coverage
       new  662fe00   Merge pull request #3250 from mjbommar/isotonic-refactor-2
       new  20b9019   ignificant refactoring, will now automatically install Python, pip, and nose. Numpy and Scipy still require manual intervention
       new  35368d4   Fixed bug with append string getting double "_py_py"
       new  89296ef   WIP: releasing the GIL in the inner loop of coordinate descent
       new  5d68c27   Replaced numpy calls with blas in dual_gap
       new  7cf73f4   Made the following fixes 1. Replaced l1_norm with dasum calls 2. Changed strides to n_tasks in ddot
       new  ec34342   Replaced multiprocessing with threading
       new  5a839e6   ENH: Release the GIL for sparse coordinate descent
       new  acd5df3   COSMIT: Replaced xrange with range
       new  3e60331   ENH: Release GIL in the gram variant
       new  d497220   Replaced C ordered contiguity and added typed memory views
       new  84cac35   Added a unit test to ensure that there are no spurious repeating values in the thresholds returned by roc_curve because of machine precision, and a quick stab at a fix.
       new  70d3cce   Added typed memory-views for H and XtA
       new  d5351bb   Changed to using wheel packages for scipy and numpy, currently hosted on the sklearn Rackspace account.
       new  e1309ab   Slightly expanded the docstring for the test_roc_nonrepeating_thresholds test. Added named slice variables to same, for readability.
       new  07ce972   Improved the treatment of the machine precision issue for calculating ROC thresholds. Added a comment explaining why the treatment is needed.
       new  75b48cd   Release GIL in the multi-task variant
       new  752eec1   Additional functions, ensuring pip is correct for python 3 and passing the proper path
       new  c212242   Registry reloading is necessary the very first time Python is installed!
       new  a319840   Fixed PATH issues for initial Python install environment
       new  40bcd6e   hange directory to $PYTHONHOME to avoid encodings error
       new  72835d2   hanging directories is *not* useful for the encoding error
       new  95d1af7   Added longer delay after python silent install
       new  e7429be   dded verbose output when setting temporarily paths
       new  2c83678   ENH: Replaced double indexing as it causes python overhead
       new  a22225e   Minor refactoring of the repeated-roc-thresholds test, for clarity.
       new  6ea9fbe   Added a backport of numpy.isclose from numpy v1.8.1. This function was not available until numpy v1.7
       new  32e3bab   Changed metrics.py to use the new backward-compatible version of np.isclose
       new  58c0a75   PEP8 fix
       new  de824b1   FIX : allow nd X for cross_val_score (was working in 0.14)
       new  67e0704   FIX: examples with more than 10 images
       new  2f07590   DOC Correct definition of multiclass log loss
       new  5fa2f9c   DOC: copyedit log loss, hint how multiclass generalizes binary
       new  31f2e07   ENH save a bit of memory in Euclidean distances
       new  1081160   Merge pull request #3102 from MechCoder/gil-enet
       new  9772d95   DOC remove HMM documentation
       new  a75f44c   Merge pull request #3280 from larsmans/remove-hmm-docs
       new  9d3b48d   DOC fix versions of requirements in README.rst
       new  f8a70bc   Sparse input support in BaggingClassfier and BaggingRegressor #3241
       new  7a1a171   Sparse support in Bagging #3241
       new  4e19ef4   Merge pull request #3286 from ilam/seconddoc
       new  cdd2279   ENH fix astype usage to prevent copying
       new  69b50e1   DOC big fat warning about multithreaded BLAS
       new  060d264   DOC correct bagging docs regarding sparse inputs
       new  2931627   FIX RidgeClassifierCV didn't have scoring parameter
       new  f458662   DOC typo in ExtraTrees
       new  d580661   DOC add the Random Kitchen Sink synonym to the RBFSampler section
       new  72bcd6a   Merge pull request #3297 from esc/doc/random_kitchen_sink
       new  ce8b214   MAINT reduce amount of boiler code using standard C operations
       new  0309793   DOC give example of binarizing binary targets
       new  204e506   Updated to include git, and install all python versions by default
       new  68b0a28   DOC Note shape of binary binarized output
       new  59d6a11   bug fix and deprectation warning added to cross-validation
       new  10507dc   COSMIT factor out split data as struct
       new  aaefdbd   Merge pull request #3165 from jnothman/links_in_examples
       new  ee6a173   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into bootstrap
       new  d85f5e7   Merge pull request #3295 from rphlypo/bootstrap_traintest
       new  950005f   Added a few new features, and an example of chocolatey-style install
       new  89885a8   Missing the r in powe(r)shell
       new  42e9f5d   DOC fix bad formatting in decomposition.rst
       new  30347e9   FIX python 3 compat for plot_bias_variance.py
       new  fa76ac3   FIX OvR with constant label for non-predict methods
       new  bd0b93e   DOC safer way to run the tests under windows
       new  51d4f75   FIX case where label is constantly absent
       new  4098883   MAINT: symeig->eigh
       new  f892622   FIX Python 3 support for datasets.species_distributions
       new  6544ad2   FIX ensure that matplotlib.use('Agg') is called first when building the docs
       new  c7bbb8f   DOC formatting fixes in whats_new.rst
       new  fa491cf   DOC one formatting fix in whats_new.rst
       new  45a9056   FIX Python 3 support for examples/applications/plot_out_of_core_classification.py
       new  1ccf358   Merge pull request #3309 from argriffing/remove-symeig
       new  e331935   FIX Python 3 support for examples/applications/plot_prediction_latency.py
       new  d2db017   TST use assert_warns and modernise test for constant predictor
       new  82864ad   FIX avoid using sequences of sequences and fix tests
       new  7aee742   Merge pull request #3308 from jnothman/ovr_constant_predict_proba
       new  916050a   DOC hmm example has been removed
       new  9789fdb   FIX Anisotropic hyperparam optimization of GP bounds
       new  b98bb34   DOC remove reference to the bootstrap target
       new  dae905c   DOC remove spurious plt.clf() from ROC example
       new  5cab6c9   DOC add missing plt.figure() for multiplot support in ROC example
       new  708d3a3   DOC Document return of fetch_olivetti_faces
       new  2bfca14   ENH sparse matrix support in label binarization
       new  7a70e0f   FIX Forced dtype to int when densifying array in label_binarize
       new  6d4fd85   DOC Improve documentation of the olivetti dataset
       new  1eaf8db   DOC typos in feature_extraction.text
       new  e6da2c1   small spelling error in doc
       new  c29ef67   Merge pull request #3318 from abhishekkrthakur/master
       new  afcb384   Merge pull request #3256 from kastnerkyle/windows_download_helper
       new  8489c33   added a new section on model persistence
       new  df27e26   model persistence doc, added improvements from ogrisel comments
       new  be315ed   Move model persistence doc inside model selection section
       new  4c51809   Remove trailing whitespace
       new  78c0f61   Simplified security and maintenance section
       new  4b79f22   Metadata information for unpickling models in future versions
       new  6460479   Merge pull request #3317 from pignacio/model_persistence_doc
       new  877d471   Convert windows_testing_downloader.ps1 from UTF-16le to UTF-8
       new  777123d   Merge pull request #3268 from jblackburne/nonrepeating_roc_thresholds
       new  a00fffd   MAINT bump joblib to 0.8.2
       new  e14b716   Merge pull request #3328 from ogrisel/joblib-0.8.2
       new  0a3a677   DOC SVMs take string class labels as well as integers
       new  68280fb   MAINT remove residual sparsefuncs*.so when compiling
       new  b4e8b3c   Release 0.15.0b2
       new  7e9c093   tree: add min_weight_fraction_leaf
       new  216204f   forest: add min_weight_fraction_leaf
       new  d72f958   min_weight_fraction_leaf and min_samples_leaf: test both best-first and depth-first tree building
       new  72c60eb   gbrt: min_weight_fraction_leaf value test
       new  72b3da8   tree: use min_weight_leaf internally while exposing min_weight_fraction_leaf
       new  39f831d   utils.testing: add assert_greater_equal and assert_less_equal
       new  9871707   forest: test min_samples_leaf and min_weight_fraction_leaf
       new  79d965a   min_weight_fraction: comments and pep8
       new  6f4b31a   test assert_greater_equal and assert_less_equal
       new  f62a6d1   min_weight_fraction_leaf: narrative docs
       new  b9f6ac4   min_weight_fraction_leaf: narrative doc update
       new  849f89f   Update what's new
       new  ac45e5d   MAINT skip tests that require large datadownload under travis
       new  03a78d5   fix typo (on -> one)
       new  82611e8   Merge pull request #3334 from ami-GS/external
       new  9a4c9f2   MAINT add docstring to explain the motivation of the fixture
       new  5be7cc2   Merge pull request #3333 from ogrisel/maint-skip-download-tests
       new  3729143   Typo
       new  4bd4e9a   MAINT Include binary_tree.pxi in source distribution
       new  e7e49a6   ENH Add partial_fit to GaussianNB
       new  3146a1f   DOC what's new: GaussianNB.partial_fit + typo in comment
       new  3b8dc3d   Update Sphinx docs for GaussianNB partial_fit
       new  08b2902   Merge pull request #3343 from ihaque/doc_partial_fit
       new  cf82012   Remove duplicate GaussianNB.fit() code
       new  d709833   Preserve public API
       new  2d12952   docstring
       new  1775095   Merge pull request #3344 from ihaque/remove_dup_fit
       new  c176ced   DOC double backticks for fixed-width (code) font
       new  5358be5   TST crank up preprocessing tests to 99% coverage
       new  108e009   FIX Gaussian KDE should return array, not scalar
       new  e4561b2   COSMIT use explicit RandomState in KDE tests
       new  e633f25   ENH make_multilabel_classification for large n_features: faster and sparse output support
       new  5ea6b3a   COSMIT in response to @arjoly's comments
       new  4ecc59f   Update what's new
       new  ab1a784   FIX failing tests due to the change of the underlying sample generator
       new  9e56b2b   DOC more explicit docstring for preprocessing.normalize
       new  96895eb   Added utility to skip tests if running on Travis
       new  f2b3a09   DOC optimize some PNG images, shaves off 24kB
       new  cfa61ab   DOC run optipng before uploading website
       new  4f98ce2   Merge pull request #3355 from larsmans/optipng
       new  482c1df   DOC run optipng before uploading website
       new  5b55f57   Fix grammar
       new  3acda36   Merge pull request #3357 from houbysoft/patch-1
       new  f5d9b5b   Fix labels of inliers vs outliers
       new  76ef30b   TST fix precision failure on windows
       new  15aff5c   TST fix precision failure on windows
       new  cddf9a4   MAINT skip joblib multiprocessing tests on travis
       new  48e8cd6   Merge pull request #3361 from ogrisel/travis-no-multiprocessing-joblib
       new  f721cb8   MAINT skip joblib multiprocessing tests on travis
       new  178d0b6   FIX assert_array_almost_equal for windows tests
       new  046d87a   FIX assert_array_almost_equal for windows tests
       new  bd0f2d5   Cosmetic: correct latex formula display
       new  7d8eaf7   DOC: write cost function of logistic regression.
       new  75a5866   Merge pull request #3365 from fabianp/master
       new  de69793   Link my name to my wordpress blog
       new  1fe371e   Merge pull request #3366 from MechCoder/link-to-blog
       new  ec05d98   fixed typo: diriclet -> dirichlet
       new  e33b3b8   Merge pull request #3368 from stevetjoa/typo-diriclet
       new  948bd77   DOC updated installation documentation
       new  ec7cad3   Merge pull request #3369 from NelleV/install_doc
       new  7022f3e   DOC: typo, envelop → envelope
       new  b289d37   Merge branch 'pr/3358'
       new  c73c9a1   MAINT Re-order argument to put deprecated one at the end
       new  851d914   DOC: add transpose and intercept to formula of logistic regression
       new  c19aa38   DOC adding return type and improving module doc
       new  5972f5a   DOC return types in datasets.lfw
       new  6668adf   DOC numpydoc convention for Bunch-returning functions
       new  fbfcdc5   Merge branch 'pr/3316'
       new  85b9499   FIX multilabel deprecation warning in RidgeClassifierCV
       new  f433f66   FIX multilabel deprecation warning in RidgeClassifierCV
       new  e90e4c9   MAINT ignore sparsefuncs_fast in diffs
       new  b65e4c8   FIX MinMaxScaler behavior on 1D inputs.
       new  f08b8c8   MAINT explicit float64 in sparsefuncs_fast
       new  f453b49   MAINT mailmap update
       new  514f073   FIX #3372: unstable input check test for RANSACRegressor
       new  e3c622e   FIX #3372: unstable input check test for RANSACRegressor
       new  1f32603   Added FIXME annotation to skipped test
       new  d8a8a8d   Merge pull request #3353 from kastnerkyle/skip_ompcv_travis
       new  9e54c8b   Corrected two typos in docstring.
       new  f01ae49   #3356 - Added an exception raising when np.nan is passed into a HashingVectorizer.
       new  5242c21   #3356 - Added an exception raising when np.nan is passed into a HashingVectorizer.
       new  cdf3b7b   Merge pull request #3375 from ldirer/hashing_fix3356
       new  6e2af35   Added a test on hashing vectorizer behavior with np.nan input.
       new  d1fe163   PEP8 line length fix.
       new  d1a9af8   Changed solver from 'dense_cholesky' to 'cholesky' to eliminate deprecation warning.
       new  2eb1e98   MAINT ensure that examples figures are displayed in the correct order
       new  cfcb53d   MAINT: more explicit glob pattern in doc generation
       new  6d528d7   DOC: typo, envelop → envelope
       new  51dc919   DOC updated installation documentation
       new  b6b86b8   Corrected two typos in docstring.
       new  494b4a0   #3356 - Added an exception raising when np.nan is passed into a HashingVectorizer.
       new  700259a   MAINT ensure that examples figures are displayed in the correct order
       new  409ba88   MAINT: more explicit glob pattern in doc generation
       new  fb62b9f   Replaced assert_raises with assert_raise_message as the point is to make sure the exception message is clear.
       new  dfeabda   Merge pull request #3378 from kastnerkyle/dense_cholesky_warnings
       new  e701377   Added utility to skip tests if running on Travis
       new  720d450   Changed solver from 'dense_cholesky' to 'cholesky' to eliminate deprecation warning.
       new  25e1181   Additional skipping for omp_cv
       new  2d7e32b   Put check_travis in the correct place so that other tests are not skipped, and changed the CCA check to be more explicit.
       new  1e777cb   Lift division out of loop in _isotonic_regression.
       new  cf25b08   DOC: more readable whats_new
       new  5b5dba1   DOC: fix classes not referenced in docs
       new  0295fda   FIX: check with tolerance on lars_path
       new  cc36ba1   Cosmetic np.abs -> abs
       new  ca7ea92   DOC Refreshed the documentation to install scikit-learn
       new  ffaea27   TEST: metrics sample_weight test: almost_equal
       new  6a99871   Merge commit '0.15.0b2-14-g720d450' into releases
       new  9b47a2d   Merge branch 'releases' into dfsg
       new  3e4fc83   Merge commit '0.15.0b2-332-g9b47a2d' (dfsg) into debian
       new  279fd37   updated changelog entry
       new  3b0aec7   Added additional skip checking for train and pickle tests
       new  59b0f61   MAINT: skip some unstable transformers test under Win 32
       new  3829f71   MAINT configure Windows CI with appveyor.com
       new  052475c   DOC various typos and fixes for the installation doc
       new  b3384b7   DOC phrasing
       new  14d03f6   Merge pull request #3381 from kastnerkyle/skip_omp_cv_regressors
       new  5174565   TST add test for minibatch k-means reassigns.
       new  d7a557a   FIX bug where random state was reset in each call to partial_fit, such that reassignment either occured never or always.
       new  3f3d8ad   FIX bug where distances were of illegal shape
       new  5141b9e   FIX crash when to_reassign.sum() > X.shape[0]
       new  f130079   TST split test for partial_fit and fit
       new  b8ca2ac   TST add additional test for batch_size > n_samples in fit.
       new  f7d14b3   MISC rename random.pyx to _random.pyx
       new  1ea8fca   ENH backport np.random.choice.
       new  4c97c6a   FIX broken tests for minibatch k-means
       new  b008e5d   FIX use choice in minibatch_kmeans
       new  b654f04   FIX another (two?) bugs: the default code path didn't ever compute distances.
       new  7a04b8d   skip doctests
       new  c2d59d2   simplify tests
       new  72d9e25   remove redundant code
       new  730683e   FIX up the rest of the stuff. Introduce .5 as a magic number. Hurray for magic numbers?
       new  7f5a784   DOC more / better documentation
       new  894c530   FIX reset counts of reassigned clusters.
       new  a39049b   ENH never reassign everything, more meaningful tests.
       new  05e65a6   TEST: fix test failing due to numeric instability
       new  e191366   COSMIT: address misc comments
       new  c8397ec   Minibach k-means: change reassignment to uniform
       new  0861f89   MISC: avoid deprecation warning
       new  40f8b49   MISC: minor changes in MBKmeans
       new  4ccfe06   DOC: spelling and phrasing
       new  45aa833   MISC: better formulation
       new  a9ad5a2   Merge pull request #3363 from ogrisel/appveyor-ci
       new  1724fc0   MAINT: skip some unstable transformers test under Win 32
       new  1d288ba   MAINT configure Windows CI with appveyor.com
       new  0597dc1   Merge pull request #3376 from GaelVaroquaux/fix_mb_kmeans
       new  e8cb36f   MAINT master to 0.16-git
       new  48b8afc   MAINT manual backport of whats_new.rst reorg
       new  b621bbb   DOC whats_new.rst format fix
       new  4316947   DOC whats_new.rst format fix
       new  7605da7   Merge pull request #3382 from ogrisel/install-doc-update
       new  e599b96   TST add test for minibatch k-means reassigns.
       new  b6368bb   FIX bug where random state was reset in each call to partial_fit, such that reassignment either occured never or always.
       new  2a1dd50   FIX bug where distances were of illegal shape
       new  e22a051   FIX crash when to_reassign.sum() > X.shape[0]
       new  6f698e9   TST split test for partial_fit and fit
       new  5002838   TST add additional test for batch_size > n_samples in fit.
       new  dc3add6   MISC rename random.pyx to _random.pyx
       new  9e9a273   ENH backport np.random.choice.
       new  bedb68e   FIX broken tests for minibatch k-means
       new  a4db49e   FIX use choice in minibatch_kmeans
       new  18cd2a8   FIX another (two?) bugs: the default code path didn't ever compute distances.
       new  c7f5646   skip doctests
       new  725937f   simplify tests
       new  610d91a   remove redundant code
       new  7f1e256   FIX up the rest of the stuff. Introduce .5 as a magic number. Hurray for magic numbers?
       new  68d9d4f   DOC more / better documentation
       new  1499bdf   FIX reset counts of reassigned clusters.
       new  15cbb1b   ENH never reassign everything, more meaningful tests.
       new  280a94a   TEST: fix test failing due to numeric instability
       new  b005f00   COSMIT: address misc comments
       new  af4b7ef   Minibach k-means: change reassignment to uniform
       new  953615b   MISC: avoid deprecation warning
       new  c1d9c30   MISC: minor changes in MBKmeans
       new  6bea881   DOC: spelling and phrasing
       new  748465f   MISC: better formulation
       new  c9ac31d   DOC Refreshed the documentation to install scikit-learn
       new  0c7ca94   DOC various typos and fixes for the installation doc
       new  150b0d2   DOC phrasing
       new  f208316   Added a comment to explain the use of a test.
       new  1a720f1   MAINT .mailmap update
       new  bd41939   MAINT whats_new: contributors for 0.15
       new  03f9e8b   MAINT whats_new: contributors for 0.15
       new  7d564b3   FIX expit bug with out != None
       new  e4e4f85   MAIN fix version in doc
       new  ccba9a0   DOC Copy and paste error, thanks @larsmans.
       new  379d730   DOC Copy and paste error, thanks @larsmans.
       new  15582e7   DOC update news on the homepage
       new  d6cccaf   DOC update news on the homepage
       new  9d1c723   BUG: fix windows pointer size problem
       new  3e98d04   Added additional skip checking for train and pickle tests
       new  d3b40d8   DOC broken formating for the People list of 0.15
       new  fd4ee4b   DOC broken formating for the People list of 0.15
       new  be643c9   FIX: try to get windows working
       new  05d8a6e   MAINT preparing the 0.15.0 release
       new  ebdeddb   Merge pull request #3385 from GaelVaroquaux/fix_intp
       new  ca7b6df   MAINT update doc version navigation
       new  ddf39dc   MAINT update doc version navigation
       new  0994ffa   BUG: fix windows pointer size problem
       new  a899155   FIX: try to get windows working
       new  c7b3951   Additional skipping for omp_cv
       new  e976967   FIX: check with tolerance on lars_path
       new  5180029   Cosmetic np.abs -> abs
       new  e86ca90   Merge pull request #3377 from ldirer/hashing_fix3356
       new  4315e91   DOC: expand documentation for logistic regression
       new  5df6f21   MAINT: minor mailmap update
       new  801c55b   MAINT update doc version links in support.rst
       new  4be0d06   MAINT update doc version links in support.rst
       new  0fe0e22   MAINT point to the configuration of the docbuilder server
       new  9c1217c   TYPO remove mutli (did you mean Muttley?)
       new  d7f4142   ENH: Return attribute n_iter_ for linear models dependent on the enet solver
       new  947da72   TST: Added test to check that warm model converges faster than a cold one
       new  059c834   TST: Better to test that warm_start runs only once after the prev model has converged
       new  e65150a   TST: Added test to check that higher alpha converges faster
       new  272e9ce   Added return_n_iter parameter for enet (and lasso) path
       new  e072fbc   Turn useless line of code into descriptive comment
       new  f7e9527   FIX potential overflow in _tree.safe_realloc
       new  ca8ee09   Merge tag '0.15.0' into releases
       new  a9e747d   Merge branch 'releases' into debian
       new  fe97948   updated changelog
       new  715bdb5   Merge branch 'releases' into dfsg
       new  4b7488a   Merge branch 'dfsg' into debian
       new  85d39c3   Quick fix on grid_scores, updating final value with len(cv) self.grid_scores_ = scores / len(cv)
       new  9248948   test case on rtfe grid_scores fix
       new  5727099   MISC: avoid overridding figure in benchs
       new  94d3bd4   MISC: <= rather than < in tol check (KMeans)
       new  7e40d0e   MAINT temporary fix to handle cython renaming in 0.15
       new  279d8ce   MAINT 32 bit unstable tests are unstable on all OSs
       new  4e438a9   [DOC] minor fixes
       new  228bc99   MAINT typo in Makefile
       new  3c7818e   ENH allow y to be a list in GridSearchCV, cross_val_score and train_test_split.
       new  5fe7c2c   add test for safe_indexing, add another test for cross_val_score
       new  c4525dc   MAINT typo in Makefile
       new  251077d   MISC: <= rather than < in tol check (KMeans)
       new  c3eb6a9   MAINT 32 bit unstable tests are unstable on all OSs
       new  d8ecd15   FIX: Fit the data in tests
       new  a130b8a   Merge pull request #3349 from MechCoder/return_niter
       new  52dcf26   scorer: add sample_weight support
       new  be5826d   BUG: Support array interface
       new  8c8da21   BUG: Support array interface
       new  d146d23   Catch ConvergenceWarning in RandomizedL1
       new  3209281   ENH improve forest testing + avoid *args
       new  ea66557   ENH improve rand_int and rand_uniform
       new  e242ebc   ENH rename parameters in MockListClassifiers.
       new  4bf824c   TST non-regression test for CV on text pipelines
       new  60bd856   Merge pull request #3400 from arjoly/forest-test-oob
       new  f33e9ac   Merge pull request #3396 from amueller/less_loud_randomized_lasso
       new  9c8f656   Merge pull request #3397 from arjoly/tree-factor-rand
       new  802c0a6   Remove deprecated 'mode' parameter from sklearn.manifold.spectral_embedding
       new  2e7ef3c   Merge pull request #2694 from amueller/allow_y_lists
       new  b0b996c   ENH allow y to be a list in GridSearchCV, cross_val_score and train_test_split.
       new  26d15aa   add test for safe_indexing, add another test for cross_val_score
       new  ba88173   ENH rename parameters in MockListClassifiers.
       new  4970f75   TST non-regression test for CV on text pipelines
       new  e107e1f   Remove deprecated 'class_weight' parameter from RidgeClassifierCV.fit
       new  c870547   use more interesting range for C in logistic l1 l2 example.
       new  90839da   COSMIT pep8 fixes
       new  ea6df6f   TST ignore sequence of sequences DeprecationWarnings
       new  52d815a   Merge pull request #3406 from jnothman/minor
       new  5b247f9   move around examples for better structure.
       new  7b69ace   Remove deprecated 'precompute_gram' parameter
       new  0a953a1   Remove deprecated 'copy_Gram', 'copy_Xy' and 'copy_X' parameters
       new  97c3c89   ENH add label ranking average precision
       new  35c9018   DOC write narrative doc for label ranking average precision
       new  9d8d774   DOC FIX error + wording
       new  1c24e73   TST invariance testing + handle degenerate case
       new  4fefe23   flake8
       new  6f00678   FIX use np.bincount
       new  c3542ba   DOC friendlier narrative documentation
       new  f4ac2b6   ENH simplify label ranking average precision (thanks @jnothman)
       new  a80c58b   ENH be backward compatible for old version of scipy
       new  32c2692   Typo
       new  b3aaf1e   pep8
       new  7d905b8   Remove deprecated 'Gram' and 'Xy' parameters from OrthogonalMatchingPursuit.fit
       new  bb93ede   DOC remove confusing mention to mean average precision
       new  bbf7ae7   BUG: Support array interface
       new  7768a30   ENH: enable y to only implement the array interface
       new  81b86f1   COSMIT: fix indentation
       new  d104a34   BUG: fix validation bug
       new  c325423   ENH: support non-ndarray subclasses in supervised estimator
       new  e42ecd1   ENH: transformers work on non ndarray subclasses
       new  6e3472b   DOC: more comments
       new  bfd546a   COSMIT:
       new  814a3ea   FIX broken test
       new  ffea8d9   Add related project to website.
       new  85ae41c   FIX DBSCAN input validation
       new  ce29c70   Removed 'Gram' and 'Xy' parameters were still in test_omp.py
       new  8935df6   Merge branch 'master' into theilsen
       new  b12d1ac   Remove deprecated 'score_func' and 'loss_func' parameters from sklearn.metrics.scorer.check_scoring. Amend the code in all the other places they were used.
       new  c7e6456   BUG: Support array interface
       new  2599365   Remove 'copy_X', 'copy_Gram' and 'copy_Xy' documentation since these parameters have been removed
       new  689058a   Merge pull request #3412 from brentp/logistic_l1_l2_ex
       new  8de00d8   Merge pull request #3411 from lesteve/scheduled-removal-from-0.15
       new  46c1c0a   MAINT faster test_weight_boosting
       new  55271f3   TST speedup test_spectral_biclustering
       new  e7d096c   TST faster check_transformer_pickle by fixing n_iter
       new  3318c9c   TST speedup test_permutation_score
       new  98f212d   TST remove joblib tests from sklearn
       new  c191771   DOC improve documentatino thanks to @vene and remove mention of relevant labels
       new  7236ade   Tidy up by removing unnecessary local variables
       new  85356dd   Merge pull request #3414 from lesteve/scheduled-removal-from-0.15
       new  3f5e691   Merge pull request #3403 from ogrisel/speedup-tests
       new  4b6768f   Merge branch 'master' into theilsen
       new  f9773ba   ENH add sample_weight support to dummy classifier
       new  b080d93   ENH don't convert dataframes in grid search and cross-validation
       new  ae2c059   Merge pull request #3418 from arjoly/dummy-weight
       new  e6d8aa3   DOC update what's new
       new  9a898da   DOC sample_weight attribute
       new  e4503b1   Fix for old Numpy 1.6.3
       new  1119e3b   Added to comments to better explain last commit
       new  94bda65   FIX: Remove raw_coef_ attribute that eats up memory
       new  9f45e37   Modified sparse OvR to handle sparse target data
       new  caff874   DOC fix docstrings; add @hamsal to authors
       new  b6ed218   Use string argument for legend's loc parameter
       new  718f6ad   split test_common.py into checks and test file. move dataset generation into estimator_checks
       new  fa50e62   Remove 'DeprecationWarning: using a non-integer number instead of an integer will result in an error in the future' warnings
       new  eb2098f   Closes #2360. Fix tiebreaking.
       new  24a0cdc   DOC more intuition about corner case
       new  56bba96   DOC add documentation to backported function
       new  2a187cd   Closes #2360. Fix tiebreaking.
       new  71d4a21   some cleanups in common_test, speedup.
       new  e6e9fb6   Make fit_transform and fit().transform() equivalent in nmf
       new  f6e3189   slight speedup some cleanup
       new  dc482a2   ENH less nested code
       new  0239310   Remove a couple more 'using a non-integer number instead of an integer DeprecationWarning'
       new  2564f24   Make everything accept lists as input.
       new  104d866   Merge pull request #3430 from amueller/test_list_input
       new  3e6892e   Merge pull request #3433 from lesteve/remove-non-integer-deprecation-warnings
       new  5189cfb   ENH: add a median absolute deviation metric
       new  434c20d   DOC: add an example of robust fitting
       new  4113dfe   Added a parameter check on n_estimators in BaseEnsemble to raise error if it is not strictly positive.
       new  8ffdf66   ENH better default for test for SelectKBest and random projection
       new  f7b94f6   MAINT tree compute feature_importance by default
       new  e23d9c9   ENH Make all decision functions have the same shape. Fixes SVC and GradientBoosting. Closes #1490.
       new  7f49bc1   pep8
       new  46613e2   Merge pull request #3439 from arjoly/test-commons
       new  b0995f9   FIX encoding issue
       new  bd1686b   Merge pull request #2804 from arjoly/lrap
       new  ce0484f   DOC update what's new
       new  e7fa78d   TST: fix doctest
       new  0807e19   MAINT deprecate fit_ovr, fit_ovo, fit_ecoc, predict_ovr, predict_ovo, predict_ecoc and predict_proba_ovr
       new  110d3f2   DOC: better documentation for robust models
       new  d3cddfe   API: naming: CamelCase class -> camel_case function
       new  6d881d3   MAINT split sklearn/metrics/metrics.py
       new  16870ff   ENH + DOC set a default scorer in the multiclass module
       new  aa54ee7   MAINT flatten metrics module and avoid nested bicluster module
       new  c03b987   DOC typo + not forgetting single output case
       new  4ec8630   Merge pull request #3444 from arjoly/explain-default-score
       new  14837ec   ENH in sklearn.metrics set bicluster as a submodule of cluster
       new  58feda0   FIX import
       new  00d7349   DOC improve documentation and distinguish each module
       new  1e42a68   DOC link example gallery scripts rather than inline
       new  e0d8c11   COSMIT
       new  937f36c   DOC show referring examples on API reference pages
       new  927de5e   DOC ensure longer underline
       new  969a060   COSMIT Use explicit if/else in scorer
       new  6598ae8   TST default scorers with sample_weight
       new  6a4aa1d   DOC Update What's New
       new  70f4d47   Merge pull request #3401 from vene/scorer_weights
       new  c91a4af   ENH add a friendly warnings before deleting the file
       new  fa26f64   Merge pull request #3442 from arjoly/split-metrics-module
       new  8f68e54   Merge pull request #3445 from arjoly/flatten-metrics
       new  c03d4bd   Merge remote-tracking branch 'gvaroquaux/pr_2949' into theilsen
       new  41d02e0   TST: fix tests on numpy 1.9.b2
       new  f7549fd   Refactor input validation.
       new  8dab222   Merge pull request #3443 from amueller/input_validation_refactoring
       new  6e2a83b   remove check_arrays stuff and old input validation
       new  2fcba22   Merge remote-tracking branch 'upstream/master' into theilsen
       new  c235c3b   ENH add allowed_sparse named argument for @ogrisel
       new  5517bad   Merge pull request #3447 from amueller/input_validation_b
       new  77c6c60   FIX classes name in OvR
       new  96b6642   DOC mention doc-building dependency on Pillow
       new  0d57c23   Merge pull request #3327 from jnothman/examples_in_apiref
       new  98f2275   DOC make neural networks example appear
       new  be08122   DOC Fix example path
       new  fb87430   DOC fix 'Return' -> 'Returns'
       new  e79d2bb   FIX Py3k support for out-of-core example
       new  06b57ae   DOC add links to github sourcecode in API reference
       new  ff82157   Merge pull request #3460 from jnothman/fix_nn_example_doc
       new  7a7fca8   Merge pull request #2777 from jnothman/doc_linkcode
       new  0b8e63c   Update outlier_detection.rst
       new  7a9ad10   DOC fix opaque background glitch when hovering example icons
       new  0ea65bb   FIX better RandomizedPCA sparse deprecation
       new  b1521e8   Merge pull request #3470 from ogrisel/fix-sparse-randomized-pca-deprecation
       new  8567deb   FIX better RandomizedPCA sparse deprecation
       new  c2b2ce4   DOC: Fix class link
       new  85c9c84   MAINT move log_loss and hinge_loss to the classification metrics
       new  24bd8f9   Merge pull request #3472 from arjoly/fix-metrics-division
       new  bbb140a   ENH: Added n_iter_ parameters across all iterative solvers
       new  c838c38   FIX: Preserve public API by introducing return_n_iter param
       new  06f0973   TST: Test that after fitting n_iter is greater than 1
       new  d72ca59   Added n_iter attribute to OMP (and CV)
       new  c508965   MAINT: Fixed a few docstrings and cosmits
       new  88b05ff   MAINT remove deprecated PCA code
       new  2fca6ce   MAINT remove deprecated code from trees and forests
       new  259b487   MAINT remove deprecated sklearn.pls module
       new  c484a4a   MAINT remove deprecated code from preprocessing
       new  be93a75   MAINT remove sklearn.test
       new  c9789c7   COSMIT clean up tests with pyflakes
       new  6e4f02c   Implementation of logistic_regression_path.
       new  bb74497   Take into account @agramfort's comments.
       new  91ed4bb   Some fixes for LogisticCV object
       new  4d2e962   Docstring of LogisticRegressionCV
       new  9139307   Docstring
       new  4cb2d93   Refactor and some bug fixing
       new  582d782   Docstring
       new  8cdbc6f   FIX missing import
       new  0171d80   FIX: bug in LogisticRegressionCV
       new  668e0d1   Make tests deterministic
       new  af6c7ed   FIX: coef.shape
       new  9fc0618   Just to be sure
       new  48a8628   Add test
       new  8cfa8ad   BUG when fit_intercept=False
       new  1443465   Fallback for failing line search
       new  228f5b0   Remove warning (not needed any more)
       new  174205e   Compatibility for old scipy
       new  02c4a28   iterate some more on line search
       new  999c232   cosmetic
       new  12adea8   COSMIT
       new  53e2593   Replaced Label Encoder with Label Binarizer
       new  b088c2a   Replaced helper function _phi by special.expit
       new  137d40b   Do away with intercept helper functions
       new  4fe04e2   Refactor fit_intercept case
       new  2e3b92f   FIX: Fixed hessian value for intercept
       new  200b925   DOC: Add docs for helper functions
       new  ddb3395   ENH: LogisticRegressionCV can now handle sparse matrices
       new  b8c6b4c   TST: Add tests to explicitly check hessian, loss and gradient for fit_intercept
       new  a4d9f77   More docs and tests for LogisticRegressionCV
       new  7c935a3   FIX: Doctests
       new  51a1831   TST: Improved tests
       new  f9871cd   ENH: Added one-vs-all fit in case of multi-class data
       new  d4b544e   ENH: Added refit parameter
       new  191119b   TST: Tests to verify OvA behavior
       new  60f454f   FIX: PEP8 and other cosmits
       new  7b3bfef   Made the following changes 1. Fixed random state 2. Object dtype support 3. Remove outdated "don't test multiclass" tests
       new  10a411f   ENH: Logistic Regression now supports newton-cg and lbfgs
       new  e411129   ENH: Weighted logistic regression for lbfgs and newton-cg
       new  5859f8b   Changed copy default from True to False, updated docstring for sample_weights
       new  4d516e8   DOC: Minor changes
       new  86dab1e   ENH: Added warnings for convergence, added support for l1 penalty if solver is liblinear
       new  21a14fd   FIX: Changed tolerance of newton-cg to be compliant with that of lbfgs
       new  8eb3d30   COSMIT: Utils imports are together
       new  0634ddf   FIX: Class weights are computed for each OvA
       new  f7db8cd   FIX: Liblinear solver for LogisticRegressionCV now works for class_weight==auto
       new  c8540df   FIX: Add DataConversionWarning
       new  a52174b   FIX: Changes due to recent refactoring of the check functions
       new  67585f6   MAINT: Improve documentation and coverage
       new  2ea804e   Update whats new!
       new  4de4849   FIX: Fixes for the cross_validation failure
       new  e9bb143   MAINT disable verbose tests on travis
       new  9dd5348   MAINT Move the build_ext --inplace call to install.sh
       new  1053950   Merge pull request #3441 from ogrisel/travis-no-verbose-tests
       new  2374a09   FIX: Increase testing accuracy
       new  56057c9   MAINT remove deprecated code
       new  ebc3f0d   DOC typo in warning
       new  384ff19   Merge pull request #2862 from MechCoder/LogCV
       new  62f1f57   Added directory checking for documentation builds, and corrected for Windows pathing
       new  8cd5d85   Merge pull request #3477 from kastnerkyle/docbuild_fix
       new  5e028a9   Improve docstring of the LogRegCV model
       new  965b109   Merge pull request #3479 from MechCoder/improve_logcv_docs
       new  d814353   MAINT More robust windows installation script
       new  0af2d8f   MAINT move skip for unstable 32bit to _check_transformer
       new  f3afd4e   FIX unstable test on 32 bit windows
       new  e93ad9a   MAINT More robust windows installation script
       new  8552c48   FIX unstable test on 32 bit windows
       new  376ac51   Merge pull request #3465 from pvnguyen/patch-1
       new  9e875bf   adding avg weights and flag to plain_sgd method
       new  2f4ea66   asgd is added
       new  67622b4   adding test that computes the average sgd
       new  d085d88   checkpoint for classifier
       new  a7a696e   fixing tests
       new  5ae2d7a   adding documentation fixes and changing avg to average
       new  4b6a5d2   fixing documentation
       new  e6d9bd7   adding comments to linear algebra operations
       new  6958f22   updating doc
       new  e0d92a4   more documentation changes
       new  11d3965   FIX numerically unstable test_logistic.py
       new  5cd2da2   DOC fix references to moved examples
       new  4aac610   TST: Fix warnings in np 1.9
       new  ccf661b   TST: Fix warnings in np 1.9
       new  db5be98   Changed _partition_estimators  signature to make it compatible with warm start.
       new  9131413   Added warm start to random forests.
       new  1f6d602   DOC whats_new.rst entry for warm_start forests
       new  2438bb8   MAINT skip tests that require large datadownload under travis
       new  b446291   MAINT add docstring to explain the motivation of the fixture
       new  82b3de7   adding test that contains a simple asgd implementation and compares it with the output of sgd_fast
       new  66ce06d   adding more precision to almost equal
       new  22604d0   ENH: enable y to only implement the array interface
       new  9e873b9   FIX 0.15.X-style fix to make the NotAnArray test pass
       new  c0afd46   DOC mistake in comment
       new  f505221   Modified the confusion matrix example: included a normalized matrix, changed the colors and added class labels.
       new  e181932   Merge pull request #3454 from ldirer/confusion_matrix_example
       new  d5430e3   DOC fix doc errors in utils.testing
       new  5f32d8d   TST stabilize check_transformer_n_iter
       new  b3bdb08   DOC fix formatting of attributes etc. in docstrings
       new  fcec373   converting to setting coef after the fit has been made
       new  09596ca   Typo
       new  0c5a777   putting const on pointer arguments
       new  1f040b4   FIX revert changes introduced by mistake in previous commit
       new  55be9f0   Merge pull request #3492 from furukama/patch-2
       new  16adbd5   MAINT: Effectively remove deprecated parameters
       new  1a2dcf3   DOC fix styling of See Also sections
       new  70f520c   DOC fix styling of method signatures
       new  5301e28   DOC: added Machinalis testimonial
       new  99c2502   COSMIT smipler & more robust check in test_kfold_valueerrors
       new  fe5423f   DOC fix see also references
       new  3893cc3   DOC move Attributes section to after Parameters and style likewise
       new  b61ab4e   Merge pull request #3491 from ogrisel/fix-warning-test-cross-val
       new  c002c5f   DOC fix see also reference
       new  1a985fc   DOC fix markup error
       new  e017490   MAINT fix prng in test_f_oneway_ints
       new  10c7527   MAINT fix prng in test_f_oneway_ints
       new  cf4cf60   ENH Stochastic Coordinate Descent for ElasticNet & Lasso
       new  63fc314   adding test for binary classifier
       new  a38577c   adding support for partial fit and adding test for partial fit
       new  e894205   FIX Implemented correct handling of multilabel y in cross_val_score
       new  dbfc8c6   FIX Implemented correct handling of multilabel y in cross_val_score
       new  adcc42c   FIX left over conflict marker from previous backport
       new  cbcae04   FIX Support unseen labels LabelBinarizer and test
       new  056d5fb   FIX Support unseen labels LabelBinarizer and test
       new  3fad0a2   adding support for averaged intercept
       new  1b2833a   Add sample_weight parameter to metrics.jaccard_similarity_score
       new  e147df5   adding test for intercept part of asgd
       new  0e31553   ENH Sparse multilabel target support
       new  a15434c   DOC label indicators are clearer as ints than floats
       new  15be75e   DOC more precise input type for multilabel metrics
       new  6a0e7f7   adding support for multiclass with a test included
       new  69961c0   changing float to double
       new  9fb2dfb   MAINT use assert_warns
       new  9a6287b   MAINT Fix regression random projections work sparse matrices
       new  195dd26   DOC whats_new.rst: missing backported fixes for 0.15.1
       new  f887cce   DOC whats_new.rst: missing backported fixes for 0.15.1
       new  7c2f2ac   adding comments to xnnz implementation
       new  bdd5cb1   Test sparse target data with dummy classifier
       new  c04ba19   Fit dummy classifier with sparse target data
       new  f7efdc9   Op directly on y.data in fit, xrange -> range, y.tocsc
       new  73f2f0c   Share array intialization sparse and dense fit
       new  c0981d2   Reorder fit and share denes sprs, Scafold sprs pred
       new  4052878   Remove temporary code from test_sparse_target_data
       new  388ac61   remove zero from random uniform return data
       new  a120827   Emulate sample weight support from dense (Rebase)
       new  0a381ed   Add four tests for sparse target dummy, one for each strategy
       new  7d9114c   Correct the class priors computation in the sparse target case
       new  8c3a986   Remove redundant indices, data, indptr appends in predict
       new  ff12d77   Test sparse stratifed corner case with 0s
       new  69adaa4   Combine dense column check in fit
       new  5c5ef1a   Update y with y.tocsc (Not an inplace op)
       new  05a8dd7   Select nonzero elements in predict stratified case
       new  3d692a9   Implement uniform and stratified w/ sparse random choices, optimize concats
       new  fd58c28   np.random.choice -> choice (from utils/random)
       new  72885ef   Use array for sparse matrix construction, fix constant check
       new  f96a6aa   Replace np.random.choice with utils.random.sample_without_replacement
       new  d0f52b6   Raise warning unifrom fit sparse target, Raise error uniform precit sparse target
       new  91a5b92   Support matrix output random_choice_csc, Update usage in dummy predict
       new  6e9d166   Validate lengths equal each element of classes and class_probabilities
       new  319d96e   Test random_choice_csc
       new  292590c   Test ValueError for length mismatch between array in classes, and probs
       new  ffd13c3   fix: i might be undefined
       new  65f4f95   Absract class prior construction to function sparse_class_distribution
       new  25f72c4   FIX six issue with module imports
       new  3edfa3c   Max iter param in Liblinear is now softcoded
       new  c8c72fd   ENH: Return n_iter_ from liblinear and print convergence warnings
       new  ad5773c   TST: Added test to check ConvergenceWarning
       new  e6b1863   Update plot_tree_regression.py
       new  6c7f029   Merge pull request #3522 from ugurthemaster/patch-1
       new  07560e4   FIX #3485: class_weight='auto' on SGDClassifier
       new  a2eb0d7   MAINT skip 32bit-unstable test in the 0.15.X branch
       new  de9de2d   FIX #3485: class_weight='auto' on SGDClassifier
       new  d490693   MAINT preparing the 0.15.1 release
       new  e1bbd9a   Test sparse_class_distribution, Correct data indexing
       new  bf43d5b   FIX wording in whats new in 0.15.1
       new  cb27e4c   FIX wording in whats new in 0.15.1
       new  44499fa   Move array intiialiaztions into dense conditional, nz_indices -> col_nonzero
       new  1f4ff67   Merge pull request #3509 from arjoly/fix-rp-sparse
       new  e1891d8   Merge pull request #3489 from jnothman/attributes-doc
       new  1491c1c   DOC remove backticks from around attribute name + PEP8
       new  e65d907   DOC A less-nested coverage of model evaluation
       new  29b57e0   Fix heisenbug due to addition of max_iter param in LSVC
       new  0a7bef6   Merge pull request #3528 from MechCoder/fix_heisen
       new  cfad359   Rever formating of warning, Reword warning for sparse uniform case
       new  2ec8e97   Correct spell 'Predicting', doument (sparse_target_input_ -> sparse_output)
       new  7bb1c01   Reverse conditionals in fit and predict, positive case first
       new  31d9b12   Make cosmetic revisions to random_choice_csc
       new  acb4e1a   Merge tag '0.15.1' into releases
       new  5122bf9   Merge branch 'releases' into dfsg
       new  d75538e   Merge branch 'dfsg' into debian
       new  0a8726a   changelog
       new  11c7538   import divison from future to give real results with two ints
       new  aa47285   Make cosmetic changes to sparse_class_distribution
       new  f5622b7   Make cosmetic changes to sparse_class_distribution
       new  431f291   Make naming changes in sparse_class_distribution
       new  2207046   Make default parameters for test_random_choice_csc, and use almost equal
       new  0ee10dc   move random_choice_csc to utils/random from utils/sparsefuncs
       new  81f8268   Move test_sparse_class_distribution to utils/multiclass from utils/sparsefuncs
       new  7e694fa   Uodate imports to correct function loactions
       new  478f9ad   Comment .eliminate_zeros() in sparse class distribution
       new  0943c92   adding faster implementation of asgd with sparse data
       new  22cafa6   Merge pull request #3527 from jnothman/unnest-model-evaluation
       new  5ba7df8   Include dense case support in class_distribution, pep8 revisions
       new  7912b1d   Make sparse target dummy tests multioutput-multiclass
       new  6c6ef87   Test class_distribution w/ multioutput-multiclass sparse and dense
       new  40d8fb3   Replace numpy.random.choice with a search sorted strategy for faster runtime
       new  f59e4da   Test random_choice_csc implicit, readability adjustments
       new  0e31472   Remove transposes from testing of random_choice_csc
       new  22ef22e   Clarify unfiorm sparse warning message, Reword sparse_output_ doc
       new  214b761   Clarify unfiorm sparse warning message, Reword sparse_output_ doc
       new  f69dcef   Use UserWarning in place of SparseEfficiencyWarning
       new  222cb87   Remove outer for loop in dense case predict
       new  69af355   FIX only use testing.ignore_warnings in tests
       new  fe8e719   Add sample_weight parameter to metrics.log_loss
       new  54774d0   DOC add link toward Jatin Shash webpage
       new  a7c9128   ENH removed manual code for parallel task batching forests
       new  aeb94c6   ENH: parallel feature importances for forests
       new  bedf14a   FIX Exact inverse transform for whitened PCA
       new  dc2bc8a   TST test whitened inverse
       new  3f6c61f   ENH components of unit length and whitening done by scaling with sqrt of explained variance
       new  83223fd   DOC What's new
       new  7b7784e   Make cosmetic changes
       new  1c01da0   ENH faster safe_indexing for common case
       new  c66a14e   Merge pull request #3539 from jnothman/fast_indexing
       new  1e1bd68   COSMIT Move imports in example
       new  8122420   FIX Verbose message in RFE to count from 1, not 0
       new  ddfb616   changing api to have average_sgd and plain_sgd
       new  9acc643   seperating out the average and plain apis in stochastic_gradient.py
       new  67a0a91   Merge branch 'pr/3535'
       new  046f7e8   MAINT NumPy 1.10-safe version comparisons in joblib
       new  94b7845   Test that class dtypes string and float fail in random_choice_csc
       new  1873ef8   Test explicit sample wieghts in the sparse case of class_distribution
       new  a5acf9d   Test insertion of 0 class in random_choice_csc
       new  acac25d   DOC fix typo
       new  553592e   adding sgd_fast.c
       new  7cfea35   Change dtype check conditional to look for everything other than int
       new  2a3f504   Make cosmetic adjustments, Fix random.py header, use message with assert_warns
       new  9783de3   Test additional corner cases with random_choice_csc, error on proabilites not summing to 1
       new  2203422   Combine sparse and dense test for class_distribution
       new  2198a84   Manage explicit zeros manualy in class_distribution, test with explicit zeros
       new  9e4aba2   FIX Using random_state.rand rather than scipy.rand consistently in Gaussian Process
       new  4698481   DOC fix typo
       new  aa53fdb   DOC specify X shape for precomputed
       new  012bc0a   FIX: Incomplete download of 20newsgroup_dataset
       new  517373f   DOC update what's new
       new  1bdd2ce   TST regression test for optimum over several random-starts of GP
       new  c604ac3   Merge branch 'pr/3545'
       new  9da134e   DOC:Typo in doc
       new  b186348   DOC GP fixes in what's new
       new  f38c1ca   Merge branch 'pr/3395'
       new  a52085f   DOC correct default value
       new  ec4f574   fixing typos
       new  143e5ff   FIX set vectorizer vocabulary outside of init
       new  8680a6b   Deprecate vectorizer fixed_vocabulary attribute
       new  9e6ce6d   adding explanation of asgd to doc
       new  a6777cd   adding if self.average logic
       new  bbf3426   DOC add missing details to what's new
       new  5202312   MAINT simplify covertype benchmark
       new  6d8ccbc   DOC correct shape of Tree.value
       new  9088aca   COSMIT update sklearn.svm.bounds
       new  39da06f   Merge pull request #3438 from hamsal/sprs-out-dmy
       new  ebbc5b1   adding test to make sure plain sgd does not have average parameters
       new  c25467e   MAINT Remove note that assert_warns comes with Numpy 1.7
       new  c149516   Update plot_kernel_approximation.py
       new  c7ba56e   Merge pull request #3572 from AndrewWalker/patch-1
       new  a46130e   DOC typo
       new  55cb591   DOC vectorizers were referring to a private function in public docs
       new  56c539e   ENH more explicit error message for ill-posed problem
       new  ca3ad7b   FIX: Memory crashes for _alpha_grid in ENetCV
       new  05aa804   Merge pull request #3573 from MechCoder/high_dimensional_enetcv
       new  0f89415   MAINT: add license info
       new  e58764f   DOC more explicit parameter descriptions in make_multilabel_classification
       new  82945e5   FIX PLSRegression again supports 1d target
       new  10e4769   FIX out-of-core example had been broken
       new  c2a88c2   MAINT: update joblib to latest release 0.8.3
       new  f0b2bbd   MAINT: better gitignore
       new  341ca93   DOC avoid the plot_ prefix in example where no plot
       new  763af96   DOC fix line references in tutorial
       new  00c7585   FIX more intuitive behavior for *SearchCV.score
       new  42fbeae   changing sgd to asgd
       new  13ca2b2   ENH multinomial logistic regression using L-BFGS
       new  d5ceb89   ENH: Add support for class weights
       new  5483cb2   FIX: Change alpha to 1./c
       new  4cd1777   ENH: Merge MultinomialLR into LR with multi_class='multinomial'
       new  e6f7ab3   ENH Added multinomial logreg to plot_classification_probability.py
       new  6b5b9e3   ENH: Added multinomial option to LogisticRegressionCV
       new  4e79127   COSMIT: Minor doc fixes
       new  b2e0e6e   TST: Tests for multinomial logistic regression
       new  89226b6   COSMIT: Made the following changes 1. Removed true from 'true multinomial loss' 2. Replaced OvA with OvR 3. Update whatsnew.rst
       new  d80ef50   DOC: Improved documentation and error messages
       new  43845d7   DOC: Made the docs for LogisticRegression clearer
       new  a92cd80   FIX: PEP8 Errors and unused imports
       new  5560236   DOC: Changed docstring style for optional arguments
       new  5024503   DOC document MultiLabelBinarizer.sparse_output param
       new  43c284f   adding comparison for ASGD
       new  5be010f   DOC: Explicit instructions for Python3
       new  7ceb307   Merge pull request #3586 from MechCoder/python3_installation
       new  ab1e724   Added test for orthogonal LDA transform.
       new  afb675c   Fixed LDA transform.
       new  798c49a   Improved precision of the LDA orthogonality test.
       new  c0b92e5   forgot to remove an import
       new  49cee4b   misc
       new  c902399   added test for transformed scatter matrix
       new  089cb07   DOC Updated what's new
       new  69ff0b2   DOC fix link embedding regexp
       new  c259c14   DOC allow link embedding to proceed when web sites unavailable
       new  787e4a0   DOC extend documentation on sample generators
       new  da7f009   Merge pull request #3001 from jnothman/doc_sample_gens
       new  8597cc4   Update gaussian_process.py
       new  ad751a3   FIX bug with set_params on minkowski 'p'.
       new  d080429   DOC deprecation of **kwargs in neighbors per 0.18
       new  4b82379   TST fix doctest for k-NN
       new  796d67f   Merge pull request #3588 from tejesh95/patch-1
       new  959a4ac   MAINT: Remove BaseLibLinear for LogisticRegression
       new  9e58bcd   DOC: Added np.sqrt since default is 'squared=False'
       new  d6bfe09   Merge pull request #3593 from MechCoder/row_norms_doc_fix
       new  34a4517   ENH: Improved verbosity and denesting
       new  a128220   FIX six issue with module imports
       new  e18a524   FIX only use testing.ignore_warnings in tests
       new  49a7fe1   MAINT: be robust to numpy's DeprecationWarning
       new  5e98a2f   FIX: Fix the imputation example
       new  2626f83   Repeated word: 'the the' -> 'the'
       new  7ff3a32   Merge pull request #3598 from GaelVaroquaux/np_warnings
       new  fc0b231   Merge pull request #3604 from rahiel/patch-1
       new  c1be2ee   Merge pull request #3603 from MechCoder/fix_imputation_example
       new  dd78f10   DOC rename "imputation.py" to "missing_values.py"
       new  be09af1   MAINT: be robust to numpy's DeprecationWarning
       new  84ac04d   fix_gnb_proba
       new  a7ccbb8   MAINT: Made the following changes 1. Use a single loop for both ovr and multinomial conditions 2. Better warning messages for penalty and loss conditions 3. Convergence warnings print only if verbose > 0 4. Used check_X_y instead of check_X
       new  eeffac6   DOC dtype parameter on validation functions
       new  75e11c0   Merge pull request #3591 from MechCoder/refactor_logreg
       new  ed0e6d7   DOC multi_output parameter on validation function
       new  6c69cb7   DOC: more explicit title
       new  0535fe9   BUG: n_samples instead of n_features in cd_fast
       new  02c71ef   BUG: n_samples instead of n_features in cd_fast
       new  1db8448   ENH: minor speed-up in k-means
       new  f37618a   ENH: Remove unused copy
       new  0cc63a7   COSMIT: explicit comment
       new  224910d   DOC tweak BaseSearchCV.score docstring notes
       new  ba7e7a4   DOC a further tweak to BaseSearchCV.score docstring
       new  dd9e512   Changed f=open() to with open() as f to eliminate ResourceWarnings. Fixes #3410.
       new  455f415   FIX #3566: redefine isfinite alias in sklearn
       new  353840c   Merge pull request #3608 from agramfort/fix_gnb_proba
       new  cf3d69b   Moved code out of context block where not required for #3612
       new  77ff749   Merge pull request #3612 from calvingiles/fix-resource-warnings
       new  8d3f4df   ENH cosmetic reorg of the confusion matrix example
       new  2c5fea1   FIX: divide each row of the CM by the true support
       new  d4e5173   Merge pull request #3580 from jnothman/gridsearch-score
       new  14eb82b   removing standard_coef and standard_intercept from plain sgd
       new  97355e9   Merge pull request #3611 from ogrisel/fix-isfinite-windows
       new  aecf2c0   FIX bump up the miniconda installation script
       new  42ca0a0   Merge pull request #3619 from ogrisel/fix-miniconda
       new  6641781   Adding newton-cg solver for multinomial case.
       new  273fdc0   Add comments to _multinomial_loss_grad_hess function.
       new  8559fbc   Fixed pep8 errors. Added docstring for _multinomial_loss.
       new  2c25322   Fixing docstring on _multinomial_loss.
       new  9f94035   FIX #3566: redefine isfinite alias in sklearn
       new  93cbcef   FIX bump up the miniconda installation script
       new  5c8b6c8   fix typo (on -> one)
       new  9adf643   DOC double backticks for fixed-width (code) font
       new  88796ef   FIX Gaussian KDE should return array, not scalar
       new  01b6fc0   COSMIT use explicit RandomState in KDE tests
       new  3163cd8   DOC optimize some PNG images, shaves off 24kB
       new  b684a88   Fix grammar
       new  82bd4ea   DOC: write cost function of logistic regression.
       new  12c298f   fixed typo: diriclet -> dirichlet
       new  8080548   Fix labels of inliers vs outliers
       new  b84e7a1   DOC adding return type and improving module doc
       new  de42bc5   DOC numpydoc convention for Bunch-returning functions
       new  7a7a4b9   FIX MinMaxScaler behavior on 1D inputs.
       new  e43a244   Lift division out of loop in _isotonic_regression.
       new  b954c52   DOC: fix classes not referenced in docs
       new  264cefd   MAINT Include binary_tree.pxi in source distribution
       new  80c388d   TEST: metrics sample_weight test: almost_equal
       new  689351e   FIX bug where distances were of illegal shape
       new  233ffdf   FIX expit bug with out != None
       new  a78aafd   TST hashing vectorizer behavior with np.nan input
       new  9024bd4   DOC: expand documentation for logistic regression
       new  b0bea23   TYPO remove mutli (did you mean Muttley?)
       new  1c13999   Turn useless line of code into descriptive comment
       new  133aefe   FIX potential overflow in _tree.safe_realloc
       new  c651319   FIX RFE + test
       new  fc7d345   MISC: avoid overridding figure in benchs
       new  e4f6029   [DOC] minor fixes
       new  591f18b   Catch ConvergenceWarning in RandomizedL1
       new  bef087e   MAINT get rid of NumPy DeprecationWarnings
       new  4c09307   Added a parameter check on n_estimators in BaseEnsemble to raise error if it is not strictly positive.
       new  88c7b4a   DOC mention doc-building dependency on Pillow
       new  e0b0ccf   FIX Py3k support for out-of-core example
       new  725981d   DOC: Fix class link
       new  647a26a   Update outlier_detection.rst
       new  4516af5   Typo
       new  0e11161   DOC fix see also reference
       new  8068f67   DOC fix markup error
       new  f9eb2d0   fix: i might be undefined
       new  6119356   DOC move Attributes section to after Parameters and style likewise
       new  39acdb4   COSMIT Move imports in example
       new  7f65304   FIX Verbose message in RFE to count from 1, not 0
       new  5515c79   MAINT NumPy 1.10-safe version comparisons in joblib
       new  7fb2aef   DOC fix typos
       new  1db2fe9   DOC specify X shape for precomputed
       new  5a4905c   FIX: Incomplete download of 20newsgroup_dataset
       new  07c1dda   DOC update what's new
       new  0b93c12   FIX: random_start feature in GaussianProcess
       new  3d1b2ee   FIX: gp_diabetes_dataset examples (theta_ attribute)
       new  bfa4d7c   FIX Using random_state.rand rather than scipy.rand consistently in Gaussian Process
       new  2e609bd   TST regression test for optimum over several random-starts of GP
       new  87750a1   DOC:Typo in doc
       new  17d327b   DOC GP fixes in what's new
       new  413a21c   DOC correct default value
       new  2510464   FIX set vectorizer vocabulary outside of init
       new  f7310ea   Deprecate vectorizer fixed_vocabulary attribute
       new  2eae002   DOC correct shape of Tree.value
       new  ed38e68   DOC typo in plot_kernel_approximation.py
       new  ff77a64   DOC typo
       new  8aa69b0   DOC vectorizers were referring to a private function in public docs
       new  3645354   MAINT: add license info
       new  ae7e147   FIX out-of-core example had been broken
       new  6a9db13   MAINT: update joblib to latest release 0.8.3
       new  18055af   MAINT: better gitignore
       new  de66f6a   DOC avoid the plot_ prefix in example where no plot
       new  9add691   DOC fix line references in tutorial
       new  cba514b   DOC: Explicit instructions for Python3
       new  db1daf5   FIX LDA transform to project onto most dicriminative direction
       new  509f249   FIX bug with set_params on minkowski 'p'.
       new  daf591d   TST fix doctest for k-NN
       new  9c71994   DOC deprecation of **kwargs in neighbors per 0.18
       new  48f2984   Update gaussian_process.py
       new  a416c18   DOC: Added np.sqrt since default is 'squared=False'
       new  d3c16a6   Repeated word: 'the the' -> 'the'
       new  f0f9668   FIX: Fix the imputation example
       new  164dc8e   DOC: more explicit title
       new  7c2cd61   ENH: minor speed-up in k-means
       new  dc5d57d   ENH: Remove unused copy in k-means
       new  46a1250   fix_gnb_proba
       new  7658dd6   Changed f=open() to with open() as f to eliminate ResourceWarnings. Fixes #3410.
       new  020e0af   MAINT optimize machinalis.png, -9kB/42%
       new  b4e41cb   FIX update sklearn.__all__ to include all end-user submodules
       new  4c692e9   FIX update sklearn.__all__ to include all end-user submodules
       new  7f72713   FIX define CC and CXX for travis
       new  c09a4ca   FIX define CC and CXX for travis
       new  a61fb87   utils.testing: add assert_greater_equal and assert_less_equal
       new  8dc8995   FIX Windows CI: use prebuilt numpy / scipy
       new  dfe642e   FIX Windows CI: use prebuilt numpy / scipy
       new  e77b8ef   MAINT remove harmless, but useless, double call to csr_matrix
       new  79646ff   MAINT use %r for better printing of regexps
       new  b1119bb   ENH friendlier message for calling predict before fit on SVMs
       new  52adb5c   ENH Add 'axis' argument to sparsefuncs.mean_variance_axis
       new  842d80a   ENH improved CSC matrix handling in VarianceThreshold
       new  dbfb25e   FIX heisenfailure in test_lasso_lars_path_length
       new  0fcace7   Merge pull request #3629 from ogrisel/fix-unstable-lars-test
       new  f848a04   FIX heisenfailure on 32 bit python + speedup
       new  ff94896   FIX finfo.eps should be used instead of tiny for equality checks
       new  37ac314   MAINT remove unused MST code
       new  775f6de   ENH upload Windows wheels to rackspace
       new  57630c4   FIX heisenfailure in test_lasso_lars_path_length
       new  9b10b3c   FIX heisenfailure on 32 bit python + speedup
       new  9c6f415   ENH upload Windows wheels to rackspace
       new  5c5e83c   DOC fixes in 0.15.2
       new  2b42e95   DOC fixes in 0.15.2
       new  fe7ff98   Release 0.15.2
       new  2068ff2   FIX appveyor.yml path
       new  f5300d1   ENH Speed up and simplify cartesian product
       new  a228e8a   FIX division by zero warning in LassoLarsIC
       new  c453711   MAINT big refactor of sklearn.cluster.bicluster
       new  cb1798a   FIX more tolerant early stopping for least angle
       new  090b1c1   MAINT clean up flake8 complaints in k-means
       new  6409357   ENH: Precompute distances only if overhead is below 100MB
       new  45e82c3   Merge branch 'pr/3613'
       new  3c92686   Merge pull request #3564 from jnothman/bounds
       new  308c7ef   FIX consensus score on non-square similarity matrices
       new  d9288f0   ENH micro-optimize gradient boosting
       new  891839c   DOC: Explain prediction when decision_function is zero
       new  181c720   TST: Cleanups in test_theil_sen
       new  e44f9c7   COSMIT: Renamed _lse to _lstsq in theil_sen.py
       new  09d44f1   ENH: Removed shared-memory parallelism in theil_sen
       new  970e4c2   COSMIT: Inlined two methods in theil_sen.py
       new  57f67d0   MAINT CI: reflect CLI change in wheelhouse-uploader tool
       new  00071bd   ENH: Use warnings instead of logging in theil_sen
       new  1314697   ENH: Removed _split_indices method in TheilSen
       new  d105145   ENH: Rewrote TheilSen._get_n_jobs as a function
       new  43b5e43   COSMIT: More explicit names for vars in theil_sen
       new  8dbec4f   Merge branch 'master' into theilsen
       new  3e70d84   FIX: usage of check_array in theil_sen
       new  030f7e1   FIX: Use check_consistent_length in theil_sen
       new  9db5ba1   ENH: Refactoring in theil_sen
       new  b28e2b6   ENH: Removed unnecessary generator in theil_sen
       new  49d3043   FIX: doctest of get_n_jobs
       new  f41352c   Fixing more pep8 errors.
       new  3e2ce55   Merge remote-tracking branch 'upstream/master' into multinomial_newtoncg
       new  6c96fbe   Merge tag '0.15.2' into releases
       new  66172bb   Merge branch 'releases' into dfsg
       new  e3292cf   Merge branch 'dfsg' into debian
       new  fdafd8d   changelog
       new  171139f   ENH add testcase for issue 2445
       new  7d2179b   ENH: Theil-Sen vs. RANSAC example
       new  3915600   debian/rules - fixed version for comparison against sphinx. Now documentation should get built on recent releases (Closes: #760730).
       new  cdfa0ce   Removed some unnecessary lines and reformatted for consistency in _multinomial_loss and _multinomial_loss_grad_hess.
       new  f69e79f   added macro-average ROC to plot_roc.py
       new  150b491   OPT: Prevent iterating across n_features twice
       new  5fc0e8f   Merge pull request #3653 from MechCoder/minor_opt
       new  149d97f   making optimal for asgd the constant learning rate
       new  8081745   fixing merge conflicts
       new  042764c   added macro-average ROC to plot_roc.py
       new  8cee66b   rebased upstream/master with master
       new  1ceb885   rephrase doc, rename unseen, fix pep8
       new  1ec01ea   Modified parameter description in docstring of _multinomial_loss_grad_hess.
       new  9af14b3   DOC restructure data transformation user guide
       new  499ed00   Merge pull request #3587 from jnothman/transforms-doc
       new  d9067fa   FIX #3370: better lars alpha path inequality checks for 32 bit support
       new  78fbd25   ENH comment on drop for good not being triggered in the tests
       new  a5852c6   changing averaging to use sparse trick
       new  3bf6975   fixed duplicated part in plot_roc.py
       new  7c29eca   DOC fix broken URL
       new  e95a7f0   adding value error for partial fit with auto weights
       new  0985ef6   adding note to whats_new.rst
       new  3bd58c9   removing miss-merge and adding github
       new  3b292b2   adding a more descriptive error message
       new  38f5b6c   adding escape characters to regex
       new  9c079cf   addressed jnothman's minor points.
       new  af4df33   Adding test cases for multinomial LogisticRegression using newton-cg solver.
       new  356c24e   typos: remove extra s's
       new  25a327c   tweak to more natural version
       new  0d9f6cc   Fix wrong attribute doc in BaggingRegressor.
       new  12f63da   Merge pull request #3667 from ajschumacher/patch-6
       new  81613d2   make IsotonicRegression pickleable
       new  dd4143b   import pickle
       new  7f4c974   Sample weights for gradient boosting
       new  f98388e   more tests
       new  ddc5123   add exponential loss to narrative documentation
       new  2503d35   cosmit: doc raise ValueError for sample_weights and robust regression
       new  7064e20   bincount supports weights
       new  5832998   more elegant PriorProbabilityEstimator
       new  6fe6c34   adressed Olivier's feedback
       new  297b36a   update tests for new exception types
       new  fd952cc   sample weight support for robust regression via weighted percentile algo
       new  93b2b7d   fix: consider sample_weights in robost init estimator and negative_gradient
       new  d992596   fix: add **kwargs to Multinomial loss' negative_gradient
       new  75f6e01   more tests for weighted percentile
       new  73a8c3a   cosmit
       new  2d50328   add GBRT sample weights to whats new
       new  50fea07   DOC document copy_X parameter from LinearRegression
       new  05fbc64   fix: sample_weight is None in Huber.deviance fix: sample_weight multiplication in Huber leaf updates more tests (non-uniform weights, weight consistency and invariance to scaling, deviance consistence)
       new  4966b12   cosmit arr -> array
       new  7485752   FIX: Make sure LogRegCV with solver=liblinear works with sparse matrices
       new  a57a306   Correct documentation for TfidfVectorizer
       new  04ee897   Merge pull request #3676 from danfrankj/master
       new  dc3b00d   Removed n_jobs parameter from the fit method and added it to the constructor
       new  d34e928   Merge pull request #3674 from MechCoder/fix_l1_penalty
       new  25c03a7   Added Deprecation warning
       new  4f2ff2d   Small PEP8 fix, n_jobs attribute is compatible with both fit method and the constructor
       new  077a1ee   Fixed a failing test in the doc
       new  ef7169a   Added tests
       new  da64670   Added seperate tests
       new  701a87f   Refactored some logic code
       new  c57e6ed   PEP8 fixes
       new  9580431   Merge pull request #3678 from akshayah3/LinearReg
       new  266fddb   Wording.
       new  862d821   Modified hessp function in _multinomial_loss_grad_hess to compute r_yhat in place. Simplied conditional statements that branches based on solver + multi_class input combinations. Extended tests on multinomial LogisticRegression to include newton-cg solver case. Updated doc string of LogisticRegression to include newton-cg solver in multinomial case.
       new  c7a002a   Update whats_new.rst. Add contact information to logistic.py.
       new  b742d58   Modified doc/modules/linear_model.rst to incorporate the newton-cg solver option for multi_class="multinomial".
       new  e6fa800   non-uniform weight test case for both reg and clf and all losses
       new  85b2c78   add probability output for exponential loss
       new  de11662   we can produce probabilistic outputs using exponential loss
       new  63ec9ac   Fix typo.
       new  7dec87c   Merge pull request #3224 from pprett/gbrt-sample-weight
       new  c5fdee4   MAINT add a deprecation version
       new  88d473c   MAINT move comment to the appropriate place
       new  41b9385   FIX don't compare strings with 'is'
       new  5335539   FIX propagate MemoryError from Tree._resize
       new  39c8701   cleaning up commented out code and previous_coef_ parameter
       new  324e3c9   ENH sort option for memory-efficient DictVectorizer
       new  7354359   MAINT refactor DictVectorizer's transform+fit_transform (1)
       new  8f18396   Merge branch 'pr/3683'
       new  cd88472   FIX error handling/memory deallocation in Liblinear wrapper
       new  2181f7c   TST more robust test for MemoryError from Tree._resize
       new  a7a0512   MAINT: handle frombuffer with empty 1st arg in utils.fixes
       new  800354b   DOC single-pass DictVectorizer in what's new
       new  3cbbe7d   ENH: Use the scipy C-based L1 distance if possible
       new  b2d6a30   IncrementalPCA implementation
       new  2887c15   + Fix the bug of calling len when using sparse matrix for fit_params in cross_validation.py.
       new  f0f26bc   fix factor of 2 in RBFSampler; make test more rigorous
       new  f1f9760   + Fix the bug of calling len when using sparse matrix for fit_params in cross_validation.py. + Add test case for the above fix.
       new  ad815a0   Merge branch 'fix_bug_in_cross_validation_when_using_sparse_matrix_for_fit_params' of https://github.com/queqichao/scikit-learn into fix_bug_in_cross_validation_when_using_sparse_matrix_for_fit_params
       new  fd4ba4d   MAINT: set attributes as last action in DictVectorizer.fit
       new  5f8271f   Updated what's new to add IncrementalPCA
       new  7426e6a   Merge pull request #3285 from kastnerkyle/incremental_pca
       new  f7af204   clarify KernelDensity.score{,_samples} docstrings
       new  5d3b42f   DOC Fix n_jobs documentation
       new  56ad4de5  Merge pull request #3697 from luispedro/cv_doc_fix
       new  b0ab714   Merge branch 'master' into theilsen
       new  a800040   COSMIT: Small changes regarding Theil-Sen
       new  3de9f22   DOC: Better documentation for Theil-Sen
       new  ef17f80   ENH: Improvements in the Theil-Sen regressor
       new  2a94ae2   Merge pull request #3687 from perimosocordiae/fast-l1-dist
       new  766fb56   FIX allow ndim>2 in shuffle
       new  2666b70   TST stronger test_shuffle_on_ndim_equals_three
       new  d1ae779   Merge branch 'pr/3696'
       new  cd3c2db   Merge pull request #3695 from dougalsutherland/kde-docs
       new  a54590b   Merge pull request #3656 from dougalsutherland/fix-rbf-samp
       new  961bf67   ENH: Shortcut for 1d case in spatial median
       new  374ac6d   ENH: Avoid trailing \ in test_theilsen imports
       new  e7baa77   Minor doc and formatting changes.
       new  3c5c4f5   fix the link of Out-of-core_algorithm
       new  5569035   Merge pull request #3705 from floydsoft/patch-1
       new  6e123c0   Fixes ovr in the binary classifier case, and adds support for lists of feature arrays in the multiclass case.
       new  11d3696   Added make_fit_parameter method
       new  16bc265   dded tests to test_boosting
       new  4d94392   Added _check_sample_weight
       new  367845c   Shifted the _check_sample_weight into BaseWeightBoosting class
       new  0241b65   GradientBoostingClassifier docstring incorrectly specified default for max_features as "auto" when it is None.
       new  30202d4   Merge pull request #3715 from trevorstephens/correct-gbc-docstring
       new  5fe405b   cleaning up floating point and unneeded todos, also removing constant learning rate for asgd
       new  c4ba5da   Merge pull request #3713 from akshayah3/Boost
       new  aa66dea   DOC Add solido testimonial
       new  b584ac4   FIX ovr predict_proba in the binary case
       new  6ea371a   Merge branch 'pr/3710'
       new  3a02da8   Raise exception for sparse inputs in the case of svd solver
       new  211dfa4   Rename k to n_clusters in docs
       new  7522923   Merge pull request #3724 from mlopezantequera/patch-1
       new  0d6a8de   removing documentation about constant learning rate for averaging
       new  74b9563   MAINT use Python versions provided by AppVeyor CI
       new  4e2960d   FIX call srand whenever random_seed >= 0 in libsvm
       new  58be184   Enable grid search with classifiers that may throw an error on individual fits.
       new  c1eb8f9   Fix sparse_cg solver when max_iter is specified.
       new  cf3b476   FIX: TheilSen -> TheilSenRegressor in docs
       new  9d9bbc0   Merge pull request #3730 from ogrisel/fix-libsvm-random-seed
       new  d204967   MAINT missing numpy / scipy for the Python 3.3 build
       new  4860126   + Fix a typo.
       new  09896dc   plot_adaboost_twoclass.py: minor improvements
       new  972fcba   Merge pull request #3731 from ndawe/master
       new  8357f17   Merge pull request #3732 from queqichao/fix_typo
       new  7b6cb91   adding add_average method and solving iteration bug by replacing definition of t
       new  009f7de   adding note for partial_fit and n_iter
       new  e5ae879   clean up and average can now be set to an int indicating the number of iterations before averaging
       new  e915fcc   Changed default argument of precompute in ElasticNet and Lasso
       new  7cb3ba1   TST: Added tests to test Deprecation warning
       new  2071518   DOC on negation of loss functions used as scoring parameters
       new  da74962   DOC reduce table width through less-qualified function names
       new  f8473af   Reformatted docstrings in multinomial functions. Refactored multinomial functions to avoid duplicating code.
       new  c3b0a69   Wrap _multinomial_loss_grad in lambda function for passing into lbfgs.
       new  984e9aa   increasing testing precision, putting comment on one line and adding asgd to regression benchmark
       new  d5a3ff9   FIX ensure that pipeline delegate classes_ to the estimator
       new  5e2d965   pep8
       new  d7c17a1   TST bagging of pipeline of classifier
       new  e579171   DOC update what's new
       new  d6c9129   changing parameters for asgd regression benchmark and changing array shape for docstring
       new  d0ea44d   adding space to shape and putting // in average division
       new  90e5b2a   DOC: Document criterion_ attribute in LassoLarsIC
       new  70d910b   Merge pull request #3740 from MechCoder/lassolarsicattribute
       new  e93d313   DOC for the Git novices: cd to the clone directory
       new  5b3208a   debian/patches - deb_skip_some_tests_on_arm -- skip two tests on armv to avoid FTBFS
       new  263f398   up_workaround_buggy_pil -- workaround a bug in PIL
       new  fd57806   MAINT remove integer warnings under Python 3 + numpy 1.8+
       new  1a29be6   fix svm.fit in feature_stacker.py
       new  fd8dc24   removing spaces for tuples of width 1
       new  0bd095f   FIX: Raise warnings in f_classif a given feature is constant throughout
       new  22cc25f   ENH: More descriptive error which prints the feature indices
       new  bee0a98   cosmit in warning message
       new  85f70f8   Update feature_stacker.py
       new  ef43e7d   DOC: Narrative doc for median_absolute_error
       new  53d1711   ENH: Reworked _modified_weiszfeld_step
       new  afda6b0   DOC: Improved _spatial_median docs
       new  89d5aa0   COSMIT: Replaced xrange by range
       new  51cef86   COSMIT: Renamed y -> x_old in _modified_weiszfeld_step
       new  d3de579   COSMIT: Renamed spmed[_old] to spatial_median[_old]
       new  1f67b56   COSMIT: Break and for .. else in _spatial_median
       new  222e23c   adding asgd to whats_new.rst
       new  9ce2444   ENH: Reworked _lstsq in theil_sen.py
       new  a0f0d46   fixing merge conflicts
       new  389d13d   ENH: Replace AssertionError by ValueError
       new  25746e2   COSMIT: Improved error message in theil_sen.py
       new  8955fd0   COSMIT: Renamed n_all to all_combinations in theil_sen.py
       new  08d98f7   Merge pull request #3480 from dsullivan7/sgd
       new  961a2ea   COSMIT: Consistent naming for n_subpop
       new  27ef88f   COSMIT: Fixed pep8 problem in theil_sen.py
       new  5618ce3   FIX: Thresholded Nearest Centroid fails with non-encoded y
       new  4633c52   DOC: Moved notes section to long description
       new  a5d2fd9   Merge branch 'master' into theilsen
       new  d826dec   Modify shape check on coef matrix in _logistic_regression_path.
       new  7ac7fdc   Merge pull request #3746 from MechCoder/nearest_neighbor_manhattan
       new  89c8175   Bug fix in computing the dataset_centroid in NearestCentroid
       new  54eb478   FIX: Remove test that tests state of warnings before and after assert_warns
       new  701cd8a   Merge pull request #3669 from pprett/fix-isotonic-pickling
       new  5776517   ENH: Added median_absolute_error to metrics
       new  afc372a   DOC: Added doc in model_evaluation.rst
       new  7f7fa1b   ENH: unit tests for median_absolute_error
       new  aadd504   DOC: Added return doc to _lstsq
       new  562f4e4   COSMIT: Better variable names for _modified_weiszfeld_step
       new  d9bdddd   COSMIT: Moved epsilon to module level
       new  9c1b0c0   COSMIT: Some empty lines for better readability
       new  0dfa4d0   Merge pull request #3742 from abhishekkrthakur/master
       new  5dfbc78   ENH: test_common.py unit tests pass
       new  c9e711b   FIX: Small typo
       new  0298d9d   Merge branch 'master' into median_absolute_error
       new  076bad6   ENH: Removed median_absolute_error due to PR #3761
       new  5b3a387   ENH: Made n_subpopulation a fit parameter
       new  bb3705c   DOC fix (Nu)SVR attribute shapes in docstring
       new  4d1692c   FIX sign flip in (Nu)SVR with linear kernel
       new  cd7b43c   TST make catch_warnings blocks more robust
       new  1e4492a   Merge pull request #3761 from FlorianWilhelm/median_absolute_error
       new  8478d3c   Merge pull request #3750 from MechCoder/bug_nearest_centroid
       new  20a5952   ENH added digits as optional arg for classification_report
       new  9a3ad6a   COSMIT: Some renamings and PEP8 compliance
       new  9ebed28   Merge branch 'master' into theilsen
       new  657c86f   ENH: Removed 1d shortcut in _spatial_median again
       new  8e5347b   COSMIT: Clearer slicing syntax in _modified_weiszfeld_step
       new  6564d23   ENH: Fixed confusing X = X.T renaming
       new  7a1a35d   COSMIT: Some renamings in _lstsq
       new  923619f   FIX: Fix of last merge with master
       new  c377ccd   COSMIT: Renamings for easier understanding
       new  ec663e6   COSMIT: Another slicing syntax cleanup
       new  859992a   TST: Test that the warning registry is empty after assert_warns
       new  29c21f0   Revert "ENH: Removed 1d shortcut in _spatial_median again"
       new  777cd87   ENH: sample without replacement
       new  85d9c59   FIX warning check in test_affinities
       new  04f6d12   fix to a minor bug with intercept
       new  031a3fc   Merge pull request #3763 from ogrisel/fix-warning-test-affinities
       new  34df94d   TST: Test that assert_warns is reset internally
       new  4213c04   DOC: Make it explicit that assert_warns clears the warning registry
       new  64fc5c9   Merge branch 'master' into theilsen
       new  e3d1c32   Merge pull request #3752 from MechCoder/remove_testing_test
       new  a5618f9   Merge pull request #3770 from dsullivan7/asgdintercept
       new  a2d7f89   fixing duplicated classifier in asgd test
       new  8d82d2a   Merge pull request #3773 from dsullivan7/sgdtestfix
       new  6744be2   Merge pull request #3646 from s8wu/multinomial_newtoncg
       new  ea1548a   Add quantile strategy to DummyRegressor. Fixes #3421.
       new  d2fa6a8   Update what's new
       new  5e1faef   Expose positive option in elasticnet and lasso path
       new  2810f5a   COSMIT reshape for all regression strategy + avoid xrange
       new  6b0b773   COSMIT  less nested constant check + use comprehension
       new  4fcd6b8   COSMIT: pep8 and renaming
       new  b8cd60b   COSMIT: replaced assert by assert_less/greater etc.
       new  b698d9f   Revert unwanted modificatoin
       new  da31344   Add sample_weight support to Dummy Regressor
       new  e0f4cd9   Use np.average instead of np.mean
       new  13c2b00   TST: Add tests and document whats_new.rst
       new  0db66ba   Merge pull request #3778 from MechCoder/expose_positive
       new  4aea364   DOC: FAQ: what to do with strings
       new  d5ec39b   TEST: No console output during unit tests
       new  f9ecbf7   ENH: Always set random_state in unit tests
       new  f92074a   ENH: Speedup of unit tests
       new  c54c17e   COSMIT: Better consistency
       new  b27ee40   Update what's new: full name of Staple
       new  2137d82   ENH: Added random_state in plot_theilsen.py
       new  beb0843   DOC: fix doc inconsistency in GMM
       new  2b6a9fa   Merge pull request #3783 from jakevdp/gmm-doc-fix
       new  535d1f6   DOC fix Isomap docstring formatting
       new  0bdaecd   bug fix for issue #3526
       new  e324fdd   bug fix for t-SNE (issue #3526) with new inputs
       new  7b01c0b   Remove trailing '\' and add test
       new  bf7993e   Describe default metric
       new  6f41bd1   ENH: Patches Nearest Centroid for metric=manhattan for sparse and dense data
       new  91450c6   FIX: Wrap csc_row_median around the _get_median imputer function
       new  9622e85   MAINT: Move _get_median into sparsefuncs to avoid circular imports
       new  3e7b816   DOC: Explain why the centroid of the manhattan metric is the median
       new  2daa028   Renamed csc_row_median to csc_median_axis_0
       new  bda6706   Warning for non-euclidean and non-manhattan metrics
       new  8fc9918   Sparse matrix conversion depending on the type of metric
       new  e871972   Update what's new.rst
       new  30619ff   Merge pull request #3772 from MechCoder/manhattan_metric
       new  d4405cb   MAINT remove deprecated oob_score_
       new  a0e5fcb   Merge pull request #3786 from AlexanderFabisch/tsne_fix
       new  92a02ad   adding example for memory wrapper
       new  51d6fa2   A short cleanup
       new  2c2e1e2   Merge pull request #3795 from dsullivan7/svm_doc
       new  9be8718   recovering deleted variable
       new  6b548f9   MAINT remove deprecated loss in gradient boosting
       new  d71c529   Handle_unknown option to OneHotEncoder
       new  afae1c7   DOC add missing public function into the references
       new  f0405a2   DOC: Remove extra # from url in fit.
       new  b7ee11b   Merge pull request #3797 from Titan-C/clean
       new  a383729   Merge pull request #3801 from cmd-ntrf/typo_vbgmm
       new  72db6ae   DOC: Make the comment slightly clearer
       new  9c1c811   Merge pull request #3800 from MechCoder/unknown_transform
       new  626f672   COSMIT: Minor opt in pairwise_distances_argmin_min
       new  b500e3e   OPT: Speed improvements by avoiding repeated calls to check_array
       new  3f49cee   Merge pull request #3811 from MechCoder/fix_repeated_checking
       new  2f275de   ENH Add cross_val_predict function
       new  6deaea0   FIX score precision in doctest
       new  8eee4bc   Implements Multiclass Hinge loss
       new  e6835a7   Merge pull request #3779 from arjoly/sw-dummy-regressor
       new  29c7181   Update what's new
       new  d3f4936   FIX #3503: use linalg.pinv to better deal with singular input data
       new  c1efa9a   FIX in KDE fit and score allow y=None, add test for use with pipeline and gridsearch
       new  2ca16ba   Fix using ground truth in silhouette score in document clustering. Closes #3806.
       new  87b23ba   DOC explain SelectorMixin strategy with multiple classes.
       new  d21d6dd   Sample weight support for hinge loss
       new  e01f225   COSMIT: Use searchsorted in weighted_percentile
       new  c03c01a   ENH Bring sparse input support to tree-based methods
       new  0bc8a98   FIX+ENH add min_weight_fraction_split support for sparse splitter
       new  7cb9a5c   Re-organize code dense splitter then sparse splitter
       new  0e86dfd   Simplify call to extract_nnz making it a method
       new  9dd87ad   ENH while -> for loop
       new  62893e3   ENH reduce number of parameters
       new  70226d0   FIX min_weight_fraction_split with random splitter
       new  6cd9333   FIX min_weight_leaf in best sparse splitter
       new  306924b   ENH remove spurious code
       new  cb9b741   cosmit
       new  c6af5c6   ENH adaboost should accept c and fortran array
       new  cb11511   COSMIT simplify function call
       new  1c26cec   ENH expand ternary operator
       new  3047cd6   Revert previous version
       new  38183c8   ENH move utils near its use
       new  06701fc   ENH add a benchmark script for sparse input data
       new  9f3f5bb   Extract non zero value extraction constant
       new  24281b1   Lower number of trees
       new  c31d565   wip benchmark
       new  bf98916   Temporarily allows to set algorithm switching through an environment variable
       new  2838a14   Benchmark: Add more estimators + uncomment text
       new  8b3b071   FIX duplicate type coercision + DOC fix inversion between csc and csr
       new  4f423d6   Remove constant print
       new  2fe9da7   DOC Add note about sample weights in export_graphviz.
       new  87dc30b   Merge pull request #3829 from amueller/feature_selection_from_model_doc
       new  9033865   FIX raise error when number of features changes in partial fit, Add common test for inconsisten number of features for partial fit.
       new  248a503   Fixed issue 3815. Discrete AdaBoostClassifier now fails early if the base classifier if worse than random.
       new  52472bc   unused variables are redefined later
       new  cf96bba   FeatureUnion example using a heterogeneous datasource.
       new  6e7b11e   DOC RidgeCV more explicit documentation of CV parameter, taken from GridSearchCV.
       new  1100616   Merge pull request #3571 from jnothman/fix-link-embedding
       new  bdd7d67   MAINT CI: collect rackspace DNS info
       new  ff0c6d5   MAINT try to avoid HTTP timeouts to rackspace
       new  158d61f   Update about page
       new  c1a429e   Remove unused distance calculation from clustering example.
       new  a642428   DOC link to NLTK website
       new  54f483a   FIX #2986: ZeroDivisionError in LinearRegression on sparse data
       new  c84260b   DOC: add funding from Paris-Saclay Centre for Data Science
       new  6efa1cd   DOC FIX: typo and minor update
       new  c706386   Merge pull request #3854 from justmarkham/doc-changes
       new  f7cea2b   Merge pull request #3833 from amueller/sgd_partial_fit_variable_length
       new  429d064   Merge pull request #3853 from ogrisel/backport-lsqr-fix
       new  651453e   Added missing GSOC student
       new  ab57964   COSMIT add Base prefix to DenseSplitter and DenseSplitter
       new  ee53f51   TST stronger non-regression test for #3815
       new  2fe50a2   Merge branch 'pr/3837'
       new  a699394   FIX make assert_raises_regex backport for 2.6 consistent with 2.7+
       new  a704528   Merge pull request #3861 from ogrisel/fix-python-2.6-test
       new  fe553b7   ENH use specific warning class for RP
       new  9e00238   Raise error when sparse matrix is supplied to predict in KNeighborsClassifier.
       new  6e904c8   gen_rst informs file with missing documentation
       new  68aa42b   fix tab error
       new  598b007   Merge pull request #3862 from Titan-C/doc_comment
       new  a9de284   Merge pull request #3852 from lesteve/add-cds-funding
       new  a65ff57   Merge pull request #3860 from ogrisel/custom-rp-warning
       new  0ce8886   DPC FIX latex error in multiclass hinge loss
       new  076c822   COSMIT: minor cleanup
       new  6137579   Small typo.
       new  24cbf1d   Merge pull request #3857 from MechCoder/test_sparse_knc
       new  65d4cc1   Merge pull request #3867 from snuderl/patch-1
       new  a049ada   DOC: Typo in dict_faces.py
       new  f0fe4af   Merge pull request #2949 from FlorianWilhelm/theilsen
       new  167e96e   MAINT remove multi-output support from meadian absolute error
       new  2504cd6   Merge pull request #3868 from MechCoder/typo_example_faces
       new  60cadce   Fix RFE n_features minimum value #3812
       new  8142664   DOC: Added whats_new for TheilSenRegressor
       new  a030734   Accelerate AffinityPropagation.
       new  8621f00   Merge pull request #3173 from arjoly/sparse-tree
       new  968ba9b   Merge pull request #3870 from FlorianWilhelm/whats_new_theilsen
       new  31d5f0d   DOC update whats_new.rst (sparse data for trees)
       new  c095dc8   Merge pull request #3826 from amueller/kde_y_none
       new  67fed69   Add random forests to document classification example.
       new  8a8ead3   Minor typo in plot_dict_face_patches.py
       new  eb660ed   FIX: Raise error when patch width/height is greater than image width/height
       new  aaf86bd   Update documentation.rst
       new  89929d8   AffinityPropagation: save memory by reusing tmp.
       new  8e66030   added n_init check for k_means()
       new  7cee54d   Merge pull request #3871 from anntzer/fast-affinity-propagation
       new  9a36668   DOC fix what's new for sparse trees
       new  dcd21f3   Merge pull request #3858 from banilo/km_init
       new  f5e7cb1   MAINT bump joblib to 0.8.4
       new  7d2caa2   DOC add documentation improvements to what's new
       new  443e590   DOC update whats_new.rst for AffinityPropagation optim
       new  2fecf6f   Fix RFE n_features minimum value #3812 and make it simpler.
       new  1d6566b   Merge pull request #3690 from queqichao/fix_bug_in_cross_validation_when_using_sparse_matrix_for_fit_params
       new  d0eb09b   FIX clip SGD gradient on linear models for stability
       new  75afb5d   WEB fix css for new version of sphinx.
       new  a605bb9   Use plot_galler=0 in makefile for sphinx >1.2.3 Also works with 1.2.3
       new  373e641   MAINT refactor gradient boosting code
       new  37f1195   TST update numerical overflow (non-regression) tests
       new  4a8c7a7   Fixes #3644
       new  e4f62b1   Added NRT for return_path propagation check
       new  1945aec   DOC: Fix LabelBinarizer docstring to rst format bug
       new  f5e0ea0   Merge pull request #3883 from ogrisel/sgd-stability
       new  b8fc0b9   DOC update whats_new.rst for SGD stability
       new  c95703a   FIX unlikely pickling error of splitters
       new  089d2f7   TST add a toy test for min_weight_samples_leaf
       new  e7bab89   Merge pull request #3823 from ragv/omp-3644
       new  b15df0d   Typo
       new  00f1bb7   Update plot_outlier_detection.py
       new  18de735   Merge pull request #3889 from jdcaballero/master
       new  ff8e1de   DOC: COSMIT: Naive Bayes long url + PEP8 cleanup
       new  b356338   Merge pull request #3892 from mrshu/mrshu/naive-bayes-small-fixes
       new  e00fd83   DOC narrative docs for grid search's robustness to failure
       new  7a59791   DOC fix layout: ensure stable width of documentwrapper
       new  1852ba9   Fix BallTree and KDTree Docs
       new  84f1134   DOC target URLs on SGD loss / penalty figures
       new  81ac899   ENH add coverage multilabel ranking metric
       new  02938a7   Remove copy paste mistake
       new  9311f12   DOC improve documentation of coverage
       new  b1ef356   TST make separate tests for coverage + remove redundant tests with commons
       new  ee8500f   DOC clarify how ties are broken for coverage_error
       new  acddda5   ENH add sample_weight support
       new  20850c2   DOC typo
       new  c698834   TST + FIX ensure that it fails if sample_weight has not proper length
       new  b351548   DOC more explicit title
       new  24b6e1e   Update what's new
       new  226fe51   Merge pull request #3882 from MechCoder/doc_tree
       new  5474983   If step is explicitly zero or negative, raise a ValueError exception.
       new  da8720b   Merge pull request #3881 from amueller/code_tag_new_sphinx
       new  23f314a   Merge pull request #3893 from jnothman/css_fix
       new  4ee7360   DOC add missing load_svmlight_files to api references
       new  e6ed1e3   MAINT enable verbose output to track random failure on appveyor
       new  fd791ca   ENH Release gil in feature importance
       new  c24ada0   ENH use threading backend for features importance parallelisation
       new  c1894b5   FIX explicit initialization of normalizer
       new  e52b763   Don’t embed hyperlinks during latex file generation.
       new  4f1c381   Merge pull request #3900 from dimazest/latexpdf
       new  0a8ede1   Minor doc enhancement: documented setup develop.
       new  8537b21   Fix example in plot_pca_3d.py (color array had wrong size)
       new  956018a   Added link to the setuptools doc and a note on the need to rebuild every time a compiled extension is modified
       new  867af08   Merge pull request #3578 from jnothman/pls-validation
       new  92f0d76   Merge pull request #3905 from jlopezpena/fix-example-plot_pca_3d
       new  84d53c6   FIX handle selection in feature names for example
       new  d96f196   DOC Fix broken class references
       new  744614d   Check if targets is a numpy array and convert it into one if it isn't
       new  a07841b   Merge pull request #3902 from JeanKossaifi/doc_enhancement
       new  daaec8a   reverted css to current stable, fixes the display bug
       new  8b414de   Merge pull request #3898 from arjoly/nogil-feature-importance
       new  ff2b244   Merge pull request #3899 from jlopezpena/fix-OvOlist
       new  60e3577   DOC "RF" -> "Random forest" in text classif example
       new  3f82c20   Merge pull request #3915 from ThomasDelteil/fix_css_examples
       new  504d222   Replace 'coefs' by 'coef_' in PLSRegression
       new  7de192b   Merge pull request #3920 from jlopezpena/pls-coef
       new  7c50e6c   various docstring fixes for web docs
       new  93bab70   ENH do not use plt.cm.jet in JL bound example
       new  969e677   DOC fix kernels documentation inconsistencies
       new  86d4f04   Fix tinyclues logo in doc/about.rst
       new  0fd4c14   Merge pull request #3923 from Lothiraldan/patch-1
       new  afb22bf   Merge pull request #3921 from trevorstephens/docstring-format-fixes
       new  2c0a6db   gen_rst to use css tooltip instruction
       new  ed2b556   remove gallery css chunk from theme css
       new  e079f38   Remove JS from examples gallery
       new  c9041e1   Examples gallery css file
       new  cc0a34a   fix gen_rst file extension
       new  935fa24   add crossvalidation
       new  65d2e0d   DOC add missing import to plot_underfitting_overfitting.py, fix plot height.
       new  21d3bc2   Merge branch 'pr/3914'
       new  47257c2   Update what's new
       new  a5ecd0b   FIX copy paste mistake
       new  cea3dc8   Merge pull request #3925 from Titan-C/cssbox
       new  c828f03   Merge pull request #3869 from borjaayerdi/rfe_n_features_to_remove
       new  5489c84   DOC more expansive description of well-established algorithms
       new  8dd7f16   adding support for class_weight in fit method
       new  6a98856   DOC typo
       new  7e1a0ee   ENH LinearSVR using liblinear code
       new  b3717d2   what_new entry for KD and BallTree doc fixes
       new  afe890b   adding warning if passing class_weight through fit
       new  f75a140   Fixed error message typo
       new  5f0278c   Merge pull request #3936 from sethdandridge/typofix
       new  b27759b   FIX random state in isomap test
       new  ea1d134   MAINT get rid of undefined variable warning in Liblinear
       new  f35116a   Allow list of strings for type_filter in all_estimators.
       new  1496804   MAINT trailing whitespace in linear.cpp
       new  63c378f   Let nr_fold = l when nr_fold > l in CV
       new  7ede0df   MAINT liblinear class label logic (disabled)
       new  f690b4a   COSMIT tabs/spaces in Liblinear
       new  f1916c8   MAINT variable initialization in Liblinear
       new  60860e4   MAINT some cleanup in Liblinear
       new  9e1ea63   MAINT use size_t in Liblinear
       new  a646615   MAINT changes to (disabled) Liblinear code, from upstream
       new  165f949   Merge pull request #3934 from amueller/common_test_slight_cleanup
       new  57ea0c0   Merge branch 'update-liblinear'
       new  b99824f   changing warn message, making it Deprecation Warning, and removing negative index for py2.7
       new  c1a3688   ENH: Birch, first commit
       new  f19aee9   add example for birch + make it more pythonic
       new  0ed517f   Made the following changes 1. Improved documentation. 2. Added sanity test. 3. Moved example to a separate folder 4. Optimised compuation of centroids. 5. Added underscore to public attributes.
       new  d5a654d   ENH: Made the following changes 1. Added partial_fit and predict methods 2. Inherit from ClusterMixin from fit_predict methods 3. Removed inplace modification of X and self.copy 4. Added final global clustering and n_clusters param.
       new  d3f4361   OPT: Set check_X_y=False internally in birch
       new  923fb2e   OPT: Increased the code speed by doing the following. 1. Prevent recomputing centroids. 2. Precompute the squared_norms as much as possible, and remove call to pairwise_distances_argmin_min 3. Cached centroids_ and squared_norms_ to prevent repeated iteration across subclusters_ 4. Compute the closest_threshold when needed. 5. Remove enumeration of X 6. In plase modification of dot products.
       new  2b3296b   OPT: Made the following changes. 1. Used np.unravel_index 2. Improved docs and removed unused imports.
       new  dcfdf6c   ENH: Added example to compare birch and minibatchkmeans
       new  bb629cb   ENH: Add arbitrary clusterer in the global clustering step and test
       new  46f2951   ENH: Sparse matrix support
       new  ab68f07   TST: Add tests for branching_factor
       new  7ed6c5f   OPT: Remove unwanted calls to np.asarray for self.centroids_
       new  f4afc7e   ENH: Add narrative documentation
       new  08d3dfe   STY: Cosmits in documentation
       new  64a4a11   Made the following changes 1. Inherit from BaseEstimator to enable pretty printing 2. Add transform function
       new  a4e7ff3   DOC: Update documentation of _CFNode
       new  abc7808   Made the following changes 1. Fix condition for checking threshold. 2. Fix the global clustering by remapping the samples to the    subcluster_centers. 3. Add tests for threshold. 4. Minor doc and cosmits.
       new  42d0dc2   Made the following changes. 1. Prevent recomputation of updating centroids. 2. Use algorithm='brute' so that sparse data can be handled.
       new  000c9b1   Made the following changes 1. Remove call to KNeighborsClassifier 2. Set default branching_factor to 50
       new  2330834   Prevent preprocssing for iterating over dense data
       new  129f371   DOC: Modification to the documentation of threshold
       new  b01e21e   Made the following changes. 1. Update doc and whatsnew.rst 2. Fix coloring
       new  c5440bf   Made the following changes: 1. Make clear that partial_fit(None) enables global clustering. 2. Add compute_labels argument.
       new  767bf26   API: Rename n_clusters to global_clusters
       new  f12fc13   Changes to narrative documentation
       new  30ebaa8   Major changes 1. Refactored node splitting into a private function called _split_node. 2. Added copy argument to copy the data once and for all. 3. Removed n_samples_ attribute in CFNode. 4. Removed X and added linear_sum as an argument to the constructor of _CFSubcluster.
       new  54e7f88   Rename global_clusters to n_clusters again
       new  815844d   FIX: Remove trailing attributes in __init__
       new  13ab4d5   Explicit error when n_dim changes during partial_fit
       new  3810578   Rewrite model if fit is being called after partial_fit or vice versa
       new  a894b2e   Update whats_new.rst for n_iter_ attribute
       new  7e40e59   DOC added note to warn users not to __init__ params with trailing _
       new  2cfdc1c   Merge pull request #3940 from ragv/doc_own_estimators
       new  ce6a01d   Tuple unpack farthest_dist
       new  8985769   DOC clarify RandomTreesEmbedding docstring
       new  081a554   FIX P/R/F metrics and scorers are now for binary problems only by default
       new  56ee99c   Merge pull request #2679 from jnothman/prf_average_explicit
       new  b9474d0   Made the following changes 1. Made radius a property. 2. Added test for compute_label. 3. Minor changes to doc and split_subcluster. 4. n_cluster -> clusterer.
       new  f86622d   MAINT: Moved compute_label test to common tests
       new  4ebdb35   MRG: Fix FeatureAgglomeration docs
       new  5a019b7   adding simplefilter for warning
       new  35e5411   facecolor/edgecolor are passed arguments to savefig if they differ from rcParam['figure.facecolor'] From nilearn
       new  9392445   DOC explain t0 in the docstring (currently somewhat confusing)
       new  bf1691b   'none' is an acceptable value for penalty
       new  9d8f3f3   Merge pull request #3955 from hammer/patch-1
       new  ae61464   Merge pull request #3954 from amueller/fix_sgd_learningrate_docs
       new  6dab7c5   Merge pull request #3802 from MechCoder/birch
       new  de35bea   add support for class_weights
       new  6cb191c   changing warning test to use assert_raises_message
       new  b8b68e3   COSMIT: Make how to use partial_fit render properly
       new  e6a464b   ENH: Use check_X_y in pairwise_distances_argmin
       new  f2c200e   Merge pull request #3962 from MechCoder/check_X_y
       new  69d393a   Merge pull request #3931 from dsullivan7/cwwarn
       new  48cbbab   CLN: Fix typo in comment
       new  2ceed3b   Merge pull request #3964 from bwignall/boostdoctypo
       new  07a94c0   FIX: ward_tree now returns children in the same order for both structured and unstructured versions
       new  dfa5567   Add forgotten newline
       new  cca417a   Cosmetic change in test
       new  b131ad2   add multioutput support & bootstrap auto mode
       new  4c77ae8   expanded class_weight dimension fix
       new  ad7e772   Fixed a bug in RFECV when step != 1
       new  6039b18   Merge pull request #3824 from nmayorov/rfecv_bugfix
       new  5da3428   Merge pull request #3953 from Titan-C/savefigarg
       new  a295697   FIX: Raise AttributeError for fit_predict
       new  eabd2f5   ENH: Add metric support to neighbors_graph
       new  0f3cf10   Replace feature with sample
       new  eb6fcb2   Remove optional input validation from pairwise_metrics
       new  f346484   Merge pull request #3946 from MechCoder/doc_fix
       new  1e8bdb8   FIX bug in fit_params assumed to be an array
       new  780fb23   DOC: make LinearSVR appear in the doc reference
       new  7dcd08b   DOC fix/expand Imputer example docs
       new  ecc19dc   Minor changes: float conversion and rename rnd into rng
       new  97e88e1   Validation of params passed to neighbors_graphs
       new  1e0b653   Merge remote-tracking branch 'upstream/master' into wtree_children
       new  b2035c6   Merge pull request #3969 from MechCoder/remove_check_X_y
       new  0ecfc0b   ENH FIX Raise ValueError when the target does not match with classes arg
       new  c81c679   TST Added NRT to test if ValueError is raised during target mismatch.
       new  de6bfab   Merge pull request #3968 from MechCoder/metric_neighbors_Graph
       new  15157f9   Merge pull request #3966 from mvdoc/wtree_children
       new  2a1d214   ward_tree returns heights (distances) between the clusters both for the structured and unstructured version. Useful for plotting the dendrogram with scipy.
       new  65335fc   ENH: `ward_tree` returns distances
       new  db04340   changed xrange to range for compatibility with Python 3
       new  83efd8a   TST, test ward_tree_distance on known dataset
       new  a57b510   FIX: structured and unstructured ward_tree return children sorted in the same way
       new  428fc94   Cosmetic change in test_hierarchical, better documentation
       new  9b2e0c7   Add ENH description in whats_new.rst
       new  5166f05   Add better description of distances, cosmetic changes
       new  73c7453   Fix git rebase problems
       new  aa04eba   Add forgotten IFs
       new  756e000   ENH: Allow connectivity to be a callable in AgglomerativeClutering et.al
       new  b341a32   Remove support for int and update whatsnew.rst
       new  8e0dc6b   Refactored LDA module (first working version).
       new  4e1c63b   Added LabelEncoder and updated most of the docstrings
       new  cadb488   Updated more docstrings
       new  981fbe0   Return self in fit()
       new  f130922   Use equal priors by default
       new  7d81299   Added simple LDA example
       new  a98e0d3   Moved some common code into helper function
       new  27f696d   Support manually set shrinkage parameter (alpha)
       new  7216260   For now only the svd solver supports the transform method
       new  d01e9a2   Updated LDA unit tests
       new  d239fc4   Simplified LDA vs. shrinkage LDA example
       new  c297502   Removed magic number
       new  b9601cc   Addressed @mblondel's comments
       new  21fbb53   Updated parameter name in help text
       new  6296194   Parameters are now consistently named; also refactored the _means_cov() function
       new  4c63712   Added different shrinkage values to test cases
       new  125fa22   Updated LDA example and removed decision plane comparison
       new  057cf64   Increased coverage
       new  c93485f   PEP8 style conventions
       new  350f973   Catch NotImplementedError
       new  e4f4f6d   Fixed doctest
       new  c6d595f   Use assert_raises() correctly
       new  4140249   Removed private _decision_function()
       new  b787f80   Revert last commit because it broke some tests
       new  fa21b5c   Assert that the coefficients of the solvers are approximately the same
       new  1ff5199   Replaced str with six.string_types
       new  4dede9a   Raise error if alpha is not the expected type
       new  84a41f5   Cleaned up docstrings.
       new  4ba7c3a   Cleaned up one more docstring.
       new  76963c0   Addressed some comments by @agramfort.
       new  764efbe   Changed format of shape parameter in docstrings.
       new  84b1008   Updated test.
       new  685c627   Use broadcasting to scale back instead of multiplying with a diagonal matrix.
       new  96e78e1   Added random seed.
       new  97ab90b   By default, estimate priors from sample.
       new  5deb80f   Fixed docstrings.
       new  a5ae150   rename alpha to shrinkage + docstring cosmits
       new  8f47351   use random state in test (no global seed) + cosmit
       new  5ca36c8   Updated example because store_covariance is now set in __init__() and not in fit().
       new  9d9eb91   Changed parametrization of coef_ and intercept_.
       new  7c1101d   Treat the binary case differently from the case when n_classes > 2.
       new  1e469b2   Introduced step size to speed up the example.
       new  7a5563d   (Hopefully) fixed Travis bug on Python 2.6.
       new  f2fcafe   Another try to fix the Travis bug.
       new  7e4356b   Inherit from LinearClassifierMixin.
       new  5cb217d   Added shrinkage LDA entry.
       new  aede3d6   Rename parameter value 'ledoit_wolf' to 'auto'.
       new  8c016e8   Test that lsqr and eigen solvers return almost exactly the same coefficients.
       new  70773bc   Added deprecation warnings to store_covariance and tol parameters in fit() method.
       new  a88ad9e   Added sentence on lsqr vs. eigen solvers.
       new  095c7a9   Now correctly use weighted average to compute the class covariance. Cleaned up code and added references.
       new  936acf7   Works with one feature.
       new  d8024bb   Updated LDA tests.
       new  71a23ef   Fix problem with np.linalg.norm for version <= 1.6.
       new  081418f   Added documentation for the three solvers.
       new  b23a55b   Updated documentation on shrinkage.
       new  4937c5c   MAINT: fix silhouette typo
       new  ae7b1a4   Merge pull request #3974 from lesteve/fix-silhouette-typo
       new  73283e9   MAINT remove probability kwarg from SVR and NuSVR
       new  12ba788   Merge branch 'pr/3944'
       new  5b90e97   TST remove SVR(probability=True) from AdaBoost tests
       new  f68b8bc   Clean code
       new  2b56343   Use check_array and remove shape check
       new  5db4a34   Clean code
       new  6db5754   DOC: Improvements to children_ attribute in AgglomerativeClustering
       new  4a02469   Merge pull request #3158 from mvdoc/clustering
       new  389a43f   Fix typo
       new  1dad4d1   Merge pull request #3979 from lmichelbacher/patch-1
       new  d4d181f   Borrow a bit from scipy definition
       new  40c03d4   COSMIT pep8 fixes to sklearn/tests
       new  40140aa   ENH: Implemented LSH forest
       new  8968b8c   TST tests for LSHForest
       new  ebd639f   ENH Added radius_neighbors method to LSHForest.
       new  d10025b   MAINT refactor LSH forest code
       new  5ba5bf7   Updated docstrings and tests. in fit, kneighbors and radius methods, None type is not checked because it is not an expected input.
       new  0a80e1b   Used bisect in _find_matching_indices method.
       new  2539e2f   Removed cached hash and replaced with numpy packbits.
       new  d69d036   Updated lsh_forest and test_lsh_forest.
       new  cd06396   Updated docstrings for tests.
       new  8098264   DOC Added example for LSHForest.
       new  b6a6439   FIX Updated tests. renamed lsh_forest to approximate.
       new  c096d64   DOC Updated the example. Updated the compute distance function to use row_norms.
       new  59b5e39   COSMIT pep8 LSH code
       new  d20e54e   DOC Updated the document of neighbors module.
       new  77333c5   FIX Completed tests for missed lines.
       new  ea3b8fa   DOC editing for LSH forest narrative documentation
       new  b862d3e   FIX Updated examples
       new  3a752d4   Removed redundant ran_seed from GaussianRandomProjection Initialization.
       new  f245815   Updated Approximate nearest neighbors in neighbors.rst
       new  cdafbb7   Inlined generate_hash into create_tree in LSHForest.
       new  179ec4e   Updated test_approximate.
       new  56e47da   Re-written Approximate nearest neighbors documentation.
       new  f7f7e95   FIX Moved random_state to fit method to pass seed to create_tree.
       new  c7827bc   FIX Updated plots in example.
       new  8981251   DOC Included information about distance measure and LSH method in the documentation.
       new  4dc7cee   Added test to check randomness of hash functions.
       new  8687547   FIX Added integer as random_state to test_hash_functions.
       new  009ce30   Updated example to avoid redundants.
       new  f3f9b9b   Example splitted into hyperparameters and scalability.
       new  89b5cae   FIX Fixed accuracy error in scalability example.
       new  ed4feb7   DOC Added new section-Mathematical description of LSH.
       new  b018524   Created benchmark for approximate nearest neighbors.
       new  f246fb6   Added scalability plots into documentation.
       new  4aa8e28   FIX Modified tests to handle cosine distances instead or euclidean.
       new  f447a9c   Updated example with cosine measures.
       new  6cb40af   Modified plots in hyper-parameters and scalability plots.
       new  6368349   FIX Added user warnings for insufficient candidates.
       new  bf4b7ec   Changed parameter descriptions of LSHForest.
       new  197699b   FIX LSH: Fill up candidates uniformly from remaining indices
       new  1dea55c   Added KNeighborsMixin, RadiusNeighborsMixin.
       new  ed802bb   Added tests for graph methods.
       new  bd0c8d3   ENH improve LSHF scalability example
       new  53e04ed   FIX Fixed benchmark.
       new  e14a317   ENH Added error bars to scalability example plots.
       new  a43b91e   FIX Fixed ordering of returned neighbors.
       new  fa70b57   DOC: explain the importance of iid index and queries in bench script
       new  f4cd6a1   ENH: ensure that max hash size and hash dtype are consistent
       new  e5cd707   DOC Modified documentation of approximate neighbors.
       new  23efc3b   FIX Fixed example.
       new  c4de952   ENH Modified hyperparameters example to calculate stadard deviations of accuracies.
       new  76454fe   ENH Modified approximate neighbors doc string.
       new  5556f86   ENH Changed doc string and comments in test_distance function.
       new  a8fc522   ENH more LSH scalability example and doc improvements
       new  7de8e3f   ENH Used GaussianRandomProjectionHash for handling hashing.
       new  0ebfff2   ENH Changed description of hyper-parameters example.
       new  4a2fecb   ENH Inlined _create_tree method.
       new  b3f93ca   DOC fix LSH forest docstrings
       new  639c879   ENH Modified kneighbors and radius_neighbors methods.
       new  02f26eb   FIX Fixed partial_fit.
       new  31dbad5   TST improve tests for approximate neighbors.
       new  32491b0   FIX Modified assert_warns_message to ignore order of warnings.
       new  8181fe9   FIX LSH benchmark: joblib import, hardcoded /tmp
       new  1081f81   Merge pull request #3980 from larsmans/lsh-forest
       new  0597c23   TST/FIX ensure correct ducktyping for metaestimators
       new  0374abf   TST extend ducktype testing to handle #2853 case
       new  94f37f4   FIX ducktyping for meta-estimators
       new  50501d5   TST more rigorous testing of delegation
       new  7ac19ac   remove _iff_attr_has_method function
       new  86f4ee1   don't give names to decorators
       new  c6b2049   DOC: what's new: LSHForest
       new  3247803   ENH: Return distances option for linkage_trees
       new  7a38062   FIX backward compatibility. force_arrays was never in a release, allow_nd was a legal option to ``train_test_split`` that raises an error in master.
       new  acc25a4   Merge return_distance tests of both linkage and ward tree
       new  1aa5a9f   Merge pull request #3523 from cle1109/slda
       new  3fdc5d5   Add documentation for return_distances option
       new  9845c04   make attribute decriptor have the right docstring.
       new  37f0e2d   add regression test for docstring delegation
       new  a1510a1   fix docstrings in sphinx
       new  cedaf1f   Merge pull request #3986 from amueller/remove_allow_nd_in_train_test_split
       new  41e2c10   rename decorator
       new  acb924f   Merge pull request #3952 from MechCoder/callable_connectivity
       new  201fb08   COSMIT: Removed XXX comment related to LIL matrices
       new  860f773   FIX: Fix bug in computing full tree when n_clusters is large
       new  ad514d6   DOC: Add note to whatsnew.rst
       new  4f0d2aa   Merge pull request #3978 from MechCoder/compute_full_tree_bug
       new  caf27db   Small typo fix in the comments: "whther" --> "whether"
       new  cbf6c7e   Merge pull request #3989 from hammer/patch-2
       new  8a197e1   Merge pull request #3976 from MechCoder/agglomerative_playground2
       new  bf62754   TST: Test ground truth in linkage trees
       new  e736d76   ENH support sparse matrices in LSHForest
       new  c53ff7c   ENH vectorise _find_longest_prefix_match
       new  2d6f1c3   MAINT remove deprecated auc_score function
       new  ed5c63f   TST LSHForest benchmark script uses vectorised queries and fixed random_state
       new  2182262   Merge pull request #3985 from MechCoder/return_distances
       new  4f2d27b   FIX not all scipy.sparse support rand with random_state
       new  91c2386   STY cleaning some LSHForest query code
       new  085d677   add class_weight to trees, expand tests & minor refactor
       new  d5c57a7   COSMIT flat > nested / avoid duplicate validation
       new  3056ec1   TEST A duplicate minimum value should not yield non-finite predictions in IsotonicRegression
       new  983e574   FIX Adding eps to minimum value in clipping in IsotonicRegression
       new  9cfb506   STY remove unnecessary variable/operation
       new  14fe83c   Fix typo in SVM docstrings
       new  6a9a43b   Merge pull request #3997 from pricingassistant/doc-typo-fix
       new  4d8cbba   STY remove unnecessary variable/operation
       new  8f95410   ENH sparse metric support for paired distances
       new  d255866   DOC Add inline comment with reference to scipy issue
       new  636502f   Merge pull request #3995 from jmetzen/fix_isotonic
       new  d6a12c4   Fix a typo on svm.rst
       new  9341b05   Merge pull request #4000 from he7d3r/patch-1
       new  a3ee284   Fix typos, no shortenings in name, simplify example.
       new  a394654   COSMIT Minor changes to address review comments
       new  3633cb3   ENH vectorize DBSCAN implementation
       new  8a240f1   ENH sparse support in DBSCAN
       new  11569e3   ENH support weighted samples in DBSCAN
       new  684cfb3   FIX DBSCAN.fit_predict now also supports sample_weight
       new  39c3417   DOC/COSMIT clean up according to review
       new  0e0e545   DOC mention sparse matrix in LSHForest param docstrings
       new  56ca8f9   Merge pull request #3994 from jnothman/vec_dbscan
       new  3d1ede9   TST add tests for nd grid-search can train_test_split
       new  6d7d25f   MAINT dev version to follow PEP440
       new  79eabdd   FIX correct name due to rebase error
       new  164381c   Spelling in heading newgroups -> newsgroups
       new  d65d8c4   Merge pull request #4003 from ClashTheBunny/patch-1
       new  2690193   FIX: Bug in sparse coordinate solver in lazy centering
       new  87b4784   Merge pull request #4010 from MechCoder/sparse_bug
       new  f79ac7c   MAINT: Remove deprecation warnings in enet_path and lasso_path
       new  8b4ddc0   Another doc typo fix
       new  17dc6ab   Merge pull request #4012 from pricingassistant/doc-typo-fix
       new  743998e   Merge pull request #4011 from MechCoder/fix_deprecation_paths
       new  8e3f21d   Merge pull request #3982 from amueller/metaestimator_delegation
       new  1c28711   Fix a bug occurring when n_samples in a class < n_features Fix case n_sample in a class == 1
       new  a72da02   warning + log of prod -> sum of log
       new  17ea0db   warning -> error though we can check if reg_param > 0 and then use spherical covariance, let's keep things simple
       new  36932d7   tests
       new  94b2e94   ignore warnings, change docstring
       new  87edc22   BUG: only symmetrize matrix when it is not already symmetric
       new  7794d90   MAINT: create ensure_symmetric utility function to check matrix symmetry
       new  4bd1811   Merge pull request #4021 from jakevdp/fix_check_symmetric
       new  4735caa   DOC add note on scipy.sparse.rand use
       new  132ac81   MAINT more informative assertion failure in test_common
       new  e77aac8   Documentation of max depth parameter
       new  963fdce   Merge pull request #4026 from akshayah3/master
       new  a21957e   rename ensure_symmetric -> test_symmetric
       new  358e163   TST: add test of check_symmetric
       new  6444582   DOC: fix docstring of check_symmetric
       new  9825d17   MAINT make it possible to use wheelhouse-uploader
       new  4ef7213   add more informative doc & error message in utils.check_symmetric
       new  46c1c42   remove deprecated ``n_bootstraps`` parameter in Bootstrap cv object, verbose in FactorAnalysis and SelectorMixin. Also adjust some deprecation strings.
       new  4e270a5   remove deprecation tests.
       new  76a2a96   Merge pull request #4024 from jakevdp/ensure_symmetric
       new  2ce01e8   Merge pull request #4013 from amueller/deprecated_0.16
       new  67ca4ef   Merge pull request #3991 from jnothman/lshforest_improvements
       new  4c16329   Add parameters to pipeline methods, add some missing docstrings.
       new  a168b12   If scoring is provided, it MUST be used. Also, a scorer might not need predict.
       new  cc8c3f8   tighten tests and adjust for new logic.
       new  c779eff   add some nonregression tests.
       new  bec7957   add fixes to whatsnew
       new  6dce077   remove bad estimator test as it was enforcing too many restrictions, instead test that we don't enforce these restrictions.
       new  ed26cae   make gridsearch_no_predict test stricter.
       new  f774536   add another bug that I just discovered existed and was fixed in the PR to whatsnew.
       new  e2d163a   address @jnothman's comments
       new  a548303   fix docstring for pipeline score func.
       new  53f3b72   make utils.check_symmetric future-proof
       new  d5c72f3   Merge pull request #4038 from jakevdp/ensure_symmetric2
       new  de72e1d   Add dependence to max_features to docstring
       new  38920a3   add 'MultiLabelBinarizer' to __all__
       new  34b1438   Merge pull request #4044 from andreasvc/patch-1
       new  e98c513   Merge pull request #4030 from amueller/add_pipeline_docstrings
       new  2b24cce   parameter validation checks & tests for errors
       new  3a33776   FIX : avoid unecessary computation sparse dual gap
       new  439c195   Merge pull request #4047 from agramfort/faster_sparse_enet_dual_gap
       new  5b96533   Sparse inputs for fit_params
       new  91c03a4   Added helper function
       new  9b8fa52   Merge pull request #4031 from amueller/check_scoring_fixes
       new  7044728   Merge pull request #4049 from akshayah3/sparse
       new  8d52051   fix for issue #4051; replaced X = np.asarray(X) with check_array(...) method
       new  43e8c37   Merge pull request #4054 from tttthomasssss/bugfix-4051-affinity-prop
       new  24c0efd   PEP8 Fix E112 and E113 errors
       new  0ec0110   COSMIT Remove trailing : in comment to facilitate pep8 autoindentation
       new  8855598   PEP8 Fix E101, E111 errors and W191, W293, W293 and W391 warnings.
       new  3c1d09c   DOC DBSCAN doesn't "initialize centers"
       new  6e70a96   DOC Birch: spaces
       new  ac9ab96   Y-org rename & whats_new update
       new  cda6438   Merge pull request #4061 from ragv/pep8_cleanup
       new  9c7566e   Merge branch 'master' into rf-class_weight
       new  d67e9c4   Add newline before bullets
       new  69567ec   Merge pull request #4042 from lmichelbacher/patch-2
       new  da8f672   Say best_estimator_ depends on refit=True. Fixes #2976.
       new  cdae4a4   Merge pull request #4071 from amueller/grid_search_document_best_estimator
       new  cad87b5   rename vars & copy sample_weight
       new  b4ea76b   FIX: typo in Pipeline error
       new  40f44ff   MANIT: pep8ize to an extent cluster module
       new  033d5e7   MANIT: Revert to py3 print version
       new  77238f3   Merge pull request #4078 from pratapvardhan/misc-docs
       new  1666274   TST slight cleanup of common tests.
       new  db89a44   Merge pull request #4075 from banilo/pipe_typo
       new  8e5bd8e   COSMIT / PEP8 Limit line length to 79
       new  b3209c1   MAINT Add new function to check if an estimator is fitted
       new  abd31d2   MAINT Make uniform the error raised for not fitted condition
       new  4fe8d93   MAINT Remove the _check_fitted test.
       new  129e319   FIX the try except block to recognize NotFittedError
       new  0fce60d   TST Add tests to assert if NotFittedError is raised appropriately
       new  35938f2   FIX support_ parameter should be tested, not the coef_
       new  0bca2ff   FIX Refactor _is_fitted method into check_is_fitted.
       new  b8ffa95   FIX Make error message more explicit in fit_inverse_transform
       new  f914afa   Merge pull request #4029 from ragv/maint_raise_uniform_error_notfitted
       new  fa0ceee   ENH add a benchmark on mnist
       new  b0dba06   ENH Improve script output display
       new  cd573f8   DOC/MAINT Update whats new entry for NotFittedError
       new  c436fdd   Merge pull request #4086 from ragv/whatsnew
       new  87a823c   Fix a typo in OneHotEncoder's docstring
       new  57544ae   Merge pull request #4089 from wujiang/patch-1
       new  b541191   Merge branch 'master' into rf-class_weight
       new  35c2535   rename cw option to subsample & refactor its implementation
       new  527ecf5   Merge pull request #3961 from trevorstephens/rf-class_weight
       new  b7d9c3e   FIX: ensure that feature importances are properly scaled
       new  6e0b95f   ENH better RBF parameters heat map
       new  f62f749   FIX: Regression in NearestCentroids
       new  46f9fc2   BUG: Correct check for local import
       new  c579244   Merge pull request #3640 from untom/fix_consensus_score
       new  871026e   does not compute
       new  124bb8c   Update lightning information.
       new  454aee9   Merge pull request #4097 from amueller/remove_unneccessary_initialization
       new  5bb3cbf   DOC add performance for all available classifiers
       new  912d142   Merge pull request #4093 from MechCoder/regresssion_nc
       new  841a51d   Merge pull request #4095 from jjhelmus/bug_local_import
       new  d3ed841   make related packages more prominent
       new  9922fa9   Merge pull request #3562 from arjoly/bench-mnist
       new  c6def01   Update seqlearn information.
       new  c1f3080   Merge pull request #4058 from amueller/cleanup_common_tests
       new  2e2eebf   Make ParameterSampler sample without replacement if all parameters are given as lists.
       new  6e0b382   TYPO in model evaluation docs
       new  93ca609   TYPO in coverage_area docstring.
       new  a6f5fe5   DOC Modified math expressions to theoretical bounds.
       new  c970fb5   COSMIT spelling
       new  08a963c   Merge pull request #4105 from amueller/minor_spelling_fixes
       new  78233b2   FIX scale - Raise ValueError when input contains non finite values.
       new  731b50b   TST Add test to check if warning is raised for integer dtype-ed inputs.
       new  48e72b5   TST Add test to check if ValueError is raised if input contains NaN
       new  ab5dd15   OPTIM Lower memory usage of spectral_embedding.
       new  407aa33   Merge pull request #4018 from ragv/fix_for_nan_scale
       new  67d128a   FIX Written in Latex format.
       new  c150608   FIX Remove unused score_overrides_loss param from check_scoring
       new  b3fbccc   FIX various mismatch between docstring and signature params
       new  a413f87   DOC minor fixes do docstrings, don't document deprecated parameters.
       new  494b8e5   TST add test for sparse matrix handling in clustering
       new  b5b1c59   DOC Add silhouette analysis plot for KMeans
       new  f04266a   DOC ENH Fill the silhouette with the corresponding cluster color
       new  ae95a0c   DOC ENH Remove the vertical x = 0 line and plot avg silhouette score line
       new  bf203de   Merge pull request #4052 from amueller/clustering_sparse_matrix
       new  a9ea55f   FIX Use slinear interpolation for isotonic regression
       new  0c9377e   ENH use parallelism for all metrics in pairwise_{kernels,distances}
       new  da90a0f   Merge pull request #4085 from jnothman/pairwise_parallel
       new  7962412   Merge pull request #4111 from ragv/slinear_isotonic
       new  5a3b9e9   MAINT Remove temporary fix #3995 in view of the change to slinear.
       new  7b36bf6   Merge pull request #4113 from ragv/isotonic
       new  7070239   DOC: copy-edits and active voice
       new  e8a8f8f   ADD @mblondel's initial implementation of KernelRidge added
       new  17b5177   ADD Example comparing kernel ridge and support vector regression
       new  0db77c2   TEST Added tests for KernelRidge based on @mblondel's code in lightning
       new  d3eb311   DOC Added documentation for kernel_ridge module
       new  6c62f87   ENH Example comparing KRR and SVR documented and slightly modified
       new  946e490   FIX Fixed assigning colors to methods in KRR example
       new  c38e14a   DOC Added narrative doc for kernelized ridge regreesion
       new  4b8a1a7   DOC Resolved minor issues in documentation of kernel_ridge
       new  12e2ee3   MISC Optimized two numpy statements in kernel_ridge.py
       new  f1b4284   FIX Add kernel_ridge to list of all submodules
       new  ddaf190   FIX Input validation in fit() of KernelRidge
       new  e5888b2   FIX Allow multi-output in KernelRidge
       new  eaaf84d   REFACTOR Backported changes in _solve_cholesky_kernel to ridge.py
       new  d240877   MISC Using RandomState object instead of setting global seed
       new  60e3d04   DOC Revised documentation of kernel ridge regression
       new  4d68c53   DOC Polishing doc of kernel_ridge
       new  34d0b8d   PEP8 Removing PEP8 violations in plot_kernel_ridge_regression.py
       new  c48c971   MISC Let sample_weight default to None
       new  f3df83f   FIX We must not use overwrite_a=True in _solve_cholesky_kernel as K might be reused
       new  a67899f   TEST Extending tests for kernel_ridge
       new  ff1ec8d   REFACTOR Using lstsq rather than manual pinv in _solve_cholesky_kernel
       new  6011193   MINOR Fixed typo in test name (test_kernel_ridge_singular_kernel)
       new  a2c8d12   TEST Use make_regression instead of make_classification when testing kernel_ridge
       new  7223efd   DOC Added KernelRidge to whats_new.rst
       new  2a2e7c6   FIX KernelRidge checks if fitted before predict
       new  5b57784   Remove singular matrix warning.
       new  cb77255   fix import error in lda.py
       new  73f0d2b   Remove unused imports.
       new  4093790   Merge pull request #4117 from chyikwei/fix_import_err
       new  9046a38   DOC Added kernel_ridge to classes.rst
       new  aa11cc3   Merge pull request #4118 from jmetzen/kernel_ridge_doc
       new  6c07cd4   Merge pull request #3850 from amueller/randomized_search_no_replacement
       new  d9ecf46   DOC what's new on parallelism of pairwise_distances
       new  22f718d   ENH Added RobustScaler
       new  444bf22   DOC ENH Simplify the example code; Add plots for n_clusters = 3 and 5
       new  22f8e56   DOC Added what's new entry for silhouette plot example
       new  bf8cfb6   Check if parameters are fitted
       new  45982e0   Merge pull request #4076 from ragv/silhouette_plot
       new  ac7c88c   TST set random state in check_classifiers_train
       new  edc9bbb   remove duplicated test, probably caused by rebase issues.
       new  bdacaba   Kernelized ridge regression -> Kernel ridge regression
       new  8e3eea8   Explain why we are somewhat selective, lower citiation rule of thumb
       new  e20ea81   Fix error in CCA docstring.
       new  6ffc685   FIX Support for negative values of n_jobs
       new  a6723ed   ENH: cross-validate over predefined splits
       new  9d5c4c6   Merge pull request #4139 from MSusik/negative_njobs
       new  edb73cc   FIX check for integer 1 before checking for floating 1 (isinstance(1, numbers.real) == True)
       new  5c5fbf5   COSMIT pep8 cleanups
       new  736aeed   Merge pull request #3836 from untom/fixedsplit
       new  0c5b0f7   add PredefinedSplit to classes.rst and whats_new.rst
       new  24c789b   Added input validation refactoring from last sprint to whatsnew. Not padding my back, just keep forgetting if that was in 0.15.2 or not.
       new  f6c7252   DOC staged_* returns generators. Fixes #3831.
       new  9ca04c2   Fix for overflow warning
       new  21e63aa   ENH: optimize DBSCAN (~10% faster on my data)
       new  44c8083   Merge pull request #4151 from larsmans/dbscan-faster
       new  247ad47   Merge pull request #4153 from hsalamin/remove_exp_warning
       new  0a4c12f   TST remove tempfiles that we create in svmlight tests and agglomerative clustering tests
       new  fc51c06   FIX agglomerative clustering w/ connectivity constraints & precomputed distances
       new  500f032   Merge pull request #4148 from amueller/clean_up_after_tests
       new  17726d2   Merge pull request #4141 from amueller/bootstrap_int_one_fix
       new  6acbb8c   Handle sample_weight upfront.
       new  59648f3   Simplify rescaling.
       new  ab88c4c   Use relative imports.
       new  7c3a86b   Add what's new entry.
       new  bdcea50   Add t-SNE to whatsnew
       new  291447f   fix nb partial_fit w class_prior #3186
       new  0fe3f5c   Fix issue #4154
       new  7f42daa   DOC Correct what's new entry of Jaccard Similarity Score
       new  de92ada   Merge pull request #4158 from ragv/whatsnew
       new  95f85f2   Merge pull request #4159 from aflaxman/gp_docs
       new  f81a88c   fix rst table layout
       new  49da1dc   Merge pull request #4160 from trevorstephens/doc-fix-4098
       new  3023053   Merge pull request #4155 from trevorstephens/nb-fix-3186
       new  4ab4fba   G is upper-triangular, so G.T should be lower-triangular.
       new  50b83ae   Add test for MSE estimate bug.
       new  ca55e74   Merge pull request #4131 from amueller/faq_no_1000_citations
       new  fd9b03a   sort labels in precision_recall_fscore_support
       new  92e1e39   Merge pull request #4147 from amueller/precision_recall_unsorted_indices
       new  d9e40d2   Merge pull request #4156 from AlexanderFabisch/fix_tsne_one_comp
       new  e936dfa   search C not gamma of linear svm in tutorial
       new  cee2217   DOC typos and consistency in SVM docstrings
       new  25db1ac   Merge pull request #4171 from amueller/statistical_learning_gridsearch
       new  9ed1df0   FIX/ENH: Fix kneighbors_graph/kneighbors and allow X=None
       new  c94d0a8   FIX: Handle and test case where n_duplicates > n_neighbors
       new  72d20a3   Add docstring and test for kneighbors
       new  ec245c8   ENH/FIX: Patch radius_neighbors to follow kneighbors convention
       new  709b41a   TST: Refactor tests for k and radius neighbors
       new  cefc2f8   ENH: Add include_self param to graph variants for backward compat
       new  8cac52d   DOC: Add what's new entry
       new  04e0f32   MAINT threshold -> threshold_ with deprecation warning.
       new  bc27546   DOC: Make docs clearer
       new  2f889b1   Merge pull request #4046 from MechCoder/allow_X_none
       new  6471e0d   Merge pull request #4174 from ragv/fixes_4173
       new  8db11a8   Make nystroem approximation robust to singular kernel.
       new  b35b2fd   Add predict_log_proba method to GradientBoostingClassifier
       new  e3f9435   TST Silence some tests.
       new  13b6aec   non-inheriting estimator added for rfe, cross validation, and pipeline tests
       new  bfef274   5 of 6 are zeros, not ones.
       new  94157fa   Merge pull request #4188 from miniyou/patch-1
       new  c5fadff   add compute_sample_weight util
       new  f75c98b   refactor forests & trees class_weight calc
       new  4fcf20a   TST/FIX in future, average='binary' iff 2 labels in y one of which is pos_label
       new  f9ec807   MAINT create a new venv in travis for "ubuntu"
       new  67dcd99   TST fix tests with numpy 1.6.1
       new  ad83fc6   DOC fixes supported numpy version
       new  7090357   MAINT improve misleading comment
       new  cedf023   Move import statement
       new  bd9d16b   Simplify the code by using len
       new  21369dd   TST Removal or modification of stop_words_ should not affect transform.
       new  965bf51   use safe in1d, use compute_sample_weight in ridge
       new  eb52bb3   Merge pull request #4037 from ragv/make_storing_stop_words_optional
       new  54d68d2   remove inplace multiplication in ridge
       new  748c750   Fix rst formatting
       new  3e9beab   DOC: fix path
       new  92b9081   FIX Ensure at least 1 feature is sampled when max_features is a float
       new  92ed5ee   FIX ensure that negative float max_features will lead to an error
       new  b8bc07e   Merge commit 'b35b2fdad3a38e29bf5ef3534aeb9468b7f20f4a' from #4181
       new  cfd68ef   Merge pull request #4198 from arjoly/gbrt-fix-percent-max_features
       new  38104ff   Merge pull request #4193 from lesteve/deactivate-travis-default-venv
       new  87c9ba7   Corrected two bugs related to 'tied' covariance_type in mixture.GMM(), added test, closes #4036
       new  364fdd6   minor fixes, addressing @agramfort's comments.
       new  6d7cb33   Merge pull request #4109 from amueller/gmm_tied_covariance
       new  af15288   website: add skll to related projects
       new  31c5497   Merge pull request #4201 from amueller/related_projects_skll
       new  b6343b8   PyMC
       new  6f5ec47   DOC: fix correct shape of children and distance
       new  1d08f08   Merge pull request #4203 from gwulfs/rel_projs
       new  f2431db   add test for None class_weights
       new  97e3b96   PEP8 compliance
       new  420daac   Merge pull request #4190 from trevorstephens/refactor_cw
       new  a168a6c   STYLE cosmetics / unused import
       new  bc5b403   FIX #4059: explicit warning for strict selectors
       new  cd63acc   Add test that score takes y, fix KMeans, FIX pipeline compatibility of clustering algorithms!
       new  0668e70   Merge pull request #4186 from xbhsu/non-inheriting_estimator_tests
       new  620301b   Merge pull request #4064 from amueller/pipeline_scoring_y_none
       new  ba2c9a7   Merge pull request #4179 from amueller/nystroem_singular_kernel
       new  eaf1e8c   Explain why we multiply mean by two
       new  2ce9eac   TST more strict selectors with empty support
       new  95681ee   Merge pull request #4206 from ogrisel/fix-strict-select-fdr
       new  a9b3965   Better error messages in MeanShift, slightly more robust to bad binning.
       new  4629366   Merge pull request #4176 from amueller/mean_shift_no_centers
       new  75365a0   DOC: Fix is->if
       new  6f9ce0b   FIX precompute_gram->precompute as in #2224
       new  c4dc60b   Merge pull request #4221 from saketkc/fixes
       new  16d8675   MAINT enable travis notifications in gitter
       new  8bd7b42   MAINT enable appveyor notifications in gitter
       new  79144e4   ENH: improve GMM convergence check #4178
       new  bdecd1f   ENH: issue #4178 (cont.)
       new  371c2f6   FIX: fix doctest
       new  12a6d19   doc: mention GMM stopping criterion change
       new  0f6f48a   a few changes according to comments by @ogrisel
       new  a3283c6   FIX tests in gmm broken by previous rebase
       new  c2d2307   Added feature to fix #1523
       new  e20d1a4   ENH decision_function now returns #votes and scaled scores.
       new  dbc5d70   Change loss names for LinearSVC and LinearSVR()
       new  6465b8d   PEP8 Clean up minor PEP8 issues.
       new  5329590   Merge pull request #4229 from ragv/4135
       new  fca4494   Merge pull request #4222 from saketkc/benchmark_fixes
       new  5c0c5c6   MAINT use absolute imports in tests
       new  1be7400   Merge pull request #4232 from lesteve/use-absolute-imports-in-tests
       new  73e5cf5   Merge pull request #3933 from fabianp/loss_liblinear
       new  270d47a   Fix gibbs sampling behavior in RBM with integral random_state.
       new  e3e0827   FIX check (and enforce) that estimators can accept different dtypes.
       new  8ef0b9a   fixes in GMM, TSNE, MDS, LSHForest, exclude SpectralEmbedding
       new  5fb923a   FIX Return np.zeros(n_alphas) when the targets are uniform
       new  7926262   TST Add enet test for the case of uniform targets
       new  3ac2da6   FIX Use resolution instead of min from np.finfo(float)
       new  8df2549   TST Check for equality in alphas_ when uniform targets are used.
       new  0c33082   DOC minor improvement in Ensemble user guide
       new  2a946dd   Merge pull request #4226 from ragv/fix_4224
       new  b34e924   Update plot_rfe_with_cross_validation.py
       new  196e321   Merge pull request #4244 from ugurcaliskan/patch-1
       new  d09894c   Merge pull request #4233 from amueller/minor_doc_improvement_ensemble
       new  7a4ac25   FIX buggy idiom found by pyflakes
       new  9f50699   FIX regularized covariance on 1D data
       new  021bf74   Merge pull request #24 from ogrisel/fix-nan-1d-regularized-covariance
       new  cd931fa   Merge pull request #4136 from amueller/test_dtypes_all
       new  d16e9ee   FIX stricter validation for non-empty input data
       new  d0b9955   Merge pull request #4214 from ogrisel/fix-empty-input-data
       new  774eec5   Merge pull request #3891 from ragv/decision_function_ovo_1523
       new  5f95154   Merge pull request #4204 from mvdoc/master
       new  cd5166e   make check_array convert object to float.
       new  414cda9   FIX TypeError not ValueError in is_fitted
       new  d90a8ae   TST test last commit
       new  0373cc0   FIX broken test under 64 bit Python 2 / Windows
       new  4f3613f   Merge pull request #4251 from ogrisel/fix-shape-format
       new  2e414d7   Add example for precision_recall_fscore_support
       new  26df432   DOC: specifies the default number of eigenvectors
       new  f20ff86   Merge pull request #4265 from harrymvr/doc_spectral_clustering
       new  b649df6   Shrink the doc for precision_recall_fscore_support
       new  7b24628   DOC python 3 fix in plot_kmeans_silhouette_analysis.py
       new  3d86d1f   Merge pull request #4266 from lesteve/python3-fix-plot-kmeans-silhouette-analysis
       new  96d4b3e   more robust check for dtype object
       new  63039b2   Added multithreading support for kneighbors search
       new  1250974   Shrink precision_recall_fscore_support's doc again
       new  1a88eb0   DOC: Reword docstring and deprecation warning for include_self.
       new  e07d420   FIX: Fix n_components in AgglomerativeClustering and friends
       new  ecfc93d   ENH Add probability calibration based on isotonic regr. and Platt's sigmoid fit + calibration-curve
       new  6bf5d7e   ENH Brier-score loss metric for classifiers
       new  6ec7a0e   TST Tests for the calibration module
       new  9cc7018   DOC Examples for the calibration of predicted probabilities and calibration-curves
       new  c340f1f   DOC Narrative doc for the calibration module
       new  6b0f75c   COSMIT : pep8 + 2 spaces
       new  7376309   TST improve coverage of calibration.py
       new  eee0e67   MAINT mark _sigmoid_calibration private
       new  70d49de   ENH ensure that a warning is raised when sample_weight is not supported
       new  c181308   DOC update contributors for the calibration features
       new  a798d9e   TST Adding brier_score_loss to test_common.py
       new  5604884   Fixed typo in CONTRIBUTING.md about how to submit changes back to the project
       new  ad54f02   Merge pull request #4277 from Celeo/contributing-request-typo
       new  44d81a0   FIX: add except* in Splitter.init definitions
       new  f6af488   Merge pull request #4282 from glouppe/tree-fix-4281
       new  0c00f1a   ENH: better convergence check
       new  1b1ec71   DOC: what's new about GMMs
       new  1add39f   [Feature Selection] Fix SelectFDR thresholding bug (#2771)
       new  700babc   extensive (excessive?) testing of FDR
       new  95f1d95   FIX DOC TST: alpha is an uppper bound on FDR
       new  f12a441   raise noise in test again, be ascii
       new  0036dd0   ENH make defensive copies in GradientBoosting*.staged_decision function.
       new  0cf9314   Merge pull request #4165 from amueller/gbrt_staged_defensive_copies
       new  3468e00   Merge pull request #4057 from amueller/dtype_object_conversion
       new  92da7c0   Merge pull request #4146 from amueller/fdr_treshold_bug_2
       new  98f65c5   Edited _get_weights to make it robust to zero distances
       new  1cf3e06   special case if dist is object in _get_weights
       new  ef0bbfb   Avoid divison by zero warning in KNeighbors estimators
       new  0d20c9f   whatsnew entry for KNeighbors zero division in weights.
       new  5c893e5   Change the way the covariance is computed to avoid problem with not positively defined covariance matrices
       new  739efd6   Added test to check that obtained covariance matrices are positive definite after learning a GMM
       new  e0536d9   Cosmetic changes in the doc string. Added an indentation level
       new  3e069d5   Clarified a comment and factorized the tests for all covariance types
       new  db2edda   Converted test_positive_definite_covars to a generator.
       new  fa3ecbc   Converted docstring of 'test_positive_definite_covars' into a comment to avoid description ptoblems with nose (Related to issue #4250)
       new  86392db   Put the error state change about underflow errors in a 'with' statement in '_covar_mstep_full'.
       new  2c3e564   DOC cosmetics in test docstring
       new  10e099e   DOC document GMM fix in whats_new.rst
       new  a59eee8   FIX Add check_symmetric to __all__
       new  4c389c6   Merge pull request #4269 from MechCoder/deprecate_n_components
       new  06bde13   Update related_projects.rst
       new  4a0e281   DOC: make it clear 'thresh' should be removed in v0.18
       new  3f5277e   Merge pull request #4284 from hbredin/dgpmm_convergence
       new  c1c8cc1   FIX loss, penalty --> penalty, loss
       new  6d23a58   EXAMPLE/FIX l2 --> squared_hinge; L1/L2 --> l1/l2
       new  9df01a8   FIX/MAINT Deprecate and support upper case values for loss/penalty
       new  50876ec   COSMIT Flake8 fixes for svm/base.py
       new  354c35c   FIX Make LinearSVC/LinearSVR support the deprecated l1/L1, l2/L2
       new  6d222fd   TST test deprecation warning and support for uppercase
       new  a926626   Merge pull request #4261 from ragv/svm_scale_c_plot
       new  07ce147   Merge pull request #4092 from ogrisel/iris-rbf-params-heatmap-example
       new  8418647   Merge pull request #4293 from kyleabeauchamp/patch-1
       new  3c3360d   FIX Make Spectral Embedding deterministic
       new  2a7a213   DOC update whats_new for deterministic spectral_embedding
       new  feceab6   FIX test for consistent handling on empty input data
       new  8f2780f   Make BaggingClassifier use if_delegate_has_method in decision_function
       new  8cf714c   remove special case for BaggingEstimator after merge of #4137
       new  97d8341   TST make TheilSenRegressor run faster in common tests, I could swear I did that before...
       new  69827c4   TST Test check_consistent_length and TypeError with ensemble arg
       new  e073432   FIX issue #4268 (bug in BernoulliNB).
       new  c31710d   TST check that Bernoulli NB is consistent with Manning et al. IR book example
       new  bdbbb02   DOC whats_new.rst for BernoulliNB fix
       new  1655a04   TST check that constructor params are not mutated by fit
       new  a5a800e   Fix docstring of mahalanobis distance in empirical covariance. Closes #4168.
       new  4ed4257   FIX inconsistent boundary handling
       new  5141f18   TST add test for radius_neighbors boundary handling
       new  2981837   DOC whats_new.rst for radius_neighbors boundary handling
       new  4893ce0   Added missing space to exception message. Simplified (trivially) code and corrected message. Updated tests.
       new  1c7f6f4   Merge pull request #4308 from terrycojones/fix-missing-space
       new  644e2a4   MAINT change "l2" to "squared_hinge" in svm.bounds
       new  acb4aa5   Merge pull request #4310 from larsmans/svm-bounds-l2
       new  b15275c   explicitly mentioned include_self
       new  dcd2bbb   explicitly passed by name
       new  6b5e63b   Added Random Kitchen Sink reference
       new  e808a3f   Update plot_multilabel.py
       new  af1906f   Merge pull request #4315 from ugurcaliskan/patch-2
       new  8820701   skip OMPCV on travis by raising SkipTest in set_fast_parameters.
       new  9bc4de8   some fixes for sphinx and in examples
       new  be69297   don't use unicode in latex. That just complicates things too much. Also, who the hell uses unicode whitespace?
       new  1800cb7   use PNG images for latex compatibility
       new  4486bde   Merge pull request #4314 from vortex-ape/RBFSampler
       new  29fb207   Merge pull request #4318 from amueller/skip_omp_cv_on_travis
       new  f1b8283   catch some warnings, be less verbose in testing.
       new  63336fc   Merge pull request #4320 from amueller/minor_doc_fixes
       new  72d4d59   Generate example rst files for inclusion in docstrings.
       new  b2e34f7   Merge pull request #4323 from amueller/generate_empty_example_rsts
       new  9e20fb6   Merge pull request #4313 from vortex-ape/spectral_clustering
       new  8ab4649   FIX Flips sign of SVM's dual_coef_ in binary case
       new  dbe4971   Updates whats new and fixes the documentation
       new  6813528   Merge pull request #4192 from jnothman/binary_iff_binary
       new  ed5519f   Merge pull request #4307 from amueller/more_quite_testing
       new  e0f5049   minor fixes in docs.
       new  deeaf0a   change default to shuffle=True in SGDClassifier and friends.
       new  5e8a0f6   Merge pull request #3965 from amueller/sgd_classifier_shuffle
       new  e01a16d   Merge pull request #4325 from amueller/more_minor_doc_fixes
       new  663c236   Removes RBF calculation by hand
       new  7a97192   Minor fix for a comment
       new  dd5c2e7   TST ensure return type of radius_neighbors is object
       new  f7efb1a   Merge pull request #4326 from Barmaley-exe/svm-dual-coef-fix
       new  114062f   Explicit encoding for opened files in gen_rst.py
       new  f4fe0f8   Support Python 2 and 3
       new  20b4eec   Remove trailing spaces
       new  0e21f93   Merge pull request #4234 from amueller/rbm_random_state_gibbs
       new  60151c0   DOC fix broken image link
       new  d39fe22   DOC add some missing API links in whats_new.rst
       new  46470b4   Fix typo of DPGMM doc
       new  0dcd7d3   STYLE trailing spaces
       new  1ca8fec   MAINT fix typo pyagm -> pygamg in SkipTest
       new  6ce0b35   Merge pull request #4338 from lesteve/fix-typo-in-pyamg-skip-test-message
       new  2b1c31e   Adding unit test to cover ties/duplicate x values in Isotonic Regression re: issue #4184
       new  ab556be   Implement "secondary" tie strategy in isotonic.
       new  c1fa16f   ENH no need for tie breaking jitter in calibration
       new  23164e7   DOC remove unnecessary backticks in CONTRIBUTING.
       new  3d8114e   use a pipeline for pre-processing feature selection, as per best practise
       new  4cc4a5d   ENH improve docstrings and test for radius_neighbors models
       new  952c160   TST boundary handling in LSHForest.radius_neighbors
       new  9ce8322   Merge pull request #4317 from ogrisel/fix-radius-queries
       new  26cd806   Merge pull request #4345 from bendavies/20newsgroups_example
       new  651c85a   add scipy2013 tutorial links to presentations on website.
       new  cebe73f   Merge pull request #4339 from amueller/scipy2013_lecture_link
       new  e51b133   Merge pull request #4342 from sotte/fix_contributing_format
       new  6ff4e14   Update documentation of predict_proba in tree module
       new  49aa9d5   Update docstring predict_proba()
       new  8dbe3f8   Merge pull request #4028 from ogrisel/wheelhouse-uploader
       new  a2087d8   Do not shuffle in DBSCAN (warn if `random_state` is used).
       new  15c9c0f   FIX/TST boundary cases in dbscan (closes #4073)
       new  6c40cd2   DOC deprecate random_state for DBSCAN
       new  ec84a00   Merge pull request #4302 from amueller/isotonic_regression_duplicate_fixes
       new  2415100   Adding fix for issue #4297, isotonic infinite loop
       new  ff942f9   MAINT use canonical PEP-440 dev version consistently
       new  b74b004   ENH Scale the sum_of_confidences to (-0.5, 0.5)
       new  342e6e4   COSMIT remove trailing whitespaces
       new  16075fb   Fix rebase conflict
       new  555b859   Fix #4351. Rendering of docs in MinMaxScaler.
       new  4cc0235   Merge pull request #4352 from amueller/issue-4297-infinite-isotonic_bak
       new  09dc09a   0.16.X branching, version 0.16b1
       new  5e4809c   Move dev branch to 0.17
       new  c23d4f3   FIX #4358: more tolerant test_scorer_sample_weight for windows
       new  9e1377a   FIX revert wrong previous fix and make weighted scorer test deterministic
       new  ab4d07d   FIX #4358: make weighted scorer test deterministic
       new  c7a04f7   Related projects linking to wrong URL
       new  88e9f90   Merge pull request #4359 from josephlewis42/master
       new  85b9ce5   DOC Fix NMF inconsistency and broken links
       new  0d6d374   add appveyor badge to alert us of failures
       new  75473a2   Merge pull request #4363 from arimbr/patch-1
       new  de5235b   Merge pull request #4364 from amueller/appveyor_badge
       new  09b5388   Merge tag '0.16b1' into releases
       new  d8a623e   Merge branch 'releases' into dfsg
       new  07226a3   Merge branch 'dfsg' into debian
       new  a99f290   Merge remote-tracking branch 'origin/0.16.X' into debian
       new  ff736b4   updated patches
       new  dea22c8   mark autoreconf for cython ge 0.20.1
       new  93b1ace   Merge remote-tracking branch 'origin/0.16.X' into dfsg
       new  23ab36d   Merge branch 'dfsg' into debian
       new  f36647b   ENH Adding sample-weight support for Gaussian naive Bayes
       new  7abb313   TST Tests for sample-weights in Gaussian naive Bayes
       new  588b3f7   add whatsnew entry for GaussianNB sample weights
       new  08b97b3   Deprecate estimator_params in RFE and RFECV #4292
       new  2bc7ffe   Add "See also" for selectors and scoring funs
       new  9c24e1a   make landscape.io much more useful
       new  9f312a7   Correct warning messages and move warnings to fit method #4292
       new  2dfdf38   Merge pull request #4331 from martin0258/uni-feature-doc-seealso
       new  613c3af   Merge pull request #4335 from martin0258/dpgmm-lowerbound-bug
       new  e73c658   Vainly add link to my GH profile in What's New.
       new  5f2faac   Merge pull request #4378 from dan-blanchard/patch-1
       new  b741aa3   FIX coding style  #4292
       new  806cb22   remove some style errors in bench code.
       new  ae72a68   FIX numpy deprecation warning from unsafe type comparison
       new  26964ab   MAINT skip unstable (arpack dependent) estimators in check_pipeline_consistency on Win32
       new  adc51f5   MAINT skip unstable (arpack dependent) estimators in check_pipeline_consistency on Win32
       new  2b3a608   MAINT preparing branch for upcoming 0.16b2
       new  2b28b27   Merge pull request #4381 from amueller/style_errors_b
       new  c8d3df4   2 new debian/patches
       new  6544a4c   MAINT: Remove reimport
       new  7fa9b0e   Support sparse matrices in KernelRidge.
       new  4a57c92   Merge pull request #4383 from saketkc/maint
       new  9d4e47f   DOC Adds SVR description to the tutorial
       new  b15941b   TST Adds sanity check for SVR's decision_function
       new  eff77f5   DOC A small change to the SVM's tutorial
       new  c09bb35   Support sparse matrices in KernelRidge.
       new  63830dc   DOC: stock_market example: follow change of symbol
       new  cf3b120   TST Fixes #4386
       new  f0d86e8   DOC: Updated set_params doc
       new  bc14636   Fix typo
       new  dcdb7e7   Merge pull request #4395 from akitty/patch-1
       new  4be6214   Merge pull request #4295 from ragv/sigmoid_decision_fn
       new  46060fe   Pass the appropriate include_self argument to kneighbors_graph everywhere.
       new  6303af4   DOC make defaults more explicit in text feature extraction.
       new  48117d5   Merge pull request #4392 from sinhrks/setparams
       new  5610dba   Merge pull request #4390 from Barmaley-exe/tests-fix
       new  27d4376   ENH optimize DBSCAN by rewriting in Cython
       new  2e99294   FIX & TST at least min_samples to be considered a core sample
       new  0779566   Merge pull request #4401 from ogrisel/rebased-pr-4157
       new  ef9cbcb   added docstrings for restrict
       new  d69f2d6   DOC whats new for DBSCAN optims
       new  da33ac7   ENH optimize DBSCAN by rewriting in Cython
       new  c6949cd   FIX & TST at least min_samples to be considered a core sample
       new  c64f841   DOC whats new for DBSCAN optims
       new  5f44a8e   Raising an error when intercept_scaling is zero
       new  ed47b5b   Directly compute polynomial features.
       new  b1b43c1   DOC fix broken image link in feature selection
       new  13c2e35   DOC fix broken image link in feature selection
       new  211f204   fix ompcv on old scipy versions
       new  a13a585   DOC remove sphinx warnings when generating the doc
       new  5d16b68   Use anonymous hyperlink targets
       new  c5a0465   DOC remove sphinx warnings when generating the doc
       new  c93c845   Use anonymous hyperlink targets
       new  05100fc   FIX : allow NaN in input of calibration if estimator handles it
       new  6beacc3   remove deprecated stuff from 0.17
       new  e1af92b   remove cross_validation.Bootstrap
       new  9896218   Merge pull request #4402 from amueller/ompcv_fix
       new  defc0a9   Merge pull request #4388 from Barmaley-exe/svr-doc-n-test
       new  3b86bb1   fix ompcv on old scipy versions
       new  cec61b5   DOC Adds SVR description to the tutorial
       new  314ff63   TST Adds sanity check for SVR's decision_function
       new  ac1ec14   DOC A small change to the SVM's tutorial
       new  677384b   Merge pull request #4407 from agramfort/calibration_nan_fix
       new  870c629   FIX : allow NaN in input of calibration if estimator handles it
       new  c4df19d   Merge pull request #4370 from amueller/0.17_deprecated
       new  88759ba   MAINT remove the deprecated n_iterations param
       new  81acb94   Merge pull request #4411 from ragv/maint_remove_n_iterations
       new  bf5fd93   MAINT Remove the deprecated n_iterations from StratifiedShuffleSplit
       new  bc5acea   Merge pull request #4412 from ragv/maint_remove_n_iterations_2
       new  a4f8f08   Specified original intercept_scaling value
       new  2638fe5   Use immutable default arguments throughtout repo
       new  ea932af   Added docstrings for kernel='precomputed'
       new  9c66a20   avoid NaN in CCA
       new  82e4415   FIX Minor shape fixes in PLS
       new  6bbd7c0   make cross_decomposition pass common tests
       new  09e6e12   fix LarsCV and LassoLarsCV fails for numpy 1.8.0
       new  7328763   Merge pull request #4416 from vortex-ape/kernel_precomputed
       new  1119a4f   Added docstrings for kernel='precomputed'
       new  119bc20   Merge pull request #4377 from vortex-ape/intercept_scaling
       new  f295b87   Raising an error when intercept_scaling is zero
       new  cd4fef9   Merge pull request #4322 from amueller/kneighbors_include_self_fixes
       new  f52611f   Pass the appropriate include_self argument to kneighbors_graph everywhere.
       new  c6f6cf3   Merge pull request #4368 from xuewei4d/deprecate_estimator_params
       new  503cf4c   Deprecate estimator_params in RFE and RFECV #4292
       new  aa60779   Merge pull request #4409 from bryandeng/immutable_defaults
       new  3847a8c   deal with y lists
       new  d133b1b   DOC fix n_jobs docs in KMeans as in 0a611193b12900dbc11f3dae4448809364161bb2.
       new  90736ef   DOC fix typo in output shape of fetch_lfw_pairs (and minor additions)
       new  c02cc3d   FIX LDA(solver="lsqr"): make sure the right error is raised on transform.
       new  3b32099   DOC fix link description to setuptools in devdocs.
       new  ea5474b   Raising an error when n_clusters <= 0
       new  cd2ee7e   MAINT docstring --> comments to prevent nose from using doc in verbose mode
       new  f763a34   MAINT use yield for cleaner output in verbose mode
       new  a454242   sklearn-theano
       new  fa616f5   fix uri
       new  2badad3   Merge pull request #4433 from gwulfs/master
       new  ad26ae4   Merge pull request #4422 from trevorstephens/y_numeric_for_np180
       new  f50bfde   fix LarsCV and LassoLarsCV fails for numpy 1.8.0
       new  04078be   FIX make StandardScaler & scale more numerically stable
       new  cbe8053   DOC whats_new entry for stabler StandardScaler
       new  8f67a81   Merge pull request #4432 from ragv/travis_ignore_docstring
       new  12b2f16   COSMIT missing newlines in metrics test
       new  b596dfc   FIX test for older versions of numpy
       new  574ebfd   Merge pull request #4436 from ogrisel/rebased-pr-3747
       new  8702801   used special.fdtrc instead of fprob
       new  dbcbdcd   FIX make StandardScaler & scale more numerically stable
       new  689388a   Merge pull request #4427 from amueller/lda_lsqr_predict
       new  9db8220   FIX LDA(solver="lsqr"): make sure the right error is raised on transform.
       new  34c85ed   [WIP] preliminary implementation of stratified train_test_split
       new  803dcc1   fixed issue when train_size < test_size for stratified
       new  5bedb73   Link to functions and fix typos in dataset docs
       new  b4a92e5   Fix class reference in twenty_newsgroups.rst
       new  6e8ad67   fix setting description in 20newsgroups
       new  0cf41c8   Moved optimize test to a new file
       new  6354e45   Merge pull request #4389 from vortex-ape/move_test
       new  6cd1cae   using StratifiedShuffleSplit instead of StratifiedKFold
       new  8ea474c   cleaned up input validation, added tests for copy keyword.
       new  5035b18   ENH minor fixes / simplifications for PLS input validation, add test for error.
       new  fc681f2   Merge pull request #4439 from bryandeng/dataset_docs
       new  8762c28   Link to functions and fix typos in dataset docs
       new  155c81a   Fix class reference in twenty_newsgroups.rst
       new  63812b7   Friendly error on KNeighbors(n_neighbors<=0 )
       new  9c57ee3   Raise ValueError when log_multivariate_normal_density has non positive definite covariance
       new  cbd7537   Merge pull request #4350 from vortex-ape/build_issue
       new  1a4adb8   used special.fdtrc instead of fprob
       new  6f335af   fix issue 4447 : min_weight_leaf not properly passed to PresortBestSplitter
       new  5018a44   add regression test case for issue 4447  - fix another test
       new  5ba271b   cosmit
       new  29a88f4   Related projects linking to wrong URL
       new  202c186   Add "See also" for selectors and scoring funs
       new  0e365ca   FIX numpy deprecation warning from unsafe type comparison
       new  87e1652   MAINT: Remove reimport
       new  ef25b77   DOC: stock_market example: follow change of symbol
       new  a870b90   ENH Scale the sum_of_confidences to (-0.5, 0.5)
       new  642ae59   DOC make defaults more explicit in text feature extraction.
       new  cad1ac1   DOC: Updated set_params doc
       new  5b2aba1   TST Fixes #4386
       new  c62238a   Directly compute polynomial features.
       new  5ff34b2   DOC fix n_jobs docs in KMeans as in 0a611193b12900dbc11f3dae4448809364161bb2.
       new  8159c65   DOC fix typo in output shape of fetch_lfw_pairs (and minor additions)
       new  3df1971   DOC fix link description to setuptools in devdocs.
       new  f9b6899   sklearn-theano
       new  1491843   DOC minor fixes in formatting, don't use deprecated n_components in Agglomerative
       new  59efcb5   DOC/MAINT remove deleted cluster.Ward from references.
       new  33d132a   DOC minor fixes in formatting, don't use deprecated n_components in Agglomerative
       new  d0db91a   add huber to non-uniform sample weight toy test case
       new  f46cd25   Merge pull request #4448 from pprett/fix-gbrt-min-leaf-weight
       new  9300848   DOC/TST docstring --> comment
       new  2ff021f   fix issue 4447 : min_weight_leaf not properly passed to PresortBestSplitter
       new  73c0283   Merge pull request #4451 from ragv/travis_ignore_docstring
       new  0dc37f5   Adding sparse support for decision function
       new  b1ddbf4   Fixes to sparse decision_function, rebase fixes.
       new  3187886   Merge pull request #4189 from amueller/sparse_decision_function
       new  b278440   Adding sparse support for decision function
       new  e95815a   MAINT docstring --> comments to prevent nose from using doc in verbose mode
       new  fa6af0f   Fixes to sparse decision_function, rebase fixes.
       new  4aaaaae   DOC added highlights to whatsnew
       new  df9f90c   REL Version 0.16.0 (adjust __version__, links in docs)
       new  82a7968   DOC / website add 0.16 release to "news".
       new  8705e6d   DOC added highlights to whatsnew
       new  e7eb90c   DOC / website add 0.16 release to "news".
       new  850fe65   DOC More version menu fixes for the website
       new  752808e   Merge tag '0.16.0' into dfsg
       new  2b05390   Merge branch 'dfsg' into debian
       new  9080325   DOC fix link to Birch in whatsnew.rst
       new  5825084   FIX one more whats_new.rst fix
       new  2b3fabc   FIX one more whats_new.rst fix
       new  0117fb5   DOC fix link to Birch in whatsnew.rst
       new  6d93057   updated changelog and github revision
       new  5e50259   Fix signature not compatible with previous declaration
       new  aa49239   MAINT: update mailmap
       new  17ee7dd   DOC & COSMIT deprecate estimator_params in RFE and RFE_CV in docstring
       new  c6d420e   Modified docs
       new  c247141   Merge pull request #4462 from xuewei4d/deprecate_estimator_params_docstring
       new  22963a0   Added import statement for Lasso example
       new  e9939c3   adjustments to make it build on wheezy etc
       new  b1348bb   Typo: PCA is not the abbreviation of Probablisitic
       new  56546a7   Merge pull request #4423 from vortex-ape/agg_clustering
       new  9e7a660   added import np statements
       new  65446a8   Merge pull request #4476 from CaMeLCa5e/master
       new  3433b71   Merge pull request #4471 from baharev/small-typo
       new  53a119a   FIX don't raise memory error in ledoit wolf
       new  e2dfd23   Merge pull request #4428 from amueller/ledoit_wolf_shrinkage_fix
       new  d89c215   Add tags to classifiers and regressors to identify them as such.
       new  188fb11   COSMIT use consistent shape description in docstring.
       new  b28d3bf   Merge pull request #4464 from vortex-ape/logregcv
       new  b18e7dd   Merge pull request #4456 from cgohlke/patch-1
       new  6e54079   Merge pull request #4356 from vortex-ape/dict_vectorizer
       new  f41f2e4   FIX: Ensure dependencies installed Fixes #1495
       new  4d7fb16   Removed array2d from developer docs
       new  98dfdc6   Make apply method of trees public. Added test for concistency with private method.
       new  cb687fd   Added docstring
       new  8bc650b   Added example demonstrating tree.apply
       new  fdb42ea   Added indentation to docstring
       new  c409ef0   Removed cruft
       new  cfe4e15   Added tests of apply() for valid and invalid inputs. Fixed style.
       new  fd3c630   Mention oob_score in narrative documentation of RandomForest*
       new  88811be   Mention oob_score in the narrative docs for Bagging*
       new  1c33a6f   Merge pull request #4489 from betatim/rf-oob-docs
       new  0f0e58d   Fix RFE - Add the possibility to use feature_importances rather than coef_ when existing, see #2121
       new  afd2b53   Merge pull request #4484 from vortex-ape/remove_array2d
       new  c4b650f   Merge pull request #4496 from vmichel/rfe_feature_importances
       new  acb21bb   DOC adding clusterer tag to dev docs.
       new  aae305c   Merge pull request #4418 from amueller/classifier_regressor_tags
       new  6c81fbc   Merge pull request #4420 from amueller/cca_low_rank
       new  03dbec8   fixed reverse sorting issue in label_binarize()
       new  a94301c   added an additional test case in test_label.test_label_binarize_with_class_order()
       new  5272512   FIX _make_unique of _isotonic.pyx uses float64 consistent internally
       new  6560a8e   Merge pull request #4505 from jmetzen/isotonic_make_unique_fix
       new  edf60c0   BUG: get_param in FeatureUnion disjoint for shallow/deep params
       new  b810c9d   CLN: Added missing __all__
       new  db665bb   Merge pull request #4514 from sinhrks/missing_all
       new  485974f   DOC try to clarify pairwise_distances docstring for Y.
       new  f8102e5   DOC fix link in text extraction narrative
       new  1b69807   DOC Grammar fix in preprocessing
       new  871ffec   Merge pull request #4519 from ClimbsRocks/patch-1
       new  9206a78   Merge pull request #4516 from amueller/distance_function_docstring
       new  bfef25a   Merge pull request #4506 from mheilman/master
       new  3f5fbbd   FIX don't raise memory error in ledoit wolf
       new  d54b856   FIX _make_unique of _isotonic.pyx uses float64 consistent internally
       new  28546c8   fixed reverse sorting issue in label_binarize()
       new  7097af0   PSL and CCA convergence check and input validation fixes
       new  72450ea   DOC GradientBoostingX.estimators_ docstring shape.
       new  5527084   FIX KMeans with fortran-aligned data.
       new  d0302d6   TST: add a test for fortran-aligned data be used in KMeans.
       new  2b66687   FIX make CalibratedClassifierCV deterministic by default.
       new  2305bdc   Merge pull request #4537 from amueller/callibrated_classifier_cv_linearsvc_seed
       new  89c1dc5   FIX be robust to columns name dtype and also to dataframes that hold dtype=object.
       new  b1eaac0   DOC typo
       new  28a0994   ENH: public apply method for decision trees
       new  1aa4249   ENH: raise NotFittedError instead of ValueError
       new  a19b1fc   DOC add warning for partial pip upgrades
       new  46aacd3   Merge pull request #4531 from zhaipro/fix
       new  86273fd   FIX KMeans with fortran-aligned data.
       new  6363ca2   TST: add a test for fortran-aligned data be used in KMeans.
       new  6ea24c7   Update what's new
       new  8d4b9d6   a simple fix to reading empty label sets in the multilabel case
       new  7ac6c9b   FIX respect astype(array, dtype, copy=True) on old numpy
       new  8c31b79   cython compiled from pyx
       new  249a96b   directly set target to empty list instead of empty string
       new  e72e38b   given that line_parts has been copied to features, parsing should be based on features rather than line_parts
       new  10a056f   cython the pyx
       new  95fb92b   TST add missing test for astype backport
       new  a7cdbfe   STYLE PEP8
       new  66f977a   Merge pull request #4542 from ogrisel/fix-doc-partial-pip-upgrades
       new  7c12089   DOC add warning for partial pip upgrades
       new  9a6e945   DOC fix fastica source matrix output shape
       new  44526cf   Merge pull request #4561 from mspacek/patch-1
       new  636d794   REL Add whatsnew for 0.16.1
       new  c0bdb51   FIX make shuffle / resample pass-through indexing utilities
       new  28765fc   Improve FastICA convergence warning message
       new  4387016   MAINT remove debug nslookup call
       new  3531c60   Merge pull request #4563 from mspacek/patch-2
       new  141a74b   use COLON to avoid byte buffer error
       new  ce5fada   Deprecated load_lfw_pairs and load_lfw_people
       new  79d9aa2   STYLE PEP8
       new  e12863f   FIX make shuffle / resample pass-through indexing utilities
       new  cbf9e75   Merge pull request #4560 from ogrisel/shuffle-passthrough
       new  f0f4c79   STYLE PEP8
       new  3001e6d   FIX make shuffle / resample pass-through indexing utilities
       new  6efb03f   Changed import statement and checked for ImportError
       new  f28b41f   add the empty label test case
       new  eff590a   add another test case with empty labelset
       new  7461b6c   Nest the for loops because they don't need to run if the condition is not true.
       new  58dfb3d   Merge pull request #4555 from ogrisel/fix-astype-copy-true
       new  29dddc7   DOC: better documentation tree.apply
       new  28c893d   Merge pull request #4488 from glouppe/tree-apply
       new  4b70b61   Merge pull request #4570 from thedrow/topic/call-enumerate-on-empty-clusters-only
       new  a0775c1   Merge pull request #4530 from amueller/gbrt_estimators_docstring
       new  5ef503c   Update whats_new.rst
       new  624c9f4   Merge pull request #4581 from kno10/patch-1
       new  e6d5098   MAINT do not ctags index non-source artifacts
       new  f9eb9cf   STYLE whitespace around operator in sklearn.utils.fixes
       new  01da5c2   ENH implement fit_predict on pipelines
       new  16d1322   verifies fit_predict is available on last step of pipeline
       new  ead38de   better docstring one-liner and use of gettattr in test
       new  205651f   Merge pull request #4584 from sdegryze/fit-predict-for-pipeline
       new  a5f0159   don't to input validation in each tree for RandomForest.predict
       new  50abfc7   Merge pull request #4588 from amueller/random_forest_predict_check
       new  011ad8d   ENH add a prior strategy for the dummy estimator
       new  f99830c   Update what's new
       new  203298e   Merge pull request #4541 from amueller/robust_input_dtype_check
       new  6d604c9   FIX be robust to columns name dtype and also to dataframes that hold dtype=object.
       new  a76b5be   don't to input validation in each tree for RandomForest.predict
       new  539f488   DOC GradientBoostingX.estimators_ docstring shape.
       new  67a8b7d   DOC / Website update whatsnew with RF speed fix and utils.shuffle pass through fix.
       new  bb39b49   REL bump version to 0.16.1, change in documentation.rst, add to news
       new  c9510ed   Fixing bug in assert_raise_message #4559
       new  af0be5f   Misc trying to make k_neighbors warning even more explicit (before it didn't say that %d was n_samples)/
       new  cb219b4   add 0.16.1 bugfixes to whatsnew.
       new  50b42c7   Merge tag '0.16.1' into releases
       new  cc9de2c   Merge branch 'releases' into dfsg
       new  4875dba   Merge branch 'dfsg' into debian
       new  adb33ff   changelog
       new  7be35c8   Merge pull request #1 from scikit-learn/master
       new  af1a74c   add py3 xrange to bicluster example
       new  02b2712   change to explicitly import function
       new  641bd93   Change xrange to range and remove patching import
       new  00409d7   DOC use standart X, y notation
       new  e7b6a7f   Merge pull request #4596 from ibayer/doc_train_test_split
       new  f3f0fa3   Merge pull request #4590 from jfraj/fix_bug4559
       new  ea8f19b   Merge pull request #4595 from sseg/bicluster-example
       new  0fe613e   Add conventions section to userguide
       new  5ad1dfe   Fix Float Resolution Bug on Gaussian Process Test
       new  4161d55   Improving Bunch Class to ensure consistent attributes
       new  3cf765f   ENH faster LSHForest: use take rather than fancy indexing
       new  1f74f1b   Merge pull request #4600 from jfraj/bug_20newsgroup
       new  0949a2e   remove order checks for sparse matrices
       new  75a0d2c   Merge pull request #4606 from ibayer/remove_order_check
       new  e4b9ff6   Adding test to ensure bunch consistency after pickle
       new  6d5328b   Merge pull request #4608 from jfraj/test_pickle_bunch
       new  ced25e9   Change urlopen and iteritems functions for Python3 compat
       new  8de3af5   Merge pull request #4604 from jnothman/faster_lshforest
       new  932726f   Merge pull request #4467 from giantoak/param_invariance
       new  4e6346e   FIX raise error properly when n_features differ in fit and apply
       new  4ffbf34   Factorize input validation
       new  4b8b200   ENH Factor out validation of X for apply/predict in forest
       new  294a621   FIX properly raised not fitted error
       new  32290ad   FIX use the new notfittederror in feature_importances_
       new  92bda97   DOC use same error message in decision tree and random forest
       new  7bbf10a   ENH optimize rfecv by eliminating the for loop
       new  c165327   Add regression test for the number of subsets of features
       new  acc1ac2   Merge pull request #4534 from xuewei4d/optimize_rfecv
       new  8d66906   ensemble classifier 1st commit
       new  2910198   votingclassifier
       new  24f88b8   added _check_proba method
       new  00abe26   property trick to raise AttributeError if voting is "hard"
       new  3f0fb6b   DOC fix libsvm docstring consistency
       new  0078b45   Merge pull request #4610 from sseg/ex-wikipedia-py3
       new  9e710ed   WIP: adding 'max' normalizer to normalize()
       new  c5358ca   ENH add ranking_loss multilabel ranking metric
       new  52244dc   incorporated amueller's feedback
       new  c97ad1d   Merge pull request #4592 from clorenz7/bugfix/gauss_proc_eps_issue
       new  6c5caa3   Merge pull request #4434 from xuewei4d/catch_LinAlgError_gmm
       new  06b78a0   Added/improved GraphLasso tests, per #4134
       new  20b188c   Minor docstring fixes to svm/classes.py
       new  c601005   Improve documentation on Diabetes dataset
       new  f8e7bd2   Typo fix in diabetes documentation :Aex: -> :Sex:
       new  29eb6bf   Update confusion_matrix docstring
       new  60a960c   Merge pull request #4613 from bnaul/graph_lasso_tests
       new  c4bbb58   Update confusion_matrix docstring
       new  f6b52d2   Merge pull request #4636 from scls19fr/scls19fr-Update_confusion_matrix_docstring
       new  fafbddb   Improve wording of SVM `crammer_singer` doc following PR feedback.
       new  027eb6d   ForestClassifier.predict docstring correction
       new  95f9be5   Merge pull request #4371 from saketkc/fix_setup
       new  e7bf670   Clarify documentation on feature scaling for diabetes data
       new  211c916   Merge pull request #4631 from Aerlinger/documentation_improvements
       new  669a4ea   ENH add high-level estimator-validation function
       new  2d4c8d7   check output, not error on sparse data handling
       new  196a7b7   fix warning message by not using assert_raise_message, which adds escaping.
       new  8ecfa3c   add comments to tests
       new  14c54c4   FIX make backport of assert_raises_regex raise the same error as the python3 version.
       new  79ea587   FIX make rfe feature importance test deterministic.
       new  ec2a89a   Merge pull request #4640 from dougalsutherland/patch-1
       new  c78f262   FIX raise ValueError if sample weight are passed but unsupported by the base estimator
       new  1197eb4   Tweak appveyor badge
       new  d3e9a7d   Merge pull request #4647 from lesteve/better-appveyor-badge
       new  b92ed8f   Merge pull request #4550 from amueller/common_test_refactoring
       new  af1b652   FIX DBSCAN fit with precomputed matrix in edge cases
       new  951b8ff   FIX TSNE.fit: didn't return self
       new  9dc8118   Merge pull request #4526 from vortex-ape/if_matplotlib
       new  36c7ddb   FIX pass percentiles to partial_dependence in plotting.
       new  fc87eb3   Add a fit_predict method for the GMM classes
       new  e1fd955   Merge pull request #4593 from clorenz7/gmm_fit_predict
       new  3b23d25   Added a verbose flag to the GMM class.
       new  dc34af7   Fixed typos.
       new  1a1a69c   Fixed doctests.
       new  1b9f4b9   Fixed pep8 warnings.
       new  eb05c53   Added tests for the verbose flag of GMM.
       new  0bb91dc   Fixed pep8 and removed unused variable.
       new  0ae3110   Adapted the verbose flag of DPGMM. Boolean flag is deprecated now.
       new  8f34306   Added tests for the adapted verbose flag of DPGMM.
       new  afeb7ee   remove useless |samples| variable out of ParameterSampler.__iter__()
       new  0a4b85b   replace itertools.repeat with six.moves.range
       new  899cc87   Merge pull request #4659 from MAnyKey/sampler-samples
       new  37a9ce1   DOC: add a testimonial by infonea
       new  542e9b9   Merge pull request #4662 from GaelVaroquaux/infonea_testimonial
       new  f10d2f4   Merge pull request #4528 from amueller/rfe_feature_importance_test
       new  0007660   inline check_proba
       new  017f289   Initial commit of plot_ensemble_oob.py
       new  43a4636   Removed the deprecation and added a test for a boolean flag value.
       new  d2382e1   Cleaned the code. pep8, pyflakes and added comments.
       new  4618cb3   Merge pull request #4653 from ssaeger/issue_4633
       new  259e983   fix minor documentation link issues. looks more complicated than it is as I replaced tabs with spaces
       new  ffb81e4   Merge pull request #4650 from lesteve/fix-4641
       new  3594925   Store tree node info for all nodes, instead of just leaves
       new  e998777   FIX remove deprecation warning
       new  c440190   Merge pull request #4642 from arjoly/error-bagging-sw
       new  5b2ee7a   Minor amendments to plot_ensemble_oob.py.
       new  b6ce322   Amended plot_ensemble_oob.py according to #4665.
       new  6d89b14   Fix string literal concatenation whitespace
       new  88e0c69   Merge pull request #4676 from zacstewart/master
       new  cdfaabf   FIX work-around for read only dataframes
       new  190fe73   FIX Include PyFuncDistance attributes while pickling.
       new  d3fba99   TST Add test to make sure BallTree remains picklable even with callable metric
       new  a685fe5   TST make sure DistMetric is picklable with both predefined as well as custom metric
       new  2f563df   Added warm_start to bagging
       new  b03af57   Fix random_state to make test reproducible
       new  4408a18   Test for unchanged n_estimator and indentation fix
       new  51a51ef   Test oob_score for warm_start'ed bagging classifiers
       new  d83a362   Removed overly deep indentation
       new  fb9e725   oob_score will only be calculated if warm_start=False
       new  919684c   Keep estimators_samples_ and estimators_features_ across warmstarts
       new  8add36b   Better assert_raises use, API compliance, better ordering
       new  3118585   Update what's new
       new  a50afc0   Merge pull request #4678 from amueller/pandas_safe_indexing_fix
       new  b17b3aa   [Pipeline] add named_steps attribute documentation #4482
       new  fb95606   add gplearn to related projects
       new  91614a7   Merge pull request #4681 from trevorstephens/gplearn-link
       new  e038cb3   Made more amendments according to PR feedback.
       new  e61520f   Merge pull request #4654 from amueller/fix_partial_dependency_percentiles
       new  6f26cf0   Typos in comments corrected
       new  dd9641f   FIX #4597: LassoLarsCV on with readonly folds
       new  2bf07b2   Merge pull request #4683 from yanlend/patch-1
       new  6bd0844   ENH make pipeline.named_steps a property, fix pipeline.named_steps doctest
       new  9eb1964   Merge pull request #4161 from rasbt/ensemble-classifier
       new  6ac6ea3   COSMIT: VotingClassifier
       new  ef94a2a   Merge pull request #4665 from kianho/plot-ensemble-oob-trace
       new  2656f7b   Refined the docstring for plot_ensemble_oob.py. - Minor change to #4665: cleaned up the writing style of the documentation.
       new  fa5bd2c   Merge pull request #4692 from kianho/plot-ensemble-oob-trace
       new  1b1e10c   ENH multioutput regression metrics
       new  e6b7d2c   DOC whatsnew for unaveraged multi-output regression metrics.
       new  007ae76   TST: covering norm='max' branches of normalize()
       new  84ee88d   DOC: updating Normalizer docstring for norm='max'
       new  5fcad7c   Fixing sparse max for older scipy
       new  8fff05c   Merge pull request #4438 from mbatchkarov/stratified_train_test_split
       new  f65d7a7   DOC add whats_new for stratified train_test_split
       new  64e730e   add OnlineLDA model and test
       new  efd5e3b   don't use sparse.block_diag since it is new in scipy 0.11.0
       new  2718499   Pretty decision trees
       new  efd5afd   fix typos and imporve documentation
       new  b7b270a   fix typo & improve doc 2
       new  794c685   testing code improvement
       new  a0ba403   move input validation and variable assignment to fit method
       new  e6e0dd7   fix _approx_bound with subsampling
       new  68fde20   Merge pull request #3735 from trevorstephens/pretty-decision-trees
       new  d6a4bc0   update export_graphviz docs
       new  360456b   native dense input support
       new  36df860   Merge pull request #4695 from perimosocordiae/patch-1
       new  64f7e54   Merge pull request #4698 from trevorstephens/doc-export_graphviz
       new  0aea95c   DOC add random_state to parameter docstring in gradient boosting
       new  64e5533   DOC document missing attributes
       new  2e87da8   ENH don't use mutable defaults in RidgeCV.
       new  e8356d3   FIX add missing numpy imports to VotingClassifier examples, pep8 fixes.
       new  da9a91d   COSMIT remove unnecessary float casting from tomography example
       new  8f3ef2a   Merge pull request #4713 from amueller/ridge_cv_nonmutable_alphas
       new  f11f5c8   TST don't check that warning was raised as whether a copy needs to be made or not depends on the pandas version. Version 0.12 and 0.13 apparently don't use cython here, and >0.17 will have a fix.
       new  ec1bba7   FIX / TST raise error when init shape doesn't match n_clusters in KMeans, check for sensible errors.
       new  69e344f   DOC fix sample_without_replacement docstring. Closes #4699.
       new  e42a0fb   Ensure dict_learning_online always has at least one batch
       new  175f50e   Check fit() returns self for all estimators
       new  1678bd5   Merge pull request #4694 from betatim/check-fit-returns-self
       new  28cd9c8   Merge pull request #4702 from betatim/online-learning-batch-size
       new  4d72563   Doc sphinx fix Return->Returns in label_ranking_loss
       new  df59b31   Fix mistake in TruncatedSVD expalation in docs
       new  ef91999   Merge pull request #4727 from Spikhalskiy/truncated-svd-doc-T
       new  c41e640   rename variable and minor pep8 fix
       new  50c0a8d   Fix documentation errors.
       new  f7ca553   Remove warn_if_not_float
       new  e595792   Remove 'copy' parameter form RobustScaler functions
       new  4ff47d9   Fix documentation in test case
       new  5a2eb01   FIX make random_choice_csc and therefor DummyClassifier deterministic.
       new  29e8377   Export robust_scale and RobustScaler
       new  6aefa53   Improved documentation of robust scaler
       new  9ffd954   Fixed decprecation warning
       new  e5fb9dd   Changed outdated test
       new  c296e49   Fixed removal of copy-parameter
       new  dd2fdbb   Removed superfluous testcase for robust scaler
       new  d747541   Merge pull request #4733 from amueller/fix_csc_random_sampling
       new  b9d8fd5   ENH support for sparse coef_ in ovr classifier
       new  2a830f6   clean up test and add error check in preplexity method
       new  3ddeb0e   remove duplicate code in _approx_bound method
       new  6b16e9d   move check_non_negative to utils/validation.py
       new  4c6ad90   Remove interquartile_scale parameter
       new  d86a629   Added robust scaling example
       new  7dbe3a8   Fix robust_scaling test
       new  2de9f37   Fixed documentation oversight.
       new  283d50b   robust_scaling example
       new  3618a02   fix params and import test coverage
       new  b2eeb33   update example and change total_samples in fit method
       new  af59fe5   typo: 'precomputed' should be in quotes
       new  ed104bd   Merge pull request #4734 from ajschumacher/patch-6
       new  90d5ef1   TST test that all default arguments are not mutable
       new  c4df418   TST: Regression test for nan parameters cloning
       new  1bf81b1   ENH: Avoid equality comparison for nans.
       new  8b386d8   DOC some fixes for the EnsembleClassifier
       new  a64b060   DOC typo in TruncatedSVD narrative
       new  25e24a7   Merge pull request #4725 from jseabold/avoid-nan-equals-comp
       new  89c1018   ENH improve check_array
       new  1abb41a   DOC add comment for np.nan equality test in clone
       new  c51f067   ENH change func_grad_hess into grad_hess
       new  9385c45   Merge pull request #4645 from TomDLT/astype_fix
       new  c5b395f   FIX / TST make cross_val_predict work on lists, test pass-through and compatibility with various input types.
       new  da021ab   FIX ransac output shape, add test for regressor output shapes
       new  8380b71   don't warn about multi-output deprecation if not using multi-output
       new  f4366d5   FIX Gram OMP check_finite on old scipy
       new  db56cf7   Merge pull request #4705 from amueller/cross_val_predict_input_types
       new  b88431b   Merge pull request #4697 from amueller/remove_warning_test_pandas_slicing
       new  797e0dc   Merge pull request #4730 from amueller/ovr_classifier_sparse_coef
       new  92de5c4   MAINT use the new container-based travis workers
       new  04b103f   Merge pull request #4743 from ogrisel/fix-check-finite-gram-omp
       new  fe38e25   ENH check that unfitted estimators raise ValueError
       new  6873c9f   Merge pull request #4684 from ogrisel/readonly-lars-path
       new  4121c30   Merge pull request #4749 from ogrisel/travis-no-sudo
       new  990a6b4   FIX n_jobs slicing bug in dict learning, add input validation to gen_even_slices
       new  20c622f   Improved example
       new  ea2e8a8   Merge pull request #4741 from amueller/multioutput_warning
       new  5029db7   Merge pull request #4679 from TomDLT/newton_cg
       new  1174e58   DOC update whats_new
       new  6a3365d   ENH use astype to avoid unnecessary copy
       new  0ac46aa   Merge pull request #4753 from TomDLT/newton_cg
       new  922d71a   FIX check parameter in LogisticRegression
       new  7e05d96   Merge pull request #4748 from TomDLT/check_maxiter
       new  d53634f   Merge pull request #4754 from TomDLT/astype_fix
       new  c12ae60   Fix documentation errors
       new  c1b2937   FIX fix test_lars_path_readonly_data failure under Windows
       new  5cc234a   Merge pull request #4716 from TomDLT/notfitted
       new  cac627a   More documentation fixes
       new  815270b   Remove redundant code
       new  1c5f35c   FIX ResourceWarning in twenty newsgroups loader
       new  3af811e   Merge pull request #4362 from rvraghav93/make_PyFuncDistance_picklable
       new  e822fa7   STYLE removed unused import, fixed comment indent and spacing
       new  0ee4d97   Merge pull request #4751 from amueller/dict_learning_n_jobs_bug
       new  543d749   Added the regularization term in the method _fit_transform
       new  da73f5c   Merge pull request #4739 from amueller/regressor_output_shape_test
       new  bb592f3   DOC fix comment in tf-idf: log+1, not log(1+x)
       new  0e5226b   unused params in gridsearchcv
       new  bf33390   Merge pull request #4762 from trevorstephens/gridsearch_unused_params
       new  1defef0   printing top terms even if LSI is used
       new  543e5a2   sidebar hide on css
       new  ee45724   delete js of sidebar
       new  a8e68ba   A responsive behavior
       new  e25da0e   Recover similar to old style
       new  46a5031   add LDA user guide
       new  6826e13   add LDA to classes.rst
       new  ca07b2a   Merge pull request #4765 from alexeygrigorev/master
       new  61efaea   Add enet_tol parameter to GraphLasso class/methods
       new  15d6f98   Merge pull request #4686 from bnaul/graph_lasso_tests
       new  0b07536   Merge pull request #4125 from untom/RobustScaler
       new  355dc7c   ENH improve parameter check in LogisticRegression
       new  f3b1208   FIX typo in RobustScaler example
       new  549ecae   DOC various fixes in whats_new.rst
       new  4eda9e6   Merge pull request #4770 from TomDLT/logistic
       new  fc9d7be   FIX use random_state in LogisticRegression
       new  0b330be   DOC fix contributor name
       new  7041ed2   Implemented parallelised version of mean_shift and test function.
       new  4c39bdb   Merge pull request #4761 from tiagofrepereira2012/fix_lle
       new  71deb4a   DOC added some bug fixes to whats_new.rst
       new  7ae688d   Merge pull request #4778 from untom/fixname
       new  2d559ba   Merge pull request #4777 from TomDLT/liblinear_randomstate
       new  21dbe72   randomforest decrease allocated memory
       new  cc84db3   refactoring to simplify
       new  61dbf54   toggle shift to left larger vertical space
       new  4c4b1f7   Change par system to joblib and n_jobs convention
       new  6eee7a5   FIX pass random_state to inner call to randomized_svd
       new  1095e7e   toggle contraction character
       new  44f17b0   Merge pull request #4663 from Titan-C/sidebar
       new  a8626b3   TST/COSMIT remove nose call boilerplate
       new  57bf371   random initialize in e-step
       new  a8f33d3   Minor appearance changes
       new  bcd7e9b   refactoring to DRY.
       new  de4577b   Trivial function name bug fixed
       new  f369d3a   Fix RFE / RFECV estimator tags
       new  0d3fa51   DOC Add coveralls badge to Readme
       new  bdef419   Merge pull request #4792 from amueller/coveralls-badge
       new  ffe74ab   change LDA example to script
       new  11aad7f   change doc, fix n_jobs, and remove fit_transform
       new  32b7075   Merge pull request #4680 from amueller/pipeline_named_steps_fix
       new  2c79a98   Merge pull request #4431 from vortex-ape/deprecate_func
       new  eedc1cd   Use more natural class_weight="auto" heuristic
       new  123a8c9   FIX add class_weight="balanced_subsample" to the forests to keep backward compatibility to 0.16
       new  2ddb503   Cosmit readability improvements and better whatsnew.
       new  75e895e   Updated _std docstring in StandardScaler to make internal handling of zero values explicit.
       new  27264e1   Merge pull request #4785 from amueller/no_noise_boilerplate
       new  4ca3878   float division
       new  85986cd   Merge pull request #4786 from amueller/estimator_tags_bug
       new  01aff46   Merge pull request #4347 from amueller/class_weight_auto
       new  4515cec   change valiable name to meaningful.
       new  02fedc6   autopep8 E501
       new  05ad745   STYLE cosmetic fixes in sklearn.mixture.gmm
       new  0650d55   DOC adding backlinks to docstrings
       new  4f12323   Typo.
       new  ec2fd72   Merge pull request #4723 from amueller/backlinks
       new  1a0711e   Add API paper.
       new  29f9dc6   Reference ECML PKDD workshop paper.
       new  5512851   Merge pull request #4810 from mblondel/cite_api_paper
       new  867da31   DOC minor sphinx fixes
       new  3be9b97   Fix deprecation of decision function in SGD
       new  5632cfc   Merge pull request #4818 from amueller/sgd_decision_function_deprecation
       new  6d4bcda   ENH refactor OVO decision function, use it in SVC for sklearn-like decision_function shape
       new  0a5d924   TST set missing random state in test
       new  f0f174b   ENH minor fixes to the tests, don't raise as many warnings in the test suite
       new  41788c3   Changed SVM to use gamma='auto' by default. Deprecated gamma=0.0.
       new  92f9c9e   fix RidgeCV when cv != None
       new  ee78879   add class_weight to PA cls, remove from PA reg
       new  3f2f225   Merge pull request #4819 from sornars/svcgamma
       new  567a956   Merge branch 'pr/4714' + pep8 fixes
       new  1274e48   TST fix failing doctest
       new  62325cb   MAINT merge _check_cv into check_cv as indices argument is removed in 0.17
       new  d1d95e8   Merge pull request #4829 from rvraghav93/remove_deprecated_check_cv
       new  019fc9b   ENH labels parameter in P/R/F may extend or reduce label set
       new  78e45fb   fixed typos and update formatting in doc/developers/index
       new  9b4204e   additional typos, formatting, and links for doc/developers/index and /utilities
       new  97ba607   Fixed factual error in descriptino of PR flow
       new  99ed481   Merge pull request #4832 from kronosapiens/update-docs
       new  8453daa   Use expit function to compute the probability in ExponentialLoss class.
       new  1676a31   add polynomial features to user guide
       new  0233cce   add connection to kernel method
       new  4c4bca1   fix SVM can be tricked into running proba() #4791
       new  2f69574   add sample_weight to RidgeClassifier
       new  a7de146   Merge pull request #4833 from tw991/feature
       new  bdd3cb7   Merge pull request #4763 from sonnyhu/RidgeCV_slice_sampleweight
       new  8b44fa1   Merge pull request #4824 from amueller/testing_less_warnings
       new  41e10c4   Better docstring for PCA, closes #3322.
       new  32b2f8e   Merge pull request #4838 from trevorstephens/ridge_sw
       new  fcf9cb4   Merge pull request #4835 from tw991/feature2
       new  98ca513   Merge pull request #4796 from joshloyal/standard-scaler-std-doc
       new  088cbd6   DOC fix rendering of code-block
       new  167324f   ENH Add option in cosine_similarity for sparse output
       new  64d3f45   Merge pull request #4847 from TomDLT/doc_dev
       new  82d50e9   DOC make contributor's guide more prominent on front page
       new  9a4b75a   ENH Document how to get valid parameters in set_params
       new  ab734b7   ENH add MaxAbsScaler
       new  891b841   DOC remove linewidth in pca/ica example
       new  59aeb54   Fix typo in model evaluation documentation
       new  4c17f1c   Merge pull request #4854 from opahk/master
       new  95c9e14   PEP8 for recent merge
       new  87aaabc   DOC backticks in attribute docstrings unnecessaru since #3489
       new  dc8578a   Merge pull request #4828 from untom/maxabs_scaler
       new  606531d   DOC minor tweaks to docs homepage
       new  99d08b5   DOC organise documentation hierarchy / table of contents
       new  3193993   fixes #4846
       new  2928c50   DOC document missing attributes
       new  1d28349   Merge pull request #4857 from christophebourguignat/master
       new  1609cfa   remove css for container-index that compensates old js behavior
       new  fbd63e6   css for container-index that compensates position of main content on index.rst and documentation.rst in small screens
       new  3396087   Update plot_ols.py
       new  2bbec7b   Merge pull request #4853 from amueller/fix_ica_pca_example
       new  38227e1   Merge pull request #4861 from ltiao/patch-1
       new  34fe4fd   Merge pull request #4859 from Titan-C/sidebar
       new  11d08e0   DOC links should not always point to dev
       new  3aeaf0d   Merge pull request #4849 from jnothman/contrib-link
       new  078a22d   silhouette_score docstring update
       new  b5a18b1   Merge pull request #4868 from rasbt/silhouette
       new  c655bb9   changed clf to regr in decision tree regression examples
       new  f5b738c   Merge pull request #4870 from rasbt/decisiontreeregressor
       new  e8c2763   Merge pull request #4862 from jnothman/fixlink
       new  ff64c3a   MAINT remove referrence to HMM in common tests
       new  733629d   Merge pull request #4872 from ogrisel/remove-hmm-ref-in-tests
       new  3d097ae   pep8 style
       new  71c004b   REF Correct the reference link for additive chi squared sampler.
       new  9d9e99a   Merge pull request #4860 from rvraghav93/additive_chi2_sampler_reflink
       new  c49aa03   Merge pull request #4851 from trevorstephens/ridge_no_copy
       new  2f1a26e   merge NMF and LDA example
       new  1886f2b   [MRG] [BUG] Pass penaly to the final logistic regression fit
       new  cd98ffc   Merge pull request #4891 from MechCoder/lrcv_bug
       new  44b3ec8   Epsilon added to x and y weights in nipals inner loop for numeric stability on windows
       new  74ae2f0   Changed addition of epsilon for cases where y_weights is near 0
       new  9f06156   add a new int array to store indices
       new  9f60b18   Merge pull request #4562 from hsuantien/emptylabelset
       new  1c07dec   Merge pull request #4894 from tw991/rf
       new  febefb0   FIX avoid memory cost when sampling from large parameter grids
       new  7488fc0   Merge pull request #4840 from jnothman/dynamicgrid
       new  fdfb808   improve LDA doc
       new  1e74e7d   fix typo and improve doc
       new  2d3600f   use n_features_
       new  36bb611   remove self.n_features_
       new  a1c3561   remove main in test
       new  567aa61   make dimension mismatch err msg more explicit
       new  093088a   change all rng to random_state
       new  c66a839   Merge pull request #4893 from jmschrei/cca_stability
       new  9612521   FIX raises memory error in depth first builder
       new  4d1c1d5   Updated joblib to 0.9.0b2
       new  7330b1c   Merge pull request #4905 from lesteve/update-joblib-0.9.0b2
       new  e5dc0e8   Merge pull request #4900 from arjoly/fix-raise-memerror
       new  49ffd7f   MAINT make sure we do not download large datasets on appveyor
       new  6a4aa3a   Merge pull request #4836 from amueller/pca_docstring
       new  26fbfe6   ENH add minmax_scale
       new  0e976e1   Change name of variable to be consistent with dataset
       new  0e5db84   Add missing scalers to classes.rst
       new  19a1528   Merge pull request #4918 from ejcaropr/doc-fix
       new  ac40d4f   MAINT PLS Remove deprecated coefs attribute
       new  e7e22d1   DOC Improve wording in minmax_scale documentation
       new  d517846   FIX #4902: SV*.predict_proba visible before fit
       new  4ab89c5   upd randomforest docstring
       new  11833f1   Merge pull request #4932 from rasbt/randomforest
       new  1cd4e7f   [docs] [trivial] Fix small error in cross-validation docs.
       new  9df1128   Merge pull request #4928 from rvraghav93/pls_deprecated
       new  ed73bbf   Merge pull request #4927 from jnothman/hiding_predict_proba
       new  a5feec2   Merge pull request #4934 from thvasilo/patch-1
       new  5b7768b   fix feature_names
       new  12859d9   improve document in _online_lda.pyx
       new  4491d10   add a example for FeatureHasher
       new  bf86954   add an example for FeatureHasher
       new  0c48570   use logsumexp
       new  243983d   try to mollify travis
       new  d9e4631   Updated graph ranges
       new  dce50be   changelog for upload to unstable
       new  7bc059f   added patch changeset_5e502596460014ed3adebe61e8189a60e72d3623.diff to "cherry-pick" 5e502596460014ed3adebe61e8189a60e72d3623 (Fix signature not compatible with previous declaration)
       new  0174be6   added patch changeset_190fe7347cfd6fab67f37c7535c4fa9b8f0870de.diff to "cherry-pick" 190fe7347cfd6fab67f37c7535c4fa9b8f0870de (FIX Include PyFuncDistance attributes while pickling.)
       new  28c9545   changelog for previous 2 patches
       new  a4b52f2   debian/control - boosted policy to 3.9.6 - properly dropped python-support
       new  25082e5   deb_skip_some_tests_on_arm fixed it up (missing nose imports). Thanks Logan Rosen!  (Closes: #773862)
       new  88e6617   LOOK INTO TODOS
       new  a1d6bb1   DOC Fix doc for NMF `init` default
       new  ad5e229   Merge pull request #4952 from stephen-hoover/fix_nmf_init_doc
       new  337fa5c   compare residual_threshold directly with 0.0
       new  8bc2fa3   Merge pull request #4947 from zyrikby/patch-1
       new  5aa5a96   Correctly match penalty and loss with arguments
       new  0e15b03   Add a space to an error message
       new  8b6a4e3   Merge pull request #4953 from mrphilroth/issue4940
       new  a1391c2   if there are no inliers, raise ValueError mentioning possible cause being residual_threshold too low along with the residual_threshold, and a test that the ValueError occurs
       new  d92d35f   Merge pull request #4783 from tokoroten/decrease_randomforest_memory
       new  3c8c492   fixed newlines, comments
       new  0241370   Fixed test to check for capital No
       new  cfd4e62   Closes #4614
       new  9346479   note about variance reduction
       new  8d4c9c3   Don't allow negative values for inverse transform, and raise error similar to transform for new labels
       new  e28b6bd   Added random seed for facial recognition example and updated the docstring
       new  b26e879   allowing for different cluster stds
       new  06f16e2   Added test
       new  4f2a93d   adding multiple cluster_std to make_blob test
       new  e03d42b   Merge pull request #4962 from vinc456/unexpected-facial-recognition
       new  3a82309   extend R^2 description
       new  e095c4b   changed format string in no inliers exception to be 2.6 compatible
       new  2781828   initial changes for python3 pkgs
       new  82ea224   Add multilabel support for dump_svmlight_file.
       new  2d4cd28   Merge pull request #4958 from rasbt/treeregressor
       new  193b3c8   Merge pull request #4966 from lazywei/multilabel-dump-svmlight
       new  44c3e64   Merge pull request #4937 from pianomania/FeatureHasher_add_example
       new  1f3ebb0   DOC FeatureHasher takes (finite) numbers as values
       new  d389142   Merge pull request #4956 from nealchau/ransacresid
       new  0156dcf   BUG Use epsilon threshold in `_samme_proba` and `_boost_real`
       new  ef5312c   Added list of multi-label supporting classifiers.
       new  8638862   Merge pull request #4915 from untom/minmax_scale
       new  a761858   adding assert statement to check the std
       new  171f933   Merge pull request #4971 from dotsdl/issue-4724
       new  ec78cc1   whitespace
       new  0af6bae   add whatsnew entry for fix of penalty passing in LogisticRegressionCV
       new  8932587   FIX dump_svmlight_file multilabel arg: should be last
       new  bc5155f   Fixed small typos in SVM examples.
       new  39dbe73   Fix #4978: Typo in CONTRIBUTING.md
       new  ee96253   Merge pull request #4979 from donnemartin/master
       new  7df3c23   Merge pull request #4977 from edsonduarte1990/master
       new  bda486d   Generated C file
       new  26b3f86   MAINT ignore Python import state in ctags
       new  605cc79   MAINT make it possible to use plain HTTP to download pre-built numpy & scipy from rackspace container
       new  792e5295  add fetch_rcv1
       new  2b87a20   Merge pull request #4983 from ogrisel/appveyor-http-rackspace
       new  f8b3473   Merge pull request #4908 from TomDLT/rcv1
       new  81de81a   ENH implement LYRL2004 train/test split of rcv1
       new  abb31d0   Merge pull request #4984 from TomDLT/rcv1-subset
       new  eca4201   DOC updated whatsnew for joblib 0.9.0b2
       new  4243412   fixes bug in oob_score when X is sparse.csc matrix [refs #4744]
       new  acb67bd   Bug fix for unnormalized laplacian
       new  2f81103   Merge pull request #4972 from ogrisel/doc-whatsnew-joblib-0.9.0b2
       new  8de86ed   fixed newline errors in documentation for SVC
       new  bd43b8e   fixed newline errors in documentation for NuSVC
       new  c1c4548   fixed newline errors in documentation for LinearSVC and LinearSVR
       new  df0dadb   Merge pull request #5001 from hlin117/svm-documentation
       new  03b09a4   Merge pull request #4954 from ankurankan/fix/4744
       new  0e87eb8   remove links to old docs
       new  9593631   Merge pull request #4933 from banilo/regr_score_docs
       new  043dc9f   - test for unnormalized Laplacian computation
       new  8b6844a   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn
       new  176f4cf   Documentation added to criterion, dense splitters.
       new  cb226d5   Reorder target_names in rcv1
       new  c151f6e   Merge pull request #5011 from TomDLT/rcv1_order
       new  840bcf3   Merge pull request #4961 from mrphilroth/issue4959
       new  a989fb6   correct comments
       new  52ed08c   Update joblib to 0.9.0b3
       new  b86af8e   DOC fix README.rst coveralls badge
       new  9f46531   Filled in missing negation
       new  73bebfe   Merge pull request #5018 from ltiao/patch-1
       new  34c4930   FIX: ensure that get_params returns VotingClassifier own params
       new  71feb53   FIX seed the initialization of NMF
       new  e250fae   Merge pull request #5020 from ogrisel/fix-nmf-rng-seeding
       new  6fedcab   Verbosity reduced, suggestions merged
       new  69994c0   Responses changed to targets
       new  7d0c734   Responses changed to targets
       new  a91aac0   PEP257 adhered to closer
       new  6c09af5   Merge pull request #5017 from lesteve/fix-coveralls-badge
       new  4c68a2d   MAINT make copy_joblib.sh Python 3 compatible
       new  552a80a   Merge pull request #5021 from lesteve/make-copy-joblib-sh-python3-compatible
       new  72eda98   Merge pull request #5005 from amueller/drop_old_doc_links
       new  15fad67   TST Increase samples for classifier tests
       new  f9e0736   Adding a new kmeans example.
       new  77d3fd9   Merge pull request #4967 from stephen-hoover/threshold-boosting-samme_proba
       new  a45a50e   Merge pull request #5019 from glouppe/voting-get_params
       new  f73fe10   DOC: Remove extra "not". [ci skip]
       new  5fa4d63   Merge pull request #5025 from jakirkham/fix_typo
       new  b52054f   Merge pull request #4957 from dotsdl/issue-4614
       new  0d4770f   Adding kmeans example to the narrative documentation
       new  253ae0e   Adding a reference to the example in the narrative documentation
       new  0b83c98   DOC tweaks for feature_extraction.text
       new  24e962c   cosmit in plot_cv_predict
       new  e2f5665   Make more clear that adding algorithms is not the preferred way to start contributing.
       new  f8985f2   Merge pull request #5038 from jnothman/textdoc
       new  7fd2206   Merge pull request #4960 from mkrump/lb_encoder
       new  e91af4c   Merge pull request #5042 from amueller/contributing_addition
       new  c3b56ac   fix shape of thetaL, thetaU, theta0 in gp
       new  2186d93   Update lfw.py
       new  f9ae0f8   Merge pull request #5048 from marktab/fixlfw
       new  0bcbf92   Merge pull request #4965 from mrphilroth/4922
       new  27f14d0   lars_path method complemented with nonnegative option.
       new  30225aa   lars_path method complemented with nonnegative option.
       new  ae1fba0   Merge branch 'nonnegative-lars' of https://github.com/michigraber/scikit-learn into nonnegative-lars
       new  9dafcb1   Merge pull request #5016 from lesteve/update-joblib-to-0.9.0b3
       new  b802ca2   ENH: add Cohen's kappa score to metrics
       new  3a135f5   Final round of revisions
       new  df14ef9   Rewrite of the documentation for LDA/QDA
       new  8b7fa9e   Small edits to LDA/QDA documentation
       new  9207639   DOC fixed short underline
       new  d857349   Merge pull request #5010 from jmschrei/_tree_doc
       new  25f8939   DOC: fixes to _tree.pyx
       new  cec3bf9   FIX assign_rows_csr: should not zero the entire out array
       new  fead69a   Bugfix : type in cd changed for read only memmap compatibility
       new  2afbc0f   FIX make sure y is 1d in svm regressor, add common test for regressor shapes.
       new  99eff75   fix TheilSenRegressor input validation, LassoLarsIC input validation
       new  35927df   added handling of 2ndim y in regressors to whatsnew.
       new  3a6c4dc   Merge pull request #5057 from amueller/fix_2d_y_svm_regressor
       new  77ecf16   Merge pull request #4775 from arthurmensch/cd_fast_readonly_array_brainfix
       new  ef3a404   Merge pull request #5052 from larsmans/kappa
       new  9127fac   FIX commit re-cythonized sparsefuncs_fast
       new  598665e   TST common metric tests for cohen_kappa_score
       new  9a0ba8c   Merge pull request #5049 from amueller/gp_reshape_fixes
       new  6bae1b9   force y to be numeric in kernel ridge.
       new  1fe83d1   Merge pull request #5047 from amueller/scipy_0.16_fixes
       new  ab7b42f   TST Fix: made the test for y 1d conversion warning more robust.
       new  2d9388f   Merge pull request #5072 from amueller/fix_fun_conversion_warning
       new  489db27   Add more extensions and related packages to related_projects.rst
       new  e324fe4   Add some newline characters to respect the 80 characters rule
       new  dd08564   Fixed minor typo
       new  cd9eb96   FIX import column_or_1d from utils.validation directly
       new  16337b2   Merge pull request #5077 from rvraghav93/fix_import_coor_des
       new  cb0916c   ENH: Adds CallableTransformer
       new  9d14f73   ENH: Renames CallableTransformer -> FunctionTransformer.
       new  ba44e7b   COMPAT: Makes test_function_transformer py2 compatible.
       new  bf8b91f   DOC expand FunctionTransform docstring
       new  662f972   DOC added whatsnew for k-means fix
       new  a2a2d83   FIX be robust to obscure callables that are false.
       new  3c6cf99   Merge pull request #5059 from amueller/function_transformer_rebase
       new  b09afb0   DOC/MAINT final touches to FunctionTransformer
       new  13c241f   Merge pull request #5076 from PGryllos/update_doc
       new  32f2e8e   Add common test that transformers don't change n_samples.
       new  208d194   Added several related projects.
       new  1aa78d1   Added a link to the deep learning software list.
       new  a8970fc   Added Testcase for LDA-Covariance
       new  1d98ca3   Fixed shrinkage covariance for LDA
       new  7befb5d   Explicit broadcasting
       new  c7c76b2   Updated whats_new.rst
       new  7a3a987   Fix memory access error in OneClassSVM
       new  9fd8149   Update whats_new.rst
       new  617d28d   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into patch-1
       new  c4dae2e   - removed incorrectly commited file
       new  e5934cf   Merge pull request #5092 from kazemakase/fix-lda-covariance
       new  d04d7cf   TST: fix undefined behavior in test
       new  250e509   ENH O(1) stop-word lookup when list provided
       new  44fd56b   Merge pull request #5093 from amueller/one_class_svm_memory_fix
       new  78bd87b   OneClassSVM sparsity regression test added
       new  7fca5a1   COSMIT prefer loops to repetition in LSHForest benchmark
       new  a6c6e73   Merge pull request #3659 from chyikwei/onlineldavb
       new  28beb84   COSMIT pep8 online LDA
       new  1a45b79   typos
       new  0586070   chmod -x
       new  d341dac   Remove shebang and utf-8 header.
       new  664d78e   MAINT Remove support for the deprecated sequence of sequences
       new  b0e44c3   MAINT return 'unknown' for seq. of seqs in type_of_target
       new  94891f0   MAINT deprecate seq of seq in metrics fully with tests
       new  9d96753   FIX bailout w/ValueError if multilabel-seq or other unsupported types
       new  72f6225   MAINT/ENH inline the logic for _is_sequence_of_sequences
       new  3bd6132   FIX make sure empty target vectors are returned as binary
       new  8f1ad44   FIX 2D binary array-like must be considered as unknown
       new  714eef2   TST Explicitly add one empty label
       new  7def0a7   TST/MAINT reintroduce mixed types test sans seq. of seq.
       new  c031b39   MAINT is_label_indicator_matrix --> is_multilabel
       new  6fb17a1   FIX revert generation of seq of seq in make_multilabel_classification
       new  7e828db   MAINT remove unused import warnings
       new  18221db   FIX type_of_target will now raise ValueError on ml-seq; [[1, 2]] is mc-mo;
       new  1da4679   TST (in metrics) change error message for seq. of seqs
       new  103f504   DOC revert the documentation changes
       new  734bead   MAINT Use check_is_fitted
       new  dd2cb65   FIX/REVERT undo removal of ignore_warnings to avoid a few warnings in tests
       new  5005d38   MAINT Remove ignore_warnings from test_matthews_corrcoef_nan
       new  89ec72e   Example to demonstrate use of tree.apply() method
       new  e4df618   ENH micro-optimize LatentDirichletAllocation + cosmetics
       new  5aab0af   added a string for FriedmanMSE (instead impurity) when exporting a dot file
       new  a08482a   missed elif
       new  4c2d61a   Merge pull request #4919 from rvraghav93/label_binarizer
       new  11cd969   fixing a typo in feature extraction documentation
       new  28452b4   Merge pull request #5117 from kylerbrown/patch-1
       new  a88f6ec   add sample_weight into LinearRegression
       new  088ee0e   deleted useless assert
       new  50e5031   Merge pull request #5074 from jdwittenauer/master
       new  e287760   DOC: happy new year!
       new  4036dd0   DOC: typo in function name
       new  4ac6a90   ENH: split LDA's n_iter_ into n_iter_ and n_batch_iter_
       new  26e47d7   remove dirichlet_component_ variable
       new  1084047   now uses isinstance and keeps original name (FriedmanMSE)
       new  03e95ef   added a test and reverted string to friedman_mse
       new  1360118   corrected test: criterion name has to be only in those nodes where there is "samples"
       new  858ce4c   took if clause into regex
       new  a7081d7   [issue #5043] fix documentation of callable kernel
       new  6d78d3d   implemented improvement suggestions
       new  edf077c   Merge pull request #5050 from AnishShah/issue5043
       new  f4fa782   Merge pull request #5111 from chyikwei/remove-unnecessary-variable
       new  5908bcc   TST: LatentDirichletAllocation behavior on empty documents
       new  ea8d092   Explicitly mention RandomTreesEmbedding in the text
       new  a946890   ENH: faster LatentDirichletAllocation (~15% on 20news)
       new  d0ce632   added newline (for pep8) and reverted to regex solution due to 0.0/-0.0 problem on windows
       new  a8976e7   fixes
       new  749f2a9   DOC n_thresholds may be < no. of unique scores
       new  25a217b   small typo fix "radius" in rbf kernel
       new  d75f8cc   "nonnegative" -> "positive" + cleanup
       new  8e3b22c   lars_path method complemented with nonnegative option.
       new  e50aaba   "nonnegative" -> "positive" + cleanup
       new  7622e0a   Merge branch 'nonnegative-lars' of https://github.com/michigraber/scikit-learn into nonnegative-lars
       new  ceb066d   cleanup
       new  d6908a9   todos for pull request
       new  b3694b1   Merge pull request #5126 from rasbt/rbf
       new  9db5dee   Merge pull request #5113 from fzalkow/master
       new  e3f46cb   docstrings added + fit_intercept reintroduced.
       new  b2cea17   cleanup
       new  7f1bf0d   positivity constrained can now also be used for the CV estimators.
       new  c369af4   doctest fix.
       new  6b22d97   positive option for Lars propagated to doc.
       new  89ddbe2   Merge pull request #5120 from larsmans/faster-lda
       new  2ed0c22   Added check_input in Lasso for DictLearning performance
       new  cf2a875   FIX: Fix for changing mode
       new  393d651   Merge pull request #5133 from arthurmensch/lasso_perf_checks
       new  ef66046   Merge pull request #5094 from pv/fix-inplace
       new  af587ec   Removed unnecessary class checks
       new  39bad0a   Merge pull request #5143 from Eric89GXL/fix-r2
       new  0015659   Merge pull request #5081 from amueller/transformers_consistent_n_samples
       new  f38aea4   Fixes #3594.
       new  11d33bc   Merge pull request #5071 from mth4saurabh/fix-issue-3594
       new  617baa1   Style fix
       new  6c1c6ec   agramfort code review input
       new  7d4e8e1   test lars_path positive constraint.
       new  535536e   merge with upstream/master
       new  7e16f9d   fit_intercept docstring comment removed from lars_path method
       new  f157606   positive option passed on to LarsCV estimator
       new  bbbca02   tests of positive option for estimator classes
       new  38451a3   tests of positive option for estimator classes condensed.
       new  7224c31   DOC rearrange related projects
       new  db1f67b   pep-8 fix
       new  8a3d97d   DOC mention intercept_ attribute in ridge docstring.
       new  98fc670   Merge pull request #5157 from amueller/ridge_intercept_docs
       new  aa2d6bf   Add check for sparse prediction in cross_val_predict
       new  48adf8b   Use a single vstack for concatenating all blocks in prediction matrix
       new  98fefd1   tests for comparison of results for Lasso and LarsLasso added.
       new  e35c90c   Use Inverted locations to reorder the predictions
       new  26d3323   Remove redundant p variable
       new  0c6b173   Add test to check sparse predictions in cross_val_predict
       new  b68f922   Fix minor indentation issues
       new  6248745   fix #5139
       new  434ee95   test for accepted sparse matrices
       new  0706636   Merge pull request #5063 from amueller/bagging_input_validation
       new  3fc4ff3   Merge pull request #5098 from olologin/OneClassSvm_sparse_test
       new  de48574   DOC FAQ on loading data as numpy arrays
       new  254b815   DOC explain fork related issues in FAQ + stopgap for Python 3.4
       new  42b9590   Merge pull request #5167 from NoonienSoong/clustering_example
       new  9486a7a   Merge pull request #5161 from beepee14/sparse_prediction_check
       new  e62e9e1   DOC add what's new entry for cross_val_predict fix
       new  ea42c55   PEP8 in cross_validation module + tests
       new  2ba124b   DOC use correct attribute in example
       new  53324a1   docstrings updated with considerations regarding small alphas when positive=True.
       new  d3dec94   comparison test of Lasso and LassoLars under positive restriction refactored and commented
       new  4170473   Merge pull request #5131 from michigraber/nonnegative-lars
       new  a8585a4   pep8 + simplify positive test
       new  c3cfebe   [BUG] predict_proba should use the softmax function in the multinomial case
       new  74d4334   DOC don't show example thumbnails in latex docs
       new  e2f3aab   DOC Updated documentation for cv parameter
       new  bcb6d3b   DOC: minor cosmetric to ROC example
       new  c779569   Merge branch 'master' into pr_3651
       new  8bc47f2   ENH exposing extra parameters in t-sne
       new  21a966a   ENH: added LabelShuffleSplit cv iterator
       new  337a6ba   DOC: add cross_validation.LabelShuffleSplit to classes.rst
       new  fb58d37   Merge branch 'master' into pr_4009
       new  388e596   Merge branch 'master' into pr_4009
       new  764826c   Merge pull request #5162 from ogrisel/faq-freeze
       new  42229fd   Merge pull request #5187 from GaelVaroquaux/pr_4009
       new  7ce32d7   Merge pull request #4253 from Flimm/doc-precision_recall_fscore_support
       new  9ec7ecd   DOC: update whats_new
       new  eb6b654   Merge pull request #5186 from sdvillal/issue5165
       new  919fc49   ENH precomputed is now a valid metric for 'brute'
       new  5867ef4   ENH precomputed is now a valid metric for 'auto' neighbors
       new  3aa4fc2   ENH support precomputed neighbors where query != index
       new  2245790   TST precomputed matrix validation in correct place
       new  f4fc275   FIX add _pairwise and test to NeighborsBase
       new  6daa0f8   DOC fix up parameter descriptions
       new  6c750ba   FIX broken rebase
       new  066a134   TST add test for precomputed metric and X=None at predict
       new  847e366   FIX test with invalid input; simplify dbscan precomputed
       new  c45b863   TST: Use a local random state
       new  aaae2db   make_multilabel_classification sparse target
       new  3e23cf0   Merge pull request #4104 from maheshakya/johnson-lindenstrauss_bound_example_fix
       new  242aaca   Merge pull request #4312 from xuewei4d/friendly_error_on_kneighbors
       new  71c99dc   FIX error introduced during rebase
       new  c4721f3   TST: No n_jobs=-1 in the test
       new  a5537aa   override predict_proba in log_reg
       new  ab8f155   Merge pull request #5188 from GaelVaroquaux/pr_4090
       new  869971f   Merge pull request #5184 from christophebourguignat/master
       new  6e735f5   Merge pull request #4767 from trevorstephens/passive-aggressive_cw
       new  55c32ef   Merge pull request #4779 from martinosorb/parallel-ms
       new  190abde   Merge pull request #4881 from sonnyhu/weighted_least_squares
       new  c85f2ad   Add non regression test
       new  0bf7536   Merge pull request #5189 from GaelVaroquaux/pr_4779
       new  8d8b434   support X_norm_squared in euclidean_distances
       new  4f713ce   Merge pull request #5182 from MechCoder/predict_proba_fix
       new  96c329f   Merge pull request #5037 from betatim/tree-feature-transform
       new  6126ece   unused import
       new  1991c43   DOC Updated documentation for cv parameter (issue #4533)
       new  53c3530   DOC Updated documentation for cv parameter (issue #4533)
       new  343f9b4   ENH sparse precomputed distance matrix in DBSCAN
       new  3393b07   Merge pull request #5174 from jnothman/dbscan_precomputed_sparse
       new  7b5eca2   DOC what's new: DBSCAN sparse precomputed
       new  201aca9   DOC tweaks to what's new
       new  4d39cf8   Merge pull request #2459 from dougalsutherland/euclidean_distances
       new  364825a   Merge pull request #5194 from christophebourguignat/master
       new  0bbdfe4   DOC resize images for latex
       new  a1ad90e   DOC hide web-specific links from latex build
       new  3bcd772   DOC prefix version headings by "Version"
       new  e19b700   DOC provide PDF documentation for download
       new  db0cbaf   MAINT bump joblib to 0.9.0b4 to use forkserver for Py3 & POSIX
       new  643e7cb   DOC update FAQ on multiprocessing forkserver
       new  150267f   added comments for y in dump_svmlight_file() method
       new  b10e78e   [DOC] Precision on random_state in KFold() doc
       new  e2e5fbd   Merge pull request #5201 from christophebourguignat/master
       new  9d01a1c   changed shape parameter
       new  4b320ce   Added check to ensure that NumPy and SciPy meet minimum version requirements (NumPy >= 1.6.1, SciPy >= 0.9)
       new  6248fee   Amended error message to remain valid when NumPy/SciPy are installed but out of date
       new  66b0e13   Added reference to SciPy/NumPy min_version vars in ImportError
       new  7496e42   Made setup.py pep8 compliant
       new  1529a1e   MAINT enable multiprocessing + kmeans test on Python 3.4
       new  d2b668c   DOC whats_new entry for forkserver
       new  5a1f183   Creation of the attribute LDA.explained_variance_ratio_, for the eigen solver. It is an analogy to PCA.explained_variance_ratio, and works in the same way.
       new  3ad9972   Addition of a piece of docstring for the attribute explained_variance_ratio_. Creation of the function test_lda_explained_variance_ratio to test this new attribute.
       new  099b483   Correction of the name of a function.
       new  cf4bc71   Changing [ to (. Formatting consistency.
       new  6e7e9d3   Modification of a line after @agramfort's suggestion.
       new  cdaad2d   Correct expectation notation in DP-GMM doc
       new  bb2aced   AdaBoostRegressor should not raise errors if the base_estimator does not support sample_weights
       new  b729cf2   DOC update AUTHORS.rst to better reflect team
       new  d39716e   Optimize MSE criterion
       new  6eb4028   Added subject independent KFold
       new  cf7bc7b   ENH: rename to LabelKFold
       new  c769a65   COSMIT: variable names, documentation, etc
       new  251fecc   ENH Faster tree-based methods by implementing reverse update of criterion
       new  e648fb9   Added the option of passing in a sparse X matrix into decision function, plus tests for sparse for all prediction functions.
       new  410e7f2   ENH: remove unnecessary assignments
       new  59a1bdd   Merge pull request #5220 from arjoly/fast-reverse-update
       new  44b1f3a   Merge pull request #5190 from glouppe/labelkfold
       new  4f6cd63   Changed predict_log_proba so that it accepts sparse matrices.
       new  2da8623   Merge pull request #5223 from mattilyra/allow_sparse_in_baggingclassifier
       new  709e6a6   Merge pull request #5195 from jnothman/example_latex
       new  6f696ca   Merge pull request #5218 from mooz/patch-1
       new  f08963b   Add numerically stable softmax function to utils.extmath
       new  d81b8ed   Add test for softmax function
       new  08bf350   Merge pull request #5199 from ogrisel/joblib-0.9.0b4
       new  7fccf45   Merge pull request #5207 from MechCoder/ensemble_sample_weight_bug
       new  5978c0b   ENH split _tree.pyx into several files
       new  ddbf727   DOC describe constitution of AUTHORS.rst list
       new  4955a64   DOC sort AUTHORS.rst by surname
       new  77c963d   Merge pull request #5230 from jmschrei/_tree_split
       new  2f09933   MAINT deprecate 1d input arrays for all estimators
       new  cb591e1   Merge pull request #5225 from MechCoder/fix_overflow
       new  3f9d33c   Merge pull request #5200 from jackzhang84/add_dump_svmlight_file_comment
       new  882c346   DOC Make cv documentation consistent across our codebase
       new  250507f   FIX deprecation message for 1d data
       new  bc225a5   ENH apply method added to Gradient Boosting
       new  470b9a4   Merge pull request #5228 from jmschrei/gb_apply
       new  994cbe1   FIX unstable test due to bootstrap and unset random state
       new  6942043   Updated whatsnew for sprint and LatentDirichletAllocation, add some missing author links
       new  4ceffe0   Adding Implementation of SAG
       new  94eb619   ENH add sag solver in LogisticRegression and Ridge
       new  cda140e   fix rebase mistake
       new  0c1afd7   [BUG] _init_centroids has an optional x_squared_norms parameter which is not exactly optional
       new  1fdfd23   Added output of out-of-date SciPy/NumPy versions to setup.py, if detected
       new  7e2453a   Merge pull request #5242 from arjoly/stable-test-2
       new  9ebc6fe   Merge pull request #5216 from JPFrancoia/master
       new  679936f   MAINT Deprecate LDA/QDA in favor of expanded names
       new  91de280   MAINT re-cythonize sklearn/tree/*.pyx
       new  3b30958   MAINT PEP8 / slightly faster tests in decomp LDA
       new  75d65dc   Merge pull request #5249 from MechCoder/init_centroids_bug
       new  62ad838   OPTIM make (Online) LDA reuse a joblib.Parallel instance
       new  84df9e9   Merge pull request #5206 from acganesh/numpy_scipy_ver_check
       new  fe09a07   FEAT Barnes-Hut t-SNE
       new  1ce9ba3   add necessary blas files.
       new  fe49161   Minor fixes, remove transform for now.
       new  6dd6f8f   minor fixes to the doc build
       new  25dbb15   TEST: stronger tests for variable importances
       new  78974de   TEST: use sklearn.fixes.bincount
       new  bcc6f1b   TEST: take comments into account
       new  5810567   Fix check in `compute_class_weight`.
       new  f01a695   Fix assertion for Python 3, use `assert_raise_message`.
       new  698b670   check_X_y: Added mention that function returns y array
       new  99f7354   Merge pull request #5263 from hlin117/check_X_y-docs
       new  4a6f96f   Merge pull request #5258 from amueller/docfixes
       new  9ec03c2   Change formatting of `assert_raise_message`.
       new  af56154   Merge pull request #5238 from rvraghav93/cvdoc
       new  5f589fb   TEST: reduce test time, variable name, etc
       new  8424c48   Merge pull request #5262 from andylamb/lamb-fix-class-weight-check
       new  3575db6   TEST: check parallel computation
       new  b099a59   Merge pull request #5261 from glouppe/check-importances
       new  0bc3a6f   Merge pull request #5245 from ogrisel/lda-acronym-deprecation
       new  0de583a   Fixed LDA typo in doc
       new  5aaf6ec   Merge pull request #5271 from vortex-ape/lda-typo
       new  1a15fdc   Added handling of sample weights in logistic.py * Updated _check_solver_option to include sample_weight check * Updated all calls to _check_solver_option() * Updated documentation of class_weight throughout logistic.py * Added sample_weight parameter to logistic_regression_path. * Added handling of sample weights to logistic_regression_path. * Added sample_weight parameter to _log_reg_scoring_path. * Added handling of sample weights to _log_reg_scoring_path. * Added  [...]
       new  942ff5c   Updated logistic regression tests with sag solver
       new  ebd5dbc   Fixed syntax and combined two test functions
       new  fac33ad   Adding checks for the input LDA prior
       new  628ed2c   ENH remove some warnings in test suit
       new  fecd355   FIX array_equal for numpy < 1.8.1
       new  b9e79bd   Go to easy _open_ issues page
       new  fbdb727   Merge pull request #5284 from tdhopper/patch-1
       new  8706e9e   Merge pull request #5280 from ogrisel/pr-4924-rebased
       new  4af0c49   vocabulary of type set now coerced to list to preserve iteration ordering after serialization
       new  4366ba5   use sklearn.utils.random
       new  0a7a1db   removed list around sorted
       new  8f0d4fc   ENH avoid slow CSR fancy indexing variant
       new  dae50f5   Updated link. Original link to ZhangIJCV06 paper "Local Features and Kernels for Classification of Texture and Object Categories: A Comprehensive Study"  was dead. Replaced all occurances of "http://eprints.pascal-network.org/archive/00002309/01/Zhang06-IJCV.pdf" with "http://research.microsoft.com/en-us/um/people/manik/projects/trade-off/papers/ZhangIJCV06.pdf"
       new  9ea9b34   Merge pull request #5289 from carrillo/master
       new  37bd275   Merge pull request #5283 from TomDLT/remove_warnings
       new  770d089   Merge pull request #4887 from amueller/tsne_fixes
       new  a24b932   DOC comment on obscure syntax
       new  b9284a7   Merge pull request #5288 from jnothman/bicluster-example-speed
       new  ceeef70   ENH refactor NMF and add CD solver
       new  db52aac   Merge pull request #4852 from TomDLT/nmf
       new  a0ce9e4   add future warning to pipeline.inverse_transform with 1d X.
       new  442b49a   ENH criterion file cleaned up
       new  95fe122   Merge pull request #5278 from jmschrei/_criterion_cleanup
       new  aa60d98   DOC: better docstring for sum_total
       new  c521783   Merge pull request #5293 from amueller/pipeline_X1d_inverse
       new  5271c04   deprecate fit params in qda and lda
       new  2a5652a   Merge pull request #5294 from trevorstephens/lda_qda_fit_params
       new  6966881   MAINT use inspect.signature for introspection
       new  3c6ef92   Merge pull request #5253 from ogrisel/online-lda-with-parallel-context-manager
       new  08460eb   ENH: optimize LDA, ~15% faster on one core
       new  218e255   Merge pull request #5301 from ogrisel/python-3.5-inspect-deprecation
       new  798aeaf   FIX criterion variable names
       new  c82ad6d   Merge pull request #5305 from jmschrei/criterion_patch
       new  ac7373a   Merge pull request #4955 from copyconstructor/preserve-vocab-ordering
       new  8fcd063   DOC: TODO notice in Hungarian algorithm, use SciPy version when released
       new  1f4bc26   FIX: Ensure correct LabelKFold folds when shuffle=True
       new  468455c   Added link to github repositories in what's new.
       new  8959375   Merge pull request #5310 from JeanKossaifi/add_repo
       new  7881ab5   Dataiku testimonial
       new  86b1ba7   Merge pull request #5311 from jereze/master
       new  46c0e85   #5309: Added verbosity argument to sparse_encode, verbosity for LassoLars and Lasso
       new  47ece78   added k-modes clustering project
       new  a00af6d   MAINT build and test with Python 3.5 on appveyor
       new  de3934f   MAINT update numpy / scipy wheels used by appveyor
       new  d28a12e   Merge pull request #5303 from larsmans/faster-lda
       new  2556ae0   Merge pull request #5300 from glouppe/fix-5292
       new  f6fe634   DOC: optimize Dataiku logo (16kB => 10kB)
       new  02c0029   ENH gbt sparse support
       new  2c758f3   Merge pull request #5252 from jmschrei/presort
       new  d70161e   MAINT configure the appveyor's cache for pip
       new  c9d8e91   Merge pull request #5302 from ogrisel/python-3.5-appveyor
       new  509bc04   Merge pull request #5317 from hlin117/verbose-sparse-encode
       new  e7ecc4f   BUG shuffle components, not samples in NMF CD
       new  47a85f3   MAINT: better timing and general fixes in NMF/LDA example
       new  f1c5924   ENH better error message for estimators with ensure_min_* checks
       new  a2d638b   Merge pull request #5334 from ogrisel/better-ensure-min-error-message
       new  c8406fb   Breast cancer dataset added
       new  46edeec   COSMIT some refactoring in NMF
       new  f8486a5   fixes scikit-learn/scikit-learn#5329
       new  9de7815   Merge pull request #5332 from larsmans/nmf-cleanup
       new  f5138ad   remove numpy's RuntimeWarning from corner case of PCA.fit
       new  c4687fd   Merge pull request #5335 from rrohan/current
       new  e7c50de   misc + pep8
       new  808fa64   FIX consistency of memory layout for linear CD solver
       new  c3a3e69   FIX ensure contiguous Gram matrix in graph lasso
       new  1025982   Merge pull request #5337 from ogrisel/fix-coordinate-descent-memory-layout
       new  a52053d   Fix sparse_encode input checks
       new  c438f78   Implemented SelectFromModel meta-transformer
       new  8a28ea8   fix test failures
       new  459cb9b   Remove warm start
       new  2416e2a   Catch filters instead of removing the tests
       new  9cee0d9   Added example to depict feature selction using SelectFromModel and Lasso
       new  3d41053   Minor doc changes and removed _set_threshold and _set_importances
       new  10176d9   Now a fitted estimator can be passed to SelectFromModel
       new  2ee718c   Add narrative docs and fix examples
       new  acf5f16   Merge SelectFromModel and L1-selection examples Add test to check the threshold can be set without refitting
       new  919968a   Fixed typo in the final ".. note::" of naive_bayes.rst
       new  b316dca   Merge pull request #5346 from Naereen/patch-1
       new  af4a2ca   DOC: Fix typo in CalibratedClassifierCV
       new  4d0861a   Merge pull request #5347 from ariddell/patch-3
       new  687af98   added predict_equal_labels test and changed kmeans single
       new  fc18756   moved condition out of the loop
       new  ddc9852   Added check for max_iter and test
       new  f4e1658   Added explanatory comment
       new  e8d2b93   Merge pull request #5231 from vighneshbirodkar/kmeans_fix
       new  74ebef5   MAINT: make GaussianNB scale-invariant
       new  83fdaa3   TST: GaussianNB scale invariance
       new  e2e33aa   GaussianNB: use var() rather than std() in epsilon determination
       new  176427e   Fix SGD partial_fit multiclass w/ average.
       new  e3b9e0b   Added entry to `doc/whats_new.rst`.
       new  03eff3a   Merge pull request #5336 from giorgiop/pca-warning
       new  69831da   ENH utility to have distinct dataset .pkl filenames
       new  1a1fe10   FIX separate filenames for RCV1
       new  b44ba51   FIX separate filenames for species distributions
       new  60eb28d   FIX separate filenames for covtype
       new  890147f   FIX separate filenames for 20 newsgroups
       new  b17bdf7   FIX separate filenames for california housing
       new  8d058d3   FIX separate filenames for Olivetti faces
       new  9e337b1   DOC what's new entry for dataset fetchers
       new  eac920d   MAINT: minor fixes to GaussianNB epsilon
       new  bb9e725   DOC: add whats_new entry for PR #5349
       new  bdc39e7   DOC FIX missing minus in Shannon's entropy formula
       new  86be478   FIX Use float64 to avoid spurious errors
       new  1d5b473   Lasso and ElasticNet should handle non-float dtypes for fit_intercept=False
       new  37c6717   Missing link for inertia (about K-Means)
       new  d77ea12   Merge pull request #4707 from amueller/k_means_init_mismatch
       new  24e6d92   Implement dropping of suboptimal thresholds in roc_curve
       new  42cc2fa   Merge pull request #5237 from gclenaghan/remove_thresholds
       new  9ff9ae7   fix merge conflict
       new  99b794c   Merge pull request #5363 from Naereen/patch-1
       new  ba3caf4   Merge pull request #5356 from rvraghav93/ridge_appveyor_failure
       new  e103c31   Fix few typos on links and doi
       new  386ecd4   Merge pull request #5365 from Naereen/patch-2
       new  60b41a8   correct optional arguments for FeatureHasher
       new  735dd1f   Fix warnings during tests
       new  f60ba72   Increase length of array indexing type in ArrayDataset
       new  8b1d37d   Merge pull request #5369 from joshloyal/fix_hashing_docstring
       new  5a0db17   1. Added parameter prefit to pass in a fitted estimator. 2. Use assert_warns instead of catch_warnings 3. Remove depracation warnings in common tests.
       new  f0121b7   removed reshaping in fast_mcd and replaced by check_array
       new  c857e1e   Merge pull request #5349 from jakevdp/naive-bayes-scale
       new  84e9e10   Merge pull request #5371 from ylow/sgd_bug_fix
       new  c805fbc   Refactor tests
       new  652b950   Merge pull request #4242 from MechCoder/select_from_model
       new  0fec61a   Changed the number of features from 10 to variable
       new  3c22232   Now plots based on the number of features in X
       new  59c0cb7   FIX mailmap
       new  f0a162d   Merge pull request #5386 from jmschrei/contrib
       new  281aebd   Merge pull request #5384 from map222/patch-1
       new  da9a7cd   Merge pull request #5362 from MechCoder/lasso_fix
       new  8d273a1   Merge pull request #5234 from vighneshbirodkar/mcd_fix
       new  46ede61   Add link to ETL.
       new  8e5b375   DOC trying to make GridSearchCV docs more accurate... not sure if better
       new  d972da2   FIX temporary fix for sparse ridge with intercept fitting
       new  db49cdd   DOC update docstring and warning
       new  6a5a2f7   partial_fit for scalers
       new  7fe902c   Merge pull request #5378 from amueller/gridsearch_docs_unsupervised
       new  90922ea   Merge pull request #5360 from TomDLT/sparseridge
       new  838570c   Fixes #4455.
       new  9564b94   Merge pull request #5104 from giorgiop/scaler-partial-fit
       new  c3625cb   DOC wrong default value in docstring
       new  1af69aa   Merge pull request #5400 from kwgoodman/doctypo
       new  e5d3cbc   MAINT update joblib to 0.9.2
       new  a601526   Added laplacian kernel
       new  276d928   DOC formatting fixes for laplacian kernel
       new  85223b9   a common test to check if classifiers fail when fed regression targets
       new  31a5a5f   Fixed warnings for DataDimensionalityWarning, decision_function and decision_function_shape.
       new  daf41d0   FIX non deterministic LSHForest doctest failure
       new  6db0037   DOC add link to the joblib 0.9.2 release notes
       new  6fe94b5   Merge pull request #4478 from amueller/fix_randomized_svd_transpose
       new  16fcb15   ENH use explicit decimal keyword & PEP8
       new  4e2f20b   DOC whats_new entry for randomized_svd heuristic
       new  1070e1b   Merge pull request #5395 from amueller/some_test_deprecation_warnings
       new  e777fa7   DOC: correct docstring
       new  02b9a72   DOC Fix plot_tomography_l1_reconstruction example
       new  d52e2a1   Merge pull request #5410 from GaelVaroquaux/fix_doc_gcv
       new  4f4c359   FIX dtypes to conform to the stricter type cast rules of numpy 1.10
       new  affa122   MAINT Update the 3rd travis build to latest versions of numpy and scipy
       new  21fab6e   FIX Move validation from helper to main function
       new  cac1d28   FIX AdaBoostRegressort test failure with numpy 1.10
       new  47f19e9   MAINT add safe_{median|mean} for np 1.10.1
       new  1b9e791   Merge pull request #5398 from rvraghav93/fix_for_numpy_10
       new  32ab5b5   DOC fix missing import in plot_lle_digits example
       new  218bd82   Merge pull request #5409 from lesteve/fix-plot-lle-digits
       new  79fa25a   MAINT don't print things in testing.
       new  67ab084   Merge pull request #5399 from lesteve/update-joblib-to-0.9
       new  c66689c   FIX Don't compare arrays to strings!!!!
       new  b1619d0   DOC : update docstring of l1_ratio for E-Net CV classes
       new  2d97609   Merge pull request #5402 from lesteve/fix-lshforest-doctest
       new  3301893   Merge pull request #5413 from amueller/cleanup_tests
       new  a870112   FIX precision to float64 across the codebase
       new  bd66a27   proposed fix for #5387
       new  c0c2c73   Merge pull request #5411 from lesteve/fix-plot-tomography-l1-reconstruction-example
       new  2300bdd   MAINT switching master to 0.18.dev0
       new  c7b1a6e   BUG: reset internal state of scaler before fitting
       new  4e64915   Merge pull request #5416 from giorgiop/fix-scaler-refit
       new  96f587a   DOC cosmit I find it confusing to say that fit resets the estimator, at it always does that.
       new  ea9da29   BUG: reset internal state of scaler before fitting
       new  1d1d0d8   DOC cosmit I find it confusing to say that fit resets the estimator, at it always does that.
       new  cd12906   MAINT Release 0.17b1
       new  c31de11   Fix broken examples using RandomTreeEmbeddings
       new  5f3cbb5   Merge pull request #5421 from shagunsodhani/proposed-fix-for-#5387
       new  18a53f2   FIX skip LDA deprecation test on python3.3 that has no reload.
       new  0d26d54   Merge pull request #5423 from MechCoder/example_fix
       new  d140e80   MAINT Use the full listing of the rackspace wheelhouse for appveyor
       new  9be4337   MAINT Use the full listing of the rackspace wheelhouse for appveyor
       new  a76d987   Fix broken examples using RandomTreeEmbeddings
       new  6db9ee0   Merge pull request #5429 from amueller/python33_reload
       new  f338a02   FIX skip LDA deprecation test on python3.3 that has no reload.
       new  fa59f82   DOC update doc/readme
       new  1bafa77   Merge tag '0.17b1' into releases
       new  a3ef5b7   Merge branch 'releases' into dfsg
       new  5e61365   Merge branch 'dfsg' into debian
       new  8e8731c   changelog
       new  b6bd07c   dropped CPed patches and refreshed one
       new  0d6b593   requires joblib >= 0.9.2
       new  7e29af0   MAINT remove deprecated stuff that will no longer be supported in 0.18
       new  008adf1   Merge pull request #5375 from rvraghav93/set_precision
       new  bf8f3f1   fixed issue #3450 for hamming loss
       new  49e8fd3   MAINT: deprecation warns from StandardScaler std_
       new  afe1f60   Naive optimization
       new  fbd4b79   - Add max_n_classes param to cluster.supervised metric - Add testing for the Value Error - Check that n_clusters,n_classes are not too high in contingency matrix
       new  5f8ef64   Pre-allocate array in _graph_is_connected
       new  e7a3675   the test to reflect issue #5433
       new  3f86d68   code fix for issue #5433
       new  e916008   Reduce test time.
       new  a29676e   Bug fix for unnormalized laplacian
       new  0543842   - test for unnormalized Laplacian computation
       new  6769d81   correct comments
       new  7f6b79b   Update whats_new.rst
       new  b97e5b4   - removed incorrectly commited file
       new  8cff95f   Revert "- removed incorrectly commited file"
       new  60cff80   Merge branch 'patch-1' of https://github.com/yanlend/scikit-learn into patch-1
       new  56796e4   Merge branch 'master' of https://github.com/scikit-learn/scikit-learn into patch-1
       new  8deaf31   Fixed the docstring (issue #5313)
       new  969337c   DOC: clarify inputs for haversine metric
       new  0cd3d63   referencing datasets documentation in toc
       new  8d2c15c   Merge pull request #5437 from rvraghav93/remove_deprecated_stuff
       new  8c9c8a5   MAINT: remove unused import
       new  40b541e   Merge pull request #5440 from giorgiop/scaler-deprecation
       new  3ab597c   Merge pull request #4995 from yanlend/patch-1
       new  19fc67c   Merge pull request #5454 from GaelVaroquaux/test_branch
       new  657190d   Add test for _graph_connected_component function
       new  a4e4cf9   FIX: remove shuffling in LabelKFold
       new  1dd5518   Correct Typo with  Gael comment
       new  9965f99   fix #4986 doc building on master with circleCI
       new  0c24fee   Merge pull request #5459 from waterponey/circleCI
       new  3fd38e9   Merge pull request #5443 from AlexandreAbraham/optimize_graph_is_connected
       new  658129a   updated test with tips from @giorgiop
       new  23e3987   fixing MaxAbsScaler according to MinMaxScaler
       new  40ba4fc   Merge pull request #5458 from glouppe/labelkfold-withoutshuffle
       new  4f580e6   Deprecate residues_ in LinearRegression
       new  2571158   add test for inverse_transform
       new  9da9b58   add fix for inverse transform
       new  a0fc0ed   changed weighting in hamming_loss in sklearn/metrics/classification.py
       new  e53bfa5   FIX: remove shuffling in LabelKFold
       new  4861730   added test to compare scipy implementation of hamming loss with ours, in sklean/metrics/tests/test_classification.py
       new  ca89b34   ADD New kernel module for GPs which supports kernel engineering
       new  ca03d52   ADD New implementation of GaussianProcessRegression
       new  b76ca72   ADD Gaussian Process classification based on Laplace approximation
       new  804c1d9   TST Tests for GP regression and classification
       new  e548d56   ADD Examples for GP classification
       new  1c5db5f   ADD Examples for GP regression
       new  b4f5463   FIX cross_correlation of anisotropic RBF kernel computed correctly
       new  e9ab6b1   FIX Fix some bugs in kernel module discovered during writing unit test
       new  9dade31   TST Test for GP kernels
       new  515defe   REFACTOR auto_correlation() -> auto(), cross_correlation() -> cross()
       new  7261032   DOC GP's kernel module documented
       new  7b4d935   DOC Revised GP examples (added doc etc.)
       new  1aa49b7   ENH GPR.predict allows returning standard-deviations of predictions
       new  226bb3a   DOC Revised examples based on @eickenberg's comments
       new  a5dc349   MISC Revised gaussian_process package based on @eickenberg's suggestions
       new  2ab097f   TST Dealing with failing tests from test_common
       new  03e3575   REFACTOR Merging kernel methods auto() and cross() in __call__()
       new  5eaa8cf   ENH Adding PairwiseKernel for GPs (wraps kernels from metrics.pairwise)
       new  e165fed   REFACTOR GaussianProcessClassification/Regression -> ...Classifier/Regressor
       new  4dcd706   ENH More efficient approximation of gradient in PairwiseKernel
       new  e05daae   ENH Avoid to create square matrix W_sr in all but one place (TODO)
       new  6b3df1b   ENH Adding RationalQuadratic and ExpSineSquared kernels
       new  722ddf1   FIX Dealing with singular kernel matrix (-inf log-marginal-likelihood)
       new  eb44a42   ENH Adding DotProduct kernel and fixing ExpSine2 kernel
       new  f534a97   DOC Initial version of Mauna Loa CO2 example
       new  d8405c2   REFACTOR Hyperparameter optimization in GPs enabled explicitly by optimizer parameter
       new  98f9774   DOC Revising CO2 example
       new  4d3eb96   TST Cover more kernels in tests
       new  209b202   ENH Analytic gradients for ExpSine2, RationalQuadratic, DotProduct kernels
       new  a13f765   DOC Extended documentation of GP kernels module
       new  cb7c424   MISC Reorganizing kernel module
       new  c259905   ENH Adding method is_stationary to GP kernels
       new  3ad82d8   TST Testing kernel stationarity
       new  49badf9   DOC Illustrating different kernels in plot_gpr_prior_posterior.py
       new  555bb11   FIX test_converged_to_local_maximum deals with convergence to boundary
       new  ef86482   ENH More efficient computation of lml-gradient in GPR using einsum
       new  d0bc5dc   FIX GPs do not modify kernel attribute but store optimized kernels in kernel_
       new  9f6e48f   REFACTOR GP kernels use separate specification of bounds and params (no joint param_space)
       new  403b5f1   REFACTOR REnamed kernel property params to theta
       new  8699a4c   ENH Kernel implements equality check and get_params() method
       new  73b32e3   TST Tests for cloning kernels with sklearn.base's clone function
       new  2feb6e0   REFACTOR Replacing all calls to deepcopy() by clone()
       new  752b6b8   TST Replace call to deepcopy() by clone()
       new  a3abc09   ENH Added Exponentiation kernel (DotProduct no longer has internal degree)
       new  0210de4   TST test_kernels tests Exponentiation kernel
       new  91b5739   FIX Using numpy-downward compatible ones_lile() instead of full_like()
       new  27bd525   DOC Reorganizing gaussian_process examples
       new  e3423b9   FIX Minor bugfixes and downward-compatibility/python3 issue
       new  141eb0b   REFACTOR Rename GPR method sample to sample_y
       new  0a23d9e   REFACTOR Compute standard deviation of predictive distribution at n points in O(n)
       new  3043a12   FIX GPC uses LabelEncoder in fit() and returns (n_samples, n_classes) array in predict_proba()
       new  ff92bc9   FIX Adapted examples and tests to GPC.predict_proba() adapted return shape
       new  40b1426   FIX GaussianProcessRegressor inherits from RegressorMixin
       new  fd2cfc8   ENH Reducing memory-consumption of GPC.predict_proba()
       new  6af3ffc   FIX Numpy downward-compatible way of computing sum over two axis
       new  fb4eded   ENH GPC can use warm-starts in _posterior_mode()
       new  1c70e52   ENH Adding diag(X) method to GP kernels.
       new  1790b94   DOC Adding GPC to plot_classifier_comparison.py script and enabling warm_start
       new  3073d7e   DOC gpr module fully documented
       new  7773707   DOC gpc module fully documented
       new  79d2194   REFACTOR Kernel hyperparamters are now set via their name
       new  4571a7f   TST Adapted tests to changes in kernels and added additional test for theta property
       new  82ec208   DOC Adapted GP examples to changes in kernel module
       new  e3413b7   FIX Kernel property theta can deal with anisotropic RBF kernel
       new  31ef645   FIX Dealing with anisotropic length-scales correctly
       new  caa4196   DOC Revised docstrings of kernels.py
       new  ebf3f05   REFACTOR Removed literal form of specifying bounds for ConstantKernel
       new  c961392   REFACTOR Changed order of l and alpha in RationalQuadratic kernel
       new  02ab93e   ENH GPR can use individual noise-levels for each datapoint
       new  7f8bd09   ENH GaussianProcessRegressor supports multiple restarts of the optimizer
       new  95d6577   TST An increasing number of random-starts of GPR only increases the LML of the chosen theta.
       new  04770b1   DOC Adapted old example of fitting with noisy targets to new GPR class
       new  22f049b   ENH GaussianProcessClassifier supports multiple restarts of the optimizer
       new  d88670c   TST An increasing number of random-starts of GPC only increases the LML of the chosen theta.
       new  54df417   FIX Fetching CO2 data from mldata
       new  d0a5ff8   FIX GP kernels check for type equality before checking attribute equality
       new  51f01d6   FIX cross_validation and pairwise_kernels treat GP kernels as special case
       new  0b1fb33   DOC Adding example comparing kernel ridge with GPR.
       new  a9a884d   DOC Revised plot_gpc example
       new  d26c355   DOC Adding documentation for GP kernels
       new  a5c1187   DOC Documentation of the CO2 example
       new  698ecf5   DOC Backported documentation to kernels module
       new  cb9935b   DOC Extending further the GP documentation
       new  66acd22   ENH GPR supports normalization of target values
       new  d778496   TST Testing normalization of output values in GPR
       new  84b49f9   ENH GPR supports target value with more than one dimension
       new  e813a73   TST Testing multi-out GPR
       new  76a1b16   FIX Not performing hyperparamter-tuning in GPR if there are no hyperparameters
       new  897da0f   ENH GPR can use an externally defined optimizer for hyperparameter tuning
       new  017a7fb   TST Testing GPR with custom external optimizer
       new  4db5142   REFACTOR GP kernel hyperparameters theta and their bounds are log-tranformed
       new  15309a4   TST Adapting GP tests to changed hyperparameter representation
       new  41cbb29   DOCAdapting GP examples to changed hyperparameter representation
       new  bcd4b72   REFACTOR Remove theta_ attribute from GPR and GPC
       new  07cd4d4   ENH GPC can use an externally defined optimizer for hyperparameter tuning
       new  78caa0b   TST Testing GPC with custom external optimizer
       new  2839f66   REFACTOR normalize_y in GPR does not modify standard-deviation
       new  74bbacb   FIX Sampling theta_intial uniformly from log-bounds in GPR
       new  8af909b   FIX Correcly composing theta vector in GP kernels
       new  0d34a34   FIX Ensuring that predicted variances are never negative
       new  daa90c2   ENH Adding Matern kernel for Gaussian Processes
       new  955aa60   DOC Adding documentation on Matern kernel
       new  588cabf   TST Added test checking that GPR can deal with different outputs for same input
       new  9443f94   FIX Checking for correct dimensionality of anisotropic kernels
       new  37a6348   FIX RBF and Matern kernel handle 1d length 1 length-scales correctly
       new  7299cb6   DOC Added further doc for GPC
       new  bf0e290   ADD CompoundKernel for multi-class GP classification
       new  a5e207d   FIX GPR's random_state is cloned correctly
       new  3a31262   ENH Support for multi-class GPC added based on OneVsRestClassifier
       new  5b47b04   DOC Adding GPC to plot_classification_probability.py example
       new  0fa050c   TST Testing GPC for multi-class classification problems
       new  27b3016   FIX Several minor changes/fixes in gaussian_process module
       new  c8f23f4   FIX Using six.string_types instead of basestring (python3 compatibility)
       new  9929b37   FIX Excluding WhiteKernel from one unit-test
       new  1f7deb7   TST Extending testing of multi-class GP
       new  0de854a   DOC Updated documentation of Gaussian process classifier
       new  f6d8686   ADD Example illustrating Gaussian process classifier on iris dataset
       new  de3889d   TST Fixing some GP-releated issues in estimator checks
       new  92a6354   TST Fix GPC and GPR such that all tests are ok
       new  768c81f   TST Further GP-related fixes
       new  59d754b   FIX Hiding method decision_function in GaussianProcessClassifier
       new  d42aa19   TST GaussianProcessRegressor added to MULTI_OUTPUT estimators
       new  adc7451   PEP8 Fixing PEP8 issues in gaussian_process package
       new  c57aa64   ENH X_train_ can be copied if requested
       new  83bc6c2   ENH Adding max_iter parameter for Newton iteration in GPC
       new  9b86186   COS Removing keyword magic from PairwiseKernel
       new  032c2a7   FIX Added missing import of ConstantKernel in gpc.py
       new  51df6e3   TST Excluded GPC from test_non_transformer_estimators_n_iter()
       new  68be69f   MISC Deprecating legacy GP implementation
       new  34506ad   COS Making BinaryGaussianProcessClassifierLaplace private
       new  5c2725b   DOC Cleaning up documenation of legacy GPs
       new  992b8c4   DOC Updating GPC XOR example (showing also results for DotProduct kernel)
       new  03718f8   FIX Remove BinaryGaussianProcessClassifierLaplace from packages __all__ list
       new  0c255e6   FIX Undoing regression_model and correlation_model  deprecation
       new  a6d564b   TST Testing n_jobs option of GaussianProcessClassifier
       new  d34b7dd   REF Realizing multi-class GPC bei delegating to OneVsRest or OneVsOne
       new  7f7455c   DOC Updating documentation of multi-class GPC
       new  c6db24e   REF Several small changes based on @eickenberg's suggestions
       new  d0eb99b   ENH More efficient computation of K_gradient for anisotropic Matern kernel
       new  65555b3   REF Renaming sigma_squared_n to alpha (consistency with Ridge)
       new  d344fa6   ENH Full support of (deep) get_params() and set_params() in GP kernels
       new  35aace4   DOC Revising GP documentation based on @kastnerkyle comments
       new  4c40f64   FIX theta_vars correctly supported in KernelOperators
       new  3f1a0f1   TST Adding two unittests for testing GP kernels
       new  eb54b49   REF Removing theta_vars attribute from kernels and using Hyperparameter instances instead
       new  c85177e   DEL Removing bounds.setter in kernels.py
       new  e87fad3   REF Renaming max_iter to max_iter_predict in GaussianProcessClassifier
       new  1813322   DOC Documenting GP kernel API
       new  4772192   DOC Fixing some minor issues in narrative doc of GPs
       new  f76e64c   FIX Fix failing tests (doctest, order of hyperparameters,  pairwise_kernels)
       new  e6f0284   FIX python3 compatibility in doctest
       new  55d26aa   FIX Fixing test_kernel_theta
       new  98e8093   FIX Further python3 related fixes
       new  266b9e9   FIX Fixing a python2.6-related issue
       new  d21b648   FIX Changing bounds in GP doctest
       new  1c4b23f   MISC copy_X_train defaults to True in GPR and GPC
       new  213a6be   FIX Fix message of ValueError for wrong shape of alpha in GPR
       new  2dd1f28   PEP8 Making pep8 and pyflakes happy
       new  e2a340a   REF More meaningful names for hyperparameters of GP kernels
       new  8f77ced   MISC Changing default for copy_X_train to True in GaussianProcessClassifier
       new  c3a41a3   FIX String comparison via equality and not identity in Hyperparameter.__init__
       new  3957fc3   ENH GPR.log_marginal_likelihood() returns the current log-likelihood if no theta vector is provided
       new  2649598   FIX Enforcing y to be numeric in GPR and fixing python3 issue with maps
       new  edc4a65   ADD Mixins for normalized and stationary kernels
       new  cea3f29   TST test_lml_precomputed() checks only for equality in first 7 digits
       new  b663f3f   MAINT: deprecation warns from StandardScaler std_
       new  5c697ba   Merge pull request #5466 from glouppe/gp_new
       new  2f15969   DOC: added entry for new GP implementation
       new  b910ab8   ENH add pomegranate to related
       new  74cfca8   Merge pull request #5452 from MaryanMorel/issue_5313
       new  e3afc0e   MAINT move custom error/warning classes into sklearn.exceptions
       new  09f0fa7   DOC/SPHINX Add template for classes which don't have init (Exception classes)
       new  e19a981   MAINT Deprecated the UndefinedMetricWarning at sklearn.metrics.base
       new  61ae8d5   MAINT move deprecated into deprecation.py
       new  f6b85d9   Merge pull request #5445 from tomMoral/devTom
       new  b10b9b3   debian/patches - deb_disable_google_anals to disable google tracking
       new  ef5a5c4   Improve travis config file (for :#5455)
       new  b459df8   BF: use systemwide underscore.js
       new  c0d5998   BF: remove minimized jquery.js
       new  24f6015   Merge branch 'dfsg' into debian
       new  857cb09   FIX/ENH Make the moved class resilient to the deprecation patching
       new  f24d94a   remove test for self.scale_
       new  69a09f6   Merge pull request #5468 from jmschrei/pomegranate
       new  056b029   doc: fix typo
       new  9758e25   FIX: gaussian_process will be deprecated in 0.19, not 0.18
       new  d271708   Merge pull request #5475 from glouppe/gp-deprecation
       new  bc15ceb   Add test before miniconda download
       new  d4e9d79   Merge pull request #4826 from rvraghav93/exceptions
       new  627e262   fontsize legend parameter is not available for matplotlib < 1.3
       new  e4332ed   DOC Add description for UndefinedMetricWarning
       new  abc032c   Merge pull request #5479 from pletelli/fix_circleci_build
       new  53bf94c   Merge pull request #5478 from rvraghav93/description_undefinedmetricwarning
       new  eaf97bf   colorbar doesn't handle label parameter
       new  74c09e5   Merge pull request #5480 from pletelli/gpc_plot
       new  baf4ade   Update my website on whats_new.rst
       new  904c667   MAINT: use xrange from six
       new  18f4a29   ENH: fixes issue #5484
       new  9865e9a   Merge pull request #5486 from jakevdp/xrange-fix
       new  17eb0b6   WIP commit - add ls and print
       new  ca0c84c   Test caching strategy for travis
       new  b7acd76   Corrected macro ROC in example plot_roc
       new  b066e37   Remove testvenv from cache
       new  92068ef   #5322: Wrote check_supervised_y_no_nan in estimator_checks
       new  b4e746b   Resolving rebase conflicts on forest.py
       new  7745319   #5322: Added check for DecisionTreeRegressor
       new  b1a1aa8   #5322: Resolving errors for SVR and NuSVR
       new  a5860a1   #5322: Fixed build for tree failures, fixing estimator_checks.check_estimator
       new  6df1fe2   #5322: Added check for MultiTaskElasticNet and MultiTaskLasso
       new  d62e316   #5322: Resolving Random Forest build errors (bad input shape)
       new  34c8399   #5322: Fixing the fix - now RFs should be okay
       new  7f82026   #5322: Fixing test in test_estimator_checks.test_check_estimator
       new  ef4faf6   #5322: check_supervised_y_no_nan raises error when error message is incorrect
       new  011dcd2   #5322: Fixed python3 build
       new  c657d77   #5322: Regression test for test_check_estimator
       new  eea5ba3   #5322: Made error message in check_supervised_y_no_nan more helpful
       new  fffa98a   MAINT Remove deprecated stuff that will be removed in 0.18
       new  52d0c89   deleted a commented line (line 1464 in sklearn/metrics/classification.py)
       new  fe1f890   slight tune up for license file -- BSD-2
       new  235f256   - deb_nopaypal  and deb_no_online_jquery
       new  3bfbe79   #5322: Making check_supervised_y_no_nan deterministic
       new  52a6acc   #5322: Addressing @MechCoder's changes
       new  7b3da77   #5322: Removing unused ConvergenceWarning import in svm/base.py
       new  f634bc2   Merge pull request #5438 from aabadie/datasets_doc
       new  8b78293   Merge pull request #5494 from dohmatob/minor-issues
       new  6a88fb6   Merge pull request #5319 from nicodv/patch-1
       new  5db2adf   MAINT Removed deprecated stuff.
       new  62633f7   Addressed comments on PR #5451
       new  95d4e23   Merge pull request #5436 from Mbompr/plot-roc-example-4976
       new  4bcfedb   FIX: GaussianProcess will be deprecated in 0.20, not 0.19
       new  bad3dd6   Merge pull request #5451 from zermelozf/deprecated-stuff
       new  b18f295   randomized_svd: power iter, normalization, benchmark
       new  718a7df   Merge pull request #5449 from Jeffrey04/5433-MaxAbsScaler-1-row-csr-fix
       new  7a2e2d8   Tweaking the legend of GP plot in plot_compare_gpr_krr.py
       new  0d340b4   MAINT disable circle ci on 0.17.X
       new  cd5688a   MAINT Removed deprecated stuff.
       new  3382733   Addressed comments on PR #5451
       new  76e0273   Deprecate residues_ in LinearRegression
       new  00996a2   FIX MaxAbsScaler on sparse matrices with 1 row
       new  0cb93b0   Merge pull request #5141 from giorgiop/power-iter-randomized-svd
       new  08b7253   Update deprecation message in metrics/regression.py (use uniform_average as default multioutput parameter of r2_score function after 0.19)
       new  888a184   Updating deprecation message in metrics.py: confirm deprecation starting from 0.16 and removal after 0.18
       new  1eb9bed   COSMIT nicer output from NMF with verbose > 0
       new  8e0f337   removing sklearn/metrics/metrics.py
       new  ee0fd9a   Merge pull request #5446 from deepcharles/charles1
       new  3697d75   MAINT/TST: public export of non_negative_factorization
       new  65aee0c   Merge pull request #5469 from rvraghav93/zero_gamma
       new  7c69c55   Merge pull request #5457 from aabadie/remove_deprecation_metrics_second_pass
       new  ebf8d3a   Merge pull request #5456 from aabadie/remove_deprecation_metrics
       new  b6a5ed2   Merge pull request #5465 from massil94/dev-massil
       new  f6d605f   Implement decision path in tree
       new  392ef50   DOC add an example highlighting the tree structure
       new  317281b   take glouppe comment
       new  4d9a955   DOC make visualizatoin aware of the tree depth
       new  bc02bd5   ENH rename decision_paths to decision_path
       new  be21b6a   Add an hardcoded test
       new  405c717   typo
       new  7d4755e   Add a decision_path function to forest estimator
       new  c9888a6   Take into account glouppe comments
       new  ec8fbef   Take into account amueller comments
       new  4d0ba3c   wording
       new  c450751   Added documentation about y nan handling in check_X_y
       new  d439dc4   FIX class_weight in LogisticRegression and LogisticRegressionCV
       new  7ad90f5   Rename since we don't plot anything
       new  290e0ea   #5322: Changed check_supervised_y_no_nan to not seed global
       new  f45f260   DOC versionadded randomized_svd
       new  3c988d5   Merge pull request #5512 from giorgiop/randomized-svd-versionadded
       new  1f48814   #5322: Using multioutput_estimator_convert_y_2d in check_supervised_y_no_nan
       new  f725485   #5322: Small change to estimator_check.py
       new  587315e   (WIP) Added Multi-layer perceptron (MLP)
       new  b52bae4   refactor mlp optimization methods into _fit_lbfgs and _fit_sgd, add Ctrl+C stop option for SGD
       new  cb4c935   minor fixes to bench_mnist.
       new  ee071e0   FIX partial fit test for MLP
       new  3c8210f   ENH better 'constant' learning rate schedule
       new  3891af8   iterate, improve. Nesterov's momentum.
       new  bde2270   fix log_loss for multiclass case and add binary_log_loss for multilabel case
       new  7fe71c1   as_float_array removed
       new  744d161   Merge pull request #5431 from hlin117/nan-targets
       new  861ac13   Merge pull request #5519 from ishank08/imp_transform
       new  3fbd3b9   Take into account ngoix comment
       new  5d3bb93   Merge pull request #5008 from TomDLT/logistic_multiclass
       new  289c0a3   FIX class_weight in LogisticRegression and LogisticRegressionCV
       new  5a58e56   Merge pull request #5487 from arjoly/example-node
       new  1c7168d   Fix fit_transform, stability issue and scale issue in PLS
       new  8888a98   Fix fit_transform, stability issue and scale issue in PLS
       new  f63ba1d   MAINT use super rather than calls to base class methods
       new  a6c2a17   undo change by @arthurmensch that was probably accidental ;)
       new  4eee24a   Merge pull request #5517 from lesteve/replace-base-class-method-calls-by-super
       new  824115a   Added colorblind compatibility to classification
       new  47a8ffa   Added colorblind compatibility
       new  efe8399   update whats_new for #5008
       new  4f7e5ff   Fixed typo..
       new  80e22b3   Patch liblinear for sample_weights in LogisticRegression(and CV)
       new  41cbfde   Add check for sample_weights
       new  9d8d7e4   Replace GETC to SAMPLE_WEIGHT
       new  8b990c3   Merge pull request #5543 from ltiao/patch-1
       new  6cb51b2   Initialize ARPACK eigsh
       new  9f8e062   fix docs and doctests for KernelPCA
       new  8ff339e   Update test_utils.py
       new  1863895   Merge pull request #5506 from RachitKansal/my-work
       new  2492dd0   Merge pull request #5012 from yanlend/patch-2
       new  dbb418b   Typo on doc for default value of solver for lda.LDA
       new  31ab442   fixed typo in LDA
       new  d4532e2   Merge pull request #5544 from Mbompr/check-PR-5417
       new  3fd1473   Added colorblind compatibility
       new  0ff1fe5   Merge pull request #5541 from johannah/module-classification-colorblind
       new  1c5d6d7   Merge pull request #5274 from MechCoder/liblinear_samples_weights
       new  b102d90   MISC : Y->y
       new  c9700e2   Changed from plot to scatter for consistency
       new  38c78ea   Merge pull request #5537 from johannah/module-linear-colorblind
       new  8aa9c19   Merge pull request #5467 from tomMoral/travis-ci
       new  844fbae   added colorblind compatibility
       new  8108bdb   Made plots colorblind friendly
       new  e8a2e27   Test edit.
       new  e82bca5   Removing test edit.
       new  a03333f   Moved Multi-task Lasso to immediately follow Lasso
       new  a614a5c   Edits to intro paragraph of Multi-task Elastic net.
       new  61f5313   Edits to math paragraph, added reference to MultiTaskElasticNetCV.
       new  6c455ae   Corrected reference to "lasso" to "ElasticNet".
       new  29b4931   Correct for Frobenius norm
       new  0b00e2b   Corrected {}
       new  07a9b8d   Corrected MultiTaskLasso objective function and moved some related definitions there as suggested by agramfort.
       new  a8880f1   Fixed typos: whitespace after '.', `` around inline code.
       new  93c3033   nitpicks
       new  55c68ef   Added colorblind compatibility
       new  d53772d   FIX serialization bug in the implementation of the pickle protocol
       new  d398034   BF: FIX OvR decision_function_shape in SVC
       new  7fc96a0   ENH don't unpickle capacity, but infer it
       new  5427e08   BF: FIX OvR decision_function_shape in SVC
       new  a1844be   DOC: added <what's new> entry regarding BF
       new  409c888   Raise appropriate error if y is sparse
       new  3f8743f   Main Commits - Major --------------------
       new  917bacb   Finish up mlp_refactoring and squash previous commits
       new  efb0179   Merge pull request #5560 from arjoly/serialization-bug-tree
       new  9e42250   HOTFIX for incorrect cast in liblinear.pyx
       new  f48d0a0   Added colorblind capability
       new  8343081   Added colorblind compatibility
       new  965a715   Merge pull request #5214 from glennq/mlp_refactoring
       new  0ba5c24   Update what's new
       new  1ce55d2   Adding shield for circleCI in README
       new  d15f322   Added colorblind compatibility
       new  b255056   Merge pull request #5565 from MechCoder/hot_fix
       new  0a20a37   doc fix lates in softmax formula
       new  c38505d   Merge pull request #5550 from johannah/module-text-colorblind
       new  34a97ec   Merge pull request #5576 from johannah/module-mixture-colorblind
       new  a94ffde   DOC: fix what's new alignment
       new  f69f895   remove _balance_weights
       new  8878088   FIX inspect.argspect is deprecated; use signature
       new  6a0b6df   Merge pull request #5579 from rvraghav93/inspect_fix
       new  06181b4   DOC: only 'sag' supports sparse input in Ridge now
       new  108d36a   DOC clearer what's new for model_selection
       new  34d2027   Merge pull request #5586 from giorgiop/patch-1
       new  a75eb63   iforest
       new  e2a53f0   Merge pull request #5582 from gclenaghan/remove_balanced_weights
       new  420e9b0   Merge pull request #5587 from jnothman/model_selection-whatsnew
       new  5c8855d   Merge pull request #4163 from ngoix/iforest
       new  c99f5db   doc: added entry for isolation forest
       new  fea218e   In tree, min_samples_split and min_samples_leaf now accept float number as percentage.
       new  a90dc38   Merge pull request #5546 from johannah/module-svm-colorblind
       new  58beae8   Merge pull request #5573 from johannah/module-model-colorblind
       new  a20e37a   Update what's new
       new  6741ec7   Merge pull request #5552 from johannah/module-tree-colorblind
       new  84346b8   Merge pull request #5556 from johannah/module-feature-colorblind
       new  07c8c58   Merge pull request #5575 from johannah/module-other-colorblind
       new  e292ac0   FIX: pass random_state to parameter distributions
       new  e30af3c   Merge pull request #5590 from glouppe/param-sampler-rng
       new  16798e3   s/labe/label
       new  c81164c   corrected typo
       new  d2d00d5   Merge pull request #5591 from antialias/master
       new  d591ac4   Merge pull request #5592 from antialias/correct-typo
       new  d01dd2f   English language changes to IsolationForest documentation
       new  be2499e   The docs have moved
       new  6541f3f   Merge pull request #5594 from betatim/isolationtree-docs
       new  d9f3277   Merge pull request #5531 from arjoly/float-min_samples
       new  adf866c   Replaced fontsize legend parameter
       new  0c29002   Added colorblind compatibility
       new  9fe25df   Merge pull request #5597 from vortex-ape/matplotlib_issue
       new  77156bf   RF: use pandas rules approach to shuffle things into -lib etc
       new  9671015   adjusted changelog, closes statements, point to debian branch for Vcs-Git
       new  c365ab9   Pushing doc to github.io
       new  ca724f5   boost debian/compat, fix .installs etc
       new  4e6829c   Corrected macro ROC in example plot_roc
       new  ff1ef8f   Remove minimized js files
       new  6915f7c   Merge branch 'dfsg' into debian
       new  dcb9b59   also use non-minimized versions for jquery scripts
       new  b52b6cb   added maphilight copyright, updated years for the others
       new  918005f   added overrides for the maphilight source
       new  e87f203   Merge commit '0.17b1-14-g4e6829c' into dfsg
       new  a715a9f   Merge branch 'dfsg' into debian
       new  b538348   updated changelog
       new  8c44bb6   REFACTOR: fixed huge code duplication in _RidgeGCV
       new  44c8519   Fix of issue #5601
       new  8b1e27e   DOC: broken links
       new  ec7d1db   Merge pull request #5607 from giorgiop/doc-broken-links
       new  8497bf3   Merge pull request #5577 from waterponey/circleCI_doc
       new  f8ecb8f   MAINT Cython will soon be a build requirement
       new  6c56b25   Merge pull request #5574 from johannah/module-decomposition-colorblind
       new  28d5345   MAINT: remove unused code from TSNE
       new  b9b32db   MAINT make nosetests verbose output more informative
       new  e3a95bf   Merge pull request #5613 from jakevdp/tsne-cleanup
       new  d46b8ab   change doc outlier_detection
       new  f67e911   DOC Modify documentation/examples for the new model_selection module
       new  425407b   Merge pull request #5617 from ngoix/doc-iforest
       new  105326d   FIX increase tolerance of class weight check for OS X
       new  53228c9   List creation could be litteral
       new  bd87003   Dictionary creation could be rewritten as litteral
       new  35ba568   Chained compararisons can be simplified
       new  efe93d4   MAINT update joblib to 0.9.3
       new  a4f125b   FIX decrease tolerance in test_logistic for appveyor failure
       new  cdeb5ca   Merge pull request #5635 from ogrisel/ogrisel-joblib-0.9.3
       new  4b1489c   Merge pull request #5633 from sieben/list_litteral
       new  0b24027   Merge pull request #5632 from sieben/chain_comparaisons
       new  ee53acc   Merge pull request #5629 from sieben/dict_litteral
       new  99530a3   Merge pull request #5566 from rvraghav93/mod_sel_doc
       new  95c54b2   Addressed comments on PR #5451
       new  79c86e9   Deprecate residues_ in LinearRegression
       new  68f4965   FIX MaxAbsScaler on sparse matrices with 1 row
       new  b1af4d5   FIX class_weight in LogisticRegression and LogisticRegressionCV
       new  2bb4d09   Fix fit_transform, stability issue and scale issue in PLS
       new  d052749   BF: FIX OvR decision_function_shape in SVC
       new  53d9bbb   Corrected macro ROC in example plot_roc
       new  339c59b   DOC: only 'sag' supports sparse input in Ridge now
       new  3ac4719   Fix of issue #5601
       new  8400e95   undo change by @arthurmensch that was probably accidental ;)
       new  f35a3ef   MAINT update joblib to 0.9.3
       new  9a0c421   fix rounding, adjust tests for 32 bit export_graphviz
       new  c95ee28   remove useless pass
       new  0c36694   Merge pull request #5645 from sieben/useless_pass
       new  598f1cb   Fix formatting issue in tree.rst
       new  f76d9c4   Merge pull request #5646 from mscherer/fix_doc
       new  45c2356   Redirecting to open issues
       new  90378af   Merge pull request #5636 from TomDLT/appveyor_logistic
       new  e478708   adding sklearn data home
       new  0f6535d   Fixed user guide link in ransac.py #5621
       new  c0e816b   Default mutable argument
       new  23f2891   Merge pull request #5648 from sieben/default_argument_mutable
       new  c60f72b   Merge pull request #5647 from sieben/patch-1
       new  f8eb351   Useless trailing semicolon
       new  31f73a6   Simplify chain comparaison
       new  2199e50   Migrate examples to avoid deprecated module
       new  47a07c8   Merge pull request #5652 from sieben/trailing_semicolon
       new  4a9ba05   Merge pull request #5655 from sieben/deprecated
       new  bc142c8   Merge pull request #5650 from shawpan/fix-user-guide-link-ransac#5621
       new  bac5e39   Merge pull request #5626 from ogrisel/fix-osx-logistic-class-weights
       new  5b602ae   FIX decrease tolerance in test_logistic for appveyor failure
       new  6f69e83   FIX increase tolerance of class weight check for OS X
       new  c37c938   Update cross_validation.py
       new  3b7ad17   DOC polish documentation of output types of train_test_split, add change to 0.16 whatsnew.
       new  3377174   DOC: Add defaults for oob_score in docstrings
       new  3a3617b   Merge pull request #5659 from ElDeveloper/doc-fix
       new  48a2329   Merge pull request #5644 from trevorstephens/export_gviz_32bit_fix
       new  ce861b4   fix rounding, adjust tests for 32 bit export_graphviz
       new  9cc3b95   TST make tsne tests 32bit save
       new  e3ff540   COSMIT don't use the deprecated residuals property of ols
       new  e232ad6   TST catch warnings in tests that the solver is changed to SAG in the sparse case.
       new  95f5e3a   TST close /dev/null in the theil sen tests.
       new  ddadf9c   Merge pull request #5658 from amueller/train_test_split_conserves_type_doc
       new  ec0a1c8   Merge pull request #5667 from amueller/theilsen_test_closes_dev_null
       new  f4f8b99   Merge pull request #5665 from amueller/no_deprecated_residuals
       new  faa6848   Merge pull request #5654 from sieben/simplify_chain_comparaisons
       new  651ade0   Remove cython warnings
       new  9717a81   Make clear that GMM is a clustering algorithm
       new  efadbd3   Calculate max_depth when max_samples is known
       new  56b51de   DOC+ENH: RidgeClassifier doesn not support multi-label
       new  b1f3ddb   Merge pull request #5657 from sieben/remove_cython_warnings
       new  d29562e   Merge pull request #5666 from amueller/sag_sparse_warnings
       new  9b0dd53   Fix docs and remove duplicated checks
       new  e37c69f   Comment on weird initialisation for scoping reasons
       new  0e55532   FIX port LDA covariance fix to decomposition module
       new  1321667   FIX we shouldn't warn the user if the solver is auto
       new  83313de   Merge pull request #5661 from amueller/tsne_test_32bit
       new  85fe5a0   TST make tsne tests 32bit save
       new  70b637a   TST catch warnings in tests that the solver is changed to SAG in the sparse case.
       new  edd2ddf   Remove "clustering" from headings
       new  00816a6   DOC polish documentation of output types of train_test_split, add change to 0.16 whatsnew.
       new  52b445c   Unify documentation in GMM module
       new  c9769dd   COSMIT don't use the deprecated residuals property of ols
       new  c26fbf1   TST close /dev/null in the theil sen tests.
       new  86974f8   Rename example to plot_gmm_covariances.py
       new  2a98fc6   Don't use floats to index numpy arrays
       new  89f8a51   OneHotEncoder warn fix
       new  ca88f91   Merge pull request #5673 from amueller/sag_dont_warn_on_auto
       new  c242c29   MAINT catch warning for deprecated allow_list option to train_test_split
       new  601d43a   FIX we shouldn't warn the user if the solver is auto
       new  1e23805   DOC: Correct confusing docs of n_values in OneHotEncoder
       new  776e53b   skip unstable tests on 32bit platform
       new  866c937   Merge pull request #5637 from MechCoder/one_hot_enc_doc
       new  4bd1286   COSMIT missing whitespace in error message
       new  13cc60a   Merge pull request #5572 from dohmatob/fix-5562
       new  1f650db   Test for max_depth calculation and re-adding max_samples_ property
       new  ee53e52   DOC - neural network module only links to unsupervised, not supervised docs
       new  149aade   Fixed string comparison on arrays in new Gaussian process described in issue #5663
       new  b9fdeb0   shortened if condition based on glouppe's suggestion
       new  2653833   DOC some fixes to the doc build.
       new  3ab249e   added warning tests for test_roc_curve method
       new  f7e886a   Merge pull request #5682 from trevorstephens/OneHotEncoder_warn_fix
       new  0188e68   Merge pull request #5684 from amueller/deprecated_float_indexing_rfe
       new  6ccdbaf   REL Make 0.17 website changes for release.
       new  912e005   Don't use floats to index numpy arrays
       new  9eb888c   OneHotEncoder warn fix
       new  61df16e   fix 1 sparse row scaling in robust scaler
       new  e2eba1f   Merge pull request #5688 from amueller/robust_scaler_1column_fix
       new  bb71d2d   fix 1 sparse row scaling in robust scaler
       new  1774f3f   fixed missing import from too much cherry picking.
       new  550b228   FIX don't compare things that can be arrays to strings.
       new  f9b92cb   Merge pull request #5668 from amueller/multioutput_string_comparison
       new  88e86f3   FIX don't compare things that can be arrays to strings.
       new  c55dc89   Merge pull request #5695 from amueller/doc_fixes
       new  22fc441   Merge pull request #5680 from amueller/lda_fix_refactoring
       new  2f2adcc   Add missing whatsnew entries for 0.17
       new  2df6bfa   DOC some fixes to the doc build.
       new  27dd394   FIX port LDA covariance fix to decomposition module
       new  36acda2   Merge pull request #5708 from amueller/0.17_whatsnew
       new  053a1e6   Add missing whatsnew entries for 0.17
       new  468b4c1   Version Added for Robust Scaler
       new  2249daa   Version added for all new classes
       new  d3a8f85   Added more documentation of version adds
       new  f2e3541   Added more versions of 0.17
       new  c6037e7   Ensure that n_nonzero_coefs is an int
       new  b4adafe   Merge pull request #5698 from gclenaghan/fix_5687
       new  bc85b78   Test max_samples_ attribute, set max_depth in BaseBagging
       new  986fb60   C file removal + cython build + Travis build cache
       new  db3eca2   DOC Add the learning curve documentation to model_selection.rst
       new  213a24c   Merge pull request #5700 from bhargav/neural-network-documentation
       new  0bdd788   Ensure that n_nonzero_coefs is an int
       new  417ee28   DOC Typo fixes in whatsnew. Thanks @jnothman
       new  42f685e   DOC Typo fixes in whatsnew. Thanks @jnothman
       new  7af49df   Fix doc string and number vs string ordering
       new  41d0f1f   split installation into simple and advanced part
       new  7d574ad   Merge pull request #5511 from KamalakerDadi/sphinx_version_added
       new  f038915   Version added for all new classes and parameters
       new  cdca3cf   Update copyright year in the doc footer (2014 → 2015)
       new  53bc02a   Merge pull request #5727 from Naereen/patch-2
       new  e3bdff6   MAINT Don't use deprecated 1d X (or deprecated matplotlib stuff) in examples.
       new  21a9994   DOC some fixes to docbuild
       new  812c3a4   REL add 0.17.0 release to news
       new  a4122f0   More doc fixes. Latex builds again.
       new  77665e3   removes extra spaces from error message
       new  8192cea   Merge pull request #5731 from ClimbsRocks/patch-2
       new  f084cb8   skip doctests on 32 bits
       new  de3a527   Merge pull request #5692 from amueller/skip_32_bit_tests
       new  9334274   skip unstable tests and doctests  on 32bit platform
       new  0c7812a   Add subpackage _build_utils
       new  7c35640   Merge pull request #5257 from amueller/installation_instructions
       new  37d18ce   split installation into simple and advanced part
       new  95d91c9   Merge pull request #5714 from rvraghav93/learning_curve_doc
       new  45ef71f   MAINT version string for 0.17. D'OH
       new  da4f480   Fix import of reload for python 3.3
       new  219fe10   BUG: use pre-parsed scipy version
       new  c957249   Fix link to dev docs.
       new  6c3b68d   Merge pull request #5740 from perimosocordiae/patch-2
       new  9cf5b52   DOC fix link to pdf documentation
       new  1e1efb5   DOC fix link to pdf documentation
       new  8bae9d9   Merge commit '0.17b1-14-g4e6829c' into releases
       new  a04972b   Merge tag '0.17' into releases
       new  f9a215c   Merge branch 'releases' into dfsg
       new  8bec97c   Merge branch 'dfsg' into debian
       new  b17b4a4   changelog
       new  2366b5c   debian/patches/deb_no_tinyclues_external_image - to not link to external images inthe documentation (privacy concern)
       new  77a8a18   DOC: fix what's new alignment
       new  82fb053   BF: for building arch packages with no arch indep ones, find should be conditioned
       new  c024978   Improve wording for best choice of calibration method
       new  886166e   Proper fix for issue #5694: added [RW2006]'s corresponding reference with link to the book's website and PDF version available online.
       new  58d2e29   Merge pull request #5749 from ethanwhite/calibration-wording
       new  34f9367   grammatical fix in tutorial.rst under basic category
       new  de5bc4a   MNT 0.19 --> 0.20; As the model_selection changes will be released in 0.18 only
       new  edb344a   Reduced nb of characters per line.
       new  5339049   Merge pull request #5753 from staranjeet/grammatical-fix
       new  d963df4   DOC - Minor documentation fixes - Fixes #5770 and #5761
       new  6f6ec7f   Reference correct variable
       new  e5f71e7   Merge pull request #5742 from amueller/pdf_doc_link
       new  ae8db69   Fix import of reload for python 3.3
       new  b767117   Use six to detect integral numbers
       new  04ca448   Typo.
       new  253e387   Merge pull request #5755 from rvraghav93/fix_depr_version
       new  299c275   BUG :#5782  check_classification_targets returns y instead of y_type
       new  1d16ec4   tests sample_weights linearreg, ridge
       new  bc3c3db   #5782 added test case for check_classification_targets()
       new  a5d6144   Merge pull request #5526 from giorgiop/fix-weight-linearregression
       new  889e2d4   Merge pull request #5678 from betatim/no-warning-iforest
       new  f49a558   Merge pull request #5492 from arthurmensch/remove_cython
       new  3fc29cd   nose-timer added to CI
       new  6e87813   Scaling a sparse matrix along axis 0 should accept a csc by default
       new  7cb9a23   Merge pull request #5733 from giorgiop/trevis-with-timer
       new  8a391be   added check for type in test_check_classification_targets
       new  ea7c5c3   fixed memory / efficiency issue in DummyClassifier(strategy='most_frequent')
       new  ce90ccb   Merge pull request #5796 from max-moroz/dummy_classifier_memory_efficiency
       new  fe4065f   Added a related PMML serialization project
       new  654bddd   FIX Explicitly specify the plot colors for each curve
       new  c6f586d   Fixed transformation of test set via RobustScaler in plot_robust_scaling.py
       new  8da0827   Distinguished more clearly between two PMML serialization projects
       new  c6d6170   #5766: Removed unused code from plotting tutorial
       new  a161f2c   predictions are float but np.bincount can only be used on integers, thus an error comes out when using it
       new  b065099   Merge pull request #5801 from hlin117/pca-plotting
       new  81e8ef7   Explained implementation details
       new  d21a297   Merge pull request #5798 from vruusmann/patch-1
       new  10445f5   Removed exit() call in topic extraction example
       new  1fe8014   Update related_projects.rst
       new  5bba3d0   Merge pull request #5806 from vighneshbirodkar/nmf_fix
       new  d644e8e   Merge pull request #5791 from MechCoder/scale_csc
       new  0289da4   Reduced size of line 167.
       new  a9858b1   debian/copyright - extended list of copyrights/licenses (Thanks Thorsten Alteholz for review)
       new  d44f939   Merge pull request #5752 from hrjn/fix-5694-proper
       new  0820731   Merge pull request #5783 from varun-kr/check_classification_bug
       new  6646e5d   Merge pull request #5809 from lmcinnes/doc_related_hdbscan
       new  5fa8b32   Merge pull request #5754 from rvraghav93/deprecated_mpl_15
       new  3c1e900   #5751: Adding explained_variance_ and clarification about components_ to docstring
       new  7fd4df8   Merge pull request #5819 from hlin117/pca-docs
       new  a20b0eb   Merge pull request #5802 from alvarouc/master
       new  3477df2   added warm_start flag to example
       new  3aeacc1   added cython to requirements in documentation
       new  2c298d0   add docs
       new  7b09113   edited cython to requirements in documentation
       new  09c3fbf   Merge pull request #5771 from bhargav/minor-doc-fixes
       new  359432c   edited cython requirement in README making obvious that it is for dev
       new  14a124b   Merge branch 'dev_requirement_documentation' of https://github.com/hugobowne/scikit-learn into dev_requirement_documentation
       new  8867e83   edited bleeding edge install instructions to include cython
       new  1f553b0   #5840: Adding my name to one of the bug fixes
       new  cb88239   Merge pull request #5846 from hlin117/whatsnew
       new  998848a   fix for #5725 (makes sure numpy array is not compared to string, which will fail in future versions of numpy)
       new  34fda74   minor grammatical fix
       new  6ec1c38   Change super call to pass presort value to super
       new  19f109c   The presort parameter is never acutally used
       new  221e7ce   Merge branch 'master' of github.com:scikit-learn/scikit-learn into gradientboost
       new  1b72433   Fix grammar in rbm docstring
       new  8558696   Merge pull request #5860 from pwgerman/rbm_doc_hidden_units
       new  2738d59   fix for missing classes found in y #4327
       new  54fa7aa   Adding email to the mailmap
       new  ad05d3d   Merge pull request #5865 from hlin117/mailmap
       new  2900c5a   Merge pull request #5855 from pieteradejong/doc-edits-transforms
       new  bc90ba0   Merge pull request #5831 from codevig/mybranch
       new  c873d51   DOC: clarify terms in documentation of Spectral Clustering
       new  2d7b5cf   DOC: further clarify spectral clustering doc
       new  bab2233   Add MLP to what's new
       new  9b5cf0f   #5877 fix confusing n_samples and n_features
       new  4d2fe36   dict.has_key() deprecated
       new  0101e2c   Merge pull request #5885 from sieben/has_key_deprecated
       new  eb23f70   Update README.rst
       new  f59dbe7   Update advanced_installation.rst
       new  bf00574   Merge pull request #5834 from hugobowne/dev_requirement_documentation
       new  55c3a3b   REL add 0.17.0 release to news, fix versions in menu
       new  a05bc8d   Merge pull request #5858 from mcculloh/gradientboost
       new  845a7a1   MAINT Use with to handle file
       new  fa45329   minor grammatical fix
       new  a72d138   Merge pull request #5828 from mth4saurabh/add-mlp-to-changelog
       new  273f1cc   Merge pull request #5895 from toastedcornflakes/master
       new  fb123ed   More doc fixes. Latex builds again.
       new  0159bc4   Merge pull request #5728 from amueller/more_doc_fixes
       new  381315d   Merge pull request #5886 from sieben/with_statement
       new  79c7583   Add user guide link to lda
       new  75d9cea   Update text.py
       new  e91dea3   Merge pull request #5896 from fayeshine/fayeshine-patch-1
       new  29e7e39   update ambiguous out of bag generalization terminology - from 'error' to 'accuracy' or 'the R^2...'.
       new  349eaca   DOC: fix shape lambda_ attribute in BayesianRidge
       new  6325173   Merge pull request #5913 from mekman/doc_fix_minor2
       new  8991bd9   website: added NYU funding to front-page footer and the about section.
       new  e17c4de   Merge pull request #5910 from amueller/nyu_funding
       new  b043d64   DOC: Comment better the importance of the mask
       new  91971be   add .tags in gitignore
       new  c8fb696   Merge pull request #5850 from JeremyNixon/jeremy-feature
       new  f0f5646   Merge pull request #5892 from mth4saurabh/add-user-guide-link-to-lda
       new  b59b0e1   Merge pull request #5882 from LeiG/fix_confusing_n_samples
       new  48ad7aa   replaced matplotlib function finance.quotes_historical_yahoo() (deprecated as of matplotlib 1.4.x) with finance.quotes_historical_yahoo_ochl(), enabling the example to work for the latest version of matplotlib (1.5.0).
       new  335536c   Merge pull request #5919 from nelson-liu/fix_stock_example
       new  1606c90   check_is_fitted validation for transform() and inverse_transform()
       new  f63ac6e   Merge pull request #5936 from olologin/MultiLabelBinarizer_check_is_fitted
       new  8a6be6b   Inconsistent indentation
       new  72d72bf   LDA: explained_variance_ratio_ can be found using svd solver
       new  114920e   edited docstring for clarity regarding the naming of the pipeline components.
       new  ba2e638    Fix coveralls report sending
       new  7602a3f   It's better to fix a comment
       new  ff31d19   repaired dead link
       new  6f7e351   residual sum of squares should be computed by summing the squared differences instead of taking the mean.
       new  50dfa86   change comment, printed comment, and code behavior to all reflect computation of mean squared error
       new  62b48b9   Merge pull request #5954 from KjellSwedin/correct_LinearRegression_example
       new  9f136ff   ENH add multinomial SAG solver for LogisticRegression
       new  43e5454   Merge pull request #5251 from TomDLT/sag_multi
       new  54b832e   Add link to multiisotonic
       new  f8f0d2d   Merge pull request #5962 from alexfields/master
       new  cecb561   Merge pull request #5944 from nelson-liu/fix_make_pipeline_doc
       new  108a0c2   Added default gamma info for laplacian_kernel and rbf_kernel
       new  159e51e   Added more gamma info for sigmoid, polynomial kernels
       new  0340c75   Changed orders of docstring descriptions for polynomial_kernel
       new  691e39d   Merge pull request #5966 from hlin117/pairwise-docs
       new  52caed6   Merge pull request #5949 from olologin/plot_svm_anova.py-and-"all-CPU's"
       new  331ca4a   debian/control - recommend python3- versions of packages for python3-sklearn
       new  d13f519   Parallelize embarrassingly parallel loop in RFECV.fit
       new  4cac986   Merge pull request #5784 from MechCoder/parallel_rfecv
       new  8d1ed03   removed the doc line 'Only applies if analyzer == 'word'' from CountVectorizer analyzer'
       new  fe4dbe3   Fix Kernel PCA docstring to reflect how remove_zero_eig defaults to false
       new  95940b5   Merge pull request #5980 from dshieble/Odd_docstring_for_CountVectorizer_analyzer
       new  fdf6648   Merge pull request #5634 from sieben/space_indentation
       new  0e5c194   fix link to contributing in website drop-down
       new  b1ecb7f   fix link to contributing in website drop-down
       new  eb48371   Merge pull request #5829 from joshuacook/repair_link
       new  21cf29c   Fixed unit test in sklearn/tree/tests/test_tree.py.
       new  4d1a466   Merge pull request #5990 from jblackburne/fix-test_no_sparse_y_support
       new  8810716   adding missing 0.16 doc link and removing broken support links
       new  5c0ea1e   Merge pull request #5923 from fabioticconi/add_link
       new  f2ac3db   adding missing 0.16 doc link and removing broken support links
       new  a680043   Merge pull request #5800 from fabioticconi/patch-1
       new  05d2161   Merge pull request #5890 from amueller/post_release_dev_doc_fixes
       new  89804c4   adding 0.17 support link
       new  34ebbd3   Merge pull request #5994 from fabioticconi/add_link
       new  1680b05   Merge pull request #5925 from lesteve/fix-coveralls
       new  9f18b94   Merge pull request #5981 from seales/master
       new  5b1aca0   MAINT Don't use deprecated 1d X (or deprecated matplotlib stuff) in examples.
       new  3bacb19   Merge pull request #5942 from hlin117/lda_svd
       new  c33fa15   Merge pull request #5773 from amueller/python3.3_fix_again
       new  6e513b6   DOC: Fix spelling
       new  82981fe   Merge pull request #5999 from jseabold/doc-fix
       new  c068ee8   update file header in affinity_propagation.py
       new  c5469c9   Merge pull request #5706 from shawpan/Fix-#5702-test-warnings-in-ranking
       new  2bbeea1   Merge pull request #6004 from leosartaj/affprop
       new  ff1ca88   Merge pull request #5721 from amueller/examples_deprecation_fixes
       new  1d4c926   Merge pull request #5578 from waterponey/circleCI_push
       new  f9ae546   HOTFIX skip partial dependence plot so we can run circleci without errors.
       new  51addc0   BUG: Fixed fetch_kddcup99 for Python 3
       new  79006fa   Provide 'self' argument to instance method
       new  e4d8988   \#6001: Adding LogisticRegression's default param for 'solver'
       new  58c386c   Merge pull request #6006 from hlin117/logistic_docs
       new  2f76310   Merge pull request #5852 from nickyj88/master
       new  8d0a299   Merge pull request #5946 from nmayorov/fetch_kddcup99_bug
       new  37c8d02   HOTFIX actually fixing the california housing problem
       new  7d6a22a   Merge pull request #6018 from proinsias/provide_argument
       new  3e4ebac   I guess it is some small mistake
       new  4b122ad   Merge pull request #6034 from alexloginov/patch-2
       new  d1800df   Fix minor typo in RandomizedSearchCV documentation
       new  2707de7   install latex required for docbuild on circleci
       new  c865b9b   Using rain_test_split and StandardScaler instead of reimplementing them
       new  eb464fb   Merge pull request #5730 from hdmetor/bugfix/5720
       new  61a58c5   Fix omitted word in SVM-Anova example
       new  66fd680   Merge pull request #6042 from kputnam/patch-2
       new  625471b   DOC: few minor typos spotted
       new  37bfe80   FIX Changed zorder so support vectors are clearly visible
       new  0e074a0   Fix bug in mini batch kmeans example
       new  2f5c86f   Merge pull request #6046 from davidthaler/plot_krr_fixup
       new  a06ddc6   Merge pull request #6044 from yarikoptic/master
       new  0a4b87c   MAINT fix the coveralls badge
       new  a8dc3a3   Merge pull request #6050 from ogrisel/coveralls-tweaks
       new  9648351   MAINT fix data home dir in circle.yml
       new  9790bc6   Merge pull request #6036 from kputnam/patch-1
       new  0485ada   General spelling fixes
       new  1d8e574   FIX #5485 Replaced missing plot_gp_regression with plot_gpr_co2 in carousel.
       new  776323b   edited FAQ to include info about deep learning
       new  167c1cb   Merge pull request #6060 from nelson-liu/add-deep-learning-to-faq
       new  fde69cb   Fixes the link to Bawa et al., "LSH Forest: Self-Tuning Indexes for Similarity Search", WWW'05 in the Nearest Neighbor section.
       new  e2f8cd2   Merge pull request #6067 from IsaacHaze/lsh_link_fix
       new  fc31b64   doc/fix: edited inaccuracies in nearest centroid classification example
       new  318b93d   Implemented keyword arguments in FunctionTransformer
       new  cd53441   Update whats_new.rst
       new  c2cee91   Merge pull request #6048 from zhengqm/fix-cluster-example-bug
       new  ad758d2   FIX: Cluster centers after fir in KMeans and MBKMeans need not be ordered in the same way
       new  a747d32   Merge pull request #6069 from MechCoder/mbkmeans_fix
       new  6ff0bbe   Merge pull request #6068 from nelson-liu/rework_nearest_centroid_classification
       new  4b21c08   debian/rules - boost minimal Cython version for recythonization to 0.23.1.  Should enable back bakckports on older Debian/Ubuntus
       new  2999a2f   * debian/tests/deb_tests_lower_precision - to lower precision of test_feature_importances to avoid FTBFS on some (BE) platforms
       new  3c546fd   debian/tests/deb_tests_lower_precision - lowering further to 4 decimals
       new  6bf63f6   Kullback-Leibler divergence as a parameter of TSNE
       new  112d7c5   Merge pull request #6084 from ogrisel/rebased-pr-3804
       new  280840f   FIX boolean indexing in test_label_kfold
       new  ea65e82   FIX deprecation warning for inspect.getargspec
       new  7cfa554   Use kwargs instead of functools.partial to define some scorers
       new  80c770f   FIX Eliminated the use of color_cycle in out of core example
       new  e152df2   DOC add version_added for multinomial sag
       new  68cf314   fixed target link to isotonic regression example
       new  5b8fc15   Smoke test the repr of scorers.
       new  1c30882   Use items instead of iteritems for python 3 compatibility.
       new  8f03dff   Merge pull request #6088 from duboism/solve6078
       new  f8d7d5f   Merge pull request #6097 from mrandrewandrade/master
       new  2ff9a4b   Remove useless normalization — fixes #6075
       new  0d597c0   Solves issue #5715: enables pandas input to log_loss
       new  4afbcce   Merge pull request #6104 from maniteja123/issue5715
       new  fb3574d   Added partial_fit method for ovo and ovr FIX issue #4167 implementing online learning where base estimators have patial_fit method.
       new  76c1eb2   First pull request of 2016!
       new  20f6866   Merge pull request #6107 from MechCoder/2016
       new  cdb9dba   Added doc, tests Added test for partial_fit when mini-batch doesn't have all target classes. Changed doc for better explaination of the use of  paramter in partial_fit method. Used generator for passing of classes in OvO in Parallel processing instead of estimator.pop().
       new  003496f   Remove extra `+` in document
       new  e464689   Merge pull request #6108 from chezou/chezou-patch-1
       new  1301692   Removed joblib comment, made compatible for python3 Removed the outdated comment about joblib overahead of thread spawining issue. Made izip compatible for both python2 and python3.
       new  3daab8a   FIX fetch_california_housing
       new  317dea8   Merge pull request #6005 from seales/SpellingFix
       new  eb4bd09   fix pplot example for uneven grids
       new  41035a0   ENH less verbose output in plot_partial_dependence example
       new  95a1437   DOC clarifies that sparse matrices are scipy
       new  7c425ab   base_estimator_ attribute is a single estimator not a list
       new  9df8a1a   add warning, new images and links in MLP docs
       new  682fd8e   Merge pull request #6102 from toastedcornflakes/denoising_doc_fix
       new  b269f2f   Change shape of target in fetch_lfw_pairs docstring to 2200 for consistence with data and pairs shape
       new  c571635   Merge pull request #6111 from betatim/base-not-a-list
       new  604d54c   Changed izip import statement, importing from six moudle
       new  7e497ba   Fixed the broken nose link in advanced installation docs.
       new  fc7192a   Add missing pairwise metrics to classes.rst
       new  d26e412   Merge pull request #6119 from Qwasser/fetch_lfw_pairs-target-shape
       new  51615b3   BUG: fix initialization in Isomap to work correctly with GridSearch
       new  99c73a8   DOC Reworded cv documentation
       new  116f42b   Fixed doc in lfw.py. Fixes #6136. Added dependence statement for data, target and pairs. Fixed minor pep8 issues.
       new  a228016   Leading underscores for warnings imported from sklearn.exceptions in other modules where warnings are deprecated.
       new  ad193c1   Merge pull request #5833 from welch/issue-5581-doc-warm-start
       new  8d17de0   Fix hyperparameter check error
       new  936ea63   Make ridge path with ill-conditioned matrices clearer
       new  bd97169   Made few tweaks based on commemnts changed docstring of partial_fit, added check_classification_target in ovo, removed classes parameter in _partial_fit_binary.
       new  eaaf2ae   Merge pull request #6148 from MechCoder/ridge_cleare
       new  aecf7bc   Merge pull request #6087 from kaichogami/binary_partial_fit
       new  49330b6   COSMIT: whitespace
       new  729be74   Use row_norms in f_regression
       new  739616a   Update whats_new for partial_fit OvR and OvO
       new  3ce41b6   Addressed comments
       new  a7e7c55   Merge pull request #6139 from dsquareindia/fetch_lfw_doc_fix
       new  9358777   Merge pull request #6150 from MechCoder/fix
       new  c7d3e0e   ENH Add sample weight support to matthews_corrcoef metric
       new  de25a6b   DOC fix FAQ link at the bottom of main page
       new  528544b   Merge pull request #6157 from lesteve/fix-faq-link
       new  4c7619b   Merge pull request #6146 from toastedcornflakes/MLP_fix
       new  6b53672   Merge pull request #5219 from scikit-learn/authors-update
       new  41526cb   Added sample_weight parameter to ransac.fit
       new  761b1f7   Deprecate residual_metric and add support for loss
       new  ac2ff4a   Merge pull request #5497 from MechCoder/ransac_residual
       new  c7ce0ab   MAINT: Refactor and speed up silhoutte_score
       new  1d9186d   ENH/FIX std error is computed for mean across a set of 3 scores, not len(X)
       new  22c687b   polish figure
       new  769127c   Merge pull request #6151 from MechCoder/silhouette_score_refactor
       new  2255ff5   TST/FIX Make model selection tests warning free!
       new  6f6d878   Upgrade joblib to 0.9.4 to fix important bugs
       new  7ff8d60   DOC entry for joblib 0.9.4 upgrade
       new  01b5b7b   Added sample weight support to confusion matrix.
       new  1487047   update url of ARD reference
       new  4e4e6c3   Merge pull request #6187 from mth4saurabh/update-paper-url
       new  d3175db   Ridge coefficients as a function of regularisation
       new  dfa516c   Merge pull request #6179 from ogrisel/joblib-0.9.4
       new  effb919   Upgrade joblib to 0.9.4 to fix important bugs
       new  02fe943   DOC entry for joblib 0.9.4 upgrade
       new  98c6edf   DOC add links to some issues in whatsnew
       new  6a7eae5   TST: More tests for Ledoit-Wolf
       new  6148686   SKF raises error if n_labels<n_folds for individual classes. Updated whats_new
       new  df08def   ENH: Implemented mutual_info function
       new  54c0783   DOC: Documentation update related to mutual_info
       new  0245fc8   MAINT: Use six.moves.zip in mutual_info
       new  c1aea3f   MAINT: Renamed module mutual_info to mutual_info_
       new  689ed0d   DOC: Example for mutual_information
       new  835102a   API: Split mutual_info into _regression and _classif
       new  8394c1b   MAINT: Add blank lines between parameters in mutual_info_.py
       new  ad2f5f5   MAINT: Add check_classification_targets to mutual_info_classif
       new  ffc4fe9   TST: Change tolerance checks in test_mutual_info.py
       new  824dda3   MAINT: Small changes to plot_f_test_vs_mi.py
       new  051d3a2   MAINT: Slightly improve logic of discrete-continuous MI estimation
       new  7869992   MAINT: Slightly improve copy logic in _estimate_mi
       new  ec17289   DOC: Add short descriptions of methods for mutual info estimation
       new  b0491be   DOC: Add a short explanation of F-test vs MI in narrative doc
       new  d3a497a   BUG: Fix copy logic for mutual info functions
       new  375b070   TST: Speed up 2 tests related to mutual info
       new  d60636a   DOC: Small fixes in mutual_info_.py documentation
       new  094a077   MAINT: Small refactoring in mutual_info_.py
       new  5b3f515   MAINT: Get rid of classes in test_mutual_info.py
       new  b48a108   DOC: Add one more reference for mutual info methods
       new  e1bc056   MAINT: Add a clarification comment in mutual_info_.py
       new  a36edf2   DOC: Modify SelectKBest and SelectPercentile docstrings slightly
       new  e716c64   MAINT: Mention mutual info methods in whats_new.rst
       new  daa73c7   BUG: Remove non-ASCII symbols from mutual_info_.py
       new  3d0c8f8   MAINT: Modify whats_new item related to mutual information
       new  f3e256f   DOC add jmetzen and tomdlt to authors.rst
       new  fa76b2e   Merge pull request #6208 from TomDLT/authors-update
       new  50b6523   FIX: doc KernelRidge
       new  38f9e1c   Merge remote-tracking branch 'upstream/pr/6182'
       new  419c925   Merge pull request #6212 from kingjr/fix_kernel_ridge_doc
       new  65ab525   Added link to profile in whats_new.rst
       new  4022c68   minor fix in RandomizedLasso API docs
       new  21d9ccc   Merge pull request #6213 from dsquareindia/add_profile_link
       new  74475bc   LabelEncoder now raises error for 0-D arrays
       new  c255a78   ENH: classification_report format supports long string labels
       new  66cf395   Merge pull request #6229 from dsquareindia/LabelEncode-error-msg-fix
       new  30e83cf   Merge pull request #6214 from mth4saurabh/minor-fix-RandomizedLasso-doc
       new  c98d5c5   Make doc for parameter consistent
       new  006aacb   MAINT: Print info message for fetch_20newsgroups
       new  fbe85eb   update note on Input Data in clustring docs
       new  ae70c55   Fix reading of bunch pickles
       new  7371c33   Added changelog entry about #6196
       new  7ed1ab3   Merge pull request #6227 from strixcuriosus/classification-report-format
       new  5a0ee9b   Merge pull request #6201 from GaelVaroquaux/more_ledoit_wold_tests
       new  05da5db   Change super call to pass presort value to super
       new  3c4ddb0   DOC fix FAQ link at the bottom of main page
       new  887a069   fix for #5725 (makes sure numpy array is not compared to string, which will fail in future versions of numpy)
       new  201e292   Typo
       new  8db4e83   Merge pull request #6235 from lesteve/fix-reading-of-bunch-pickle-generated-with-0.16
       new  98622be   Merge pull request #6230 from yenchenlin1994/fix-Hierarchical-parameter-doc
       new  cb1a4bd   Merge pull request #6238 from amlankar/patch-1
       new  84ea56c   Merge pull request #6218 from mth4saurabh/update-input-data-note-clustering-docs
       new  3569e27   Fix reading of bunch pickles
       new  3933ff4   Added changelog entry about #6196
       new  3030ad6   TST non-regression test for #6147, roc_auc on memmap data
       new  9c40e85   FIX 6147: ensure that AUC is always a float
       new  3427d20   DOC add fix for #6147 to the changelog
       new  c0e4832   remove shuffling and standardization fix layout
       new  1b4a9f2   Merge pull request #6225 from ogrisel/fix-6147-scoring-memmap
       new  b50d3ed   Fixes make_sparse_spd_matrix documentation
       new  84ef805   Merge pull request #6244 from harrymvr/patch-1
       new  613f1ad   TST non-regression test for #6147, roc_auc on memmap data
       new  497ddf3   FIX 6147: ensure that AUC is always a float
       new  3bfbdb0   DOC add fix for #6147 to the changelog
       new  2e270a7   DOC missing changelog entry for #5857
       new  96d8e3c   DOC missing changelog entry for #5857
       new  bb0a632   move title in the bottom
       new  c3d8493   Rewordin algorithm requirements in FAQ
       new  51480fc   Merge pull request #6152 from halwai/faq
       new  e20e003   Fix axis label of multioutput regression tree example
       new  0a3eaf5   Fix typo in parameter's doc
       new  da5d898   Merge pull request #6241 from louridas/fix-iris-decision-tree
       new  79dea7b   BUG: fix pickling of DistanceMetric instances (fixes #6269(
       new  4ae03d1   Merge pull request #5701 from ziky90/gp-comparison-fixes
       new  d1fb39e   added LatentDA to the Incremental learning page
       new  37080d2   Merge pull request #6272 from jakevdp/fix6269
       new  46ec927   Merge pull request #6277 from rasbt/latendDAupdate
       new  864f1c0   Merge pull request #6264 from yenchenlin1994/fix-multioutput-regression-tree-example
       new  070ff21   Merge pull request #6232 from maniteja123/fetch_20newsgroups
       new  637cebb   FIX restore previous title layout in examples/tree/plot_iris.py
       new  9d0a489   Merge pull request #6267 from yenchenlin1994/fix-LDA-doc-error
       new  4f2cd80   raise NotFittedError in VotingClassifier and unit test improvements
       new  a94c04b   Merge pull request #5794 from rasbt/check_votingclassifier
       new  aec8dd0   Add doc to clarify the use of DictVectorizer when categorical features are represented as numeric values
       new  d472a6a   much faster isotonic regression prediction (involved re-setting interpolation to linear)
       new  c694b9b   change to test_isotonic_regression_ties_min
       new  956205c   Merge pull request #6286 from ogrisel/pr-6206-followup
       new  604e1e0   Fix check_cv error message
       new  ef307ec   Correct typo in ``decomposition.rst``. Merge latest changes.
       new  8d270fb   MAINT speed up test_gpc
       new  dd6ab4e   MAINT speed up test_gpr
       new  6470a83   MAINT speed up test_kernels
       new  6eca250   MAINT speed up test_extmath
       new  9ee3f50   MAINT speed up test_coordinate_descent
       new  d3eb01e   MAINT speed up tests in test_neighbors
       new  1814e84   MAINT speed up test_split
       new  58c859e   Limit the number of thread and force install libgfortran
       new  23f5105   Merge pull request #6270 from ogrisel/pr-5711-followup
       new  19024a8   FIX performance issue in _graph_connected_component
       new  b0d071b   Merge pull request #6268 from ogrisel/fix-spectral-embedding-stopping-condition
       new  4e56161   MAINT display pip version in appveyor
       new  8eacee4   MAING Upgrade pip in appveyor
       new  2b8904d   Merge pull request #6299 from ogrisel/fix-appveyor-pip
       new  9cbceb3   MAINT split test_ridge.test_dense_sparse with generator
       new  795b051   Merge pull request #6296 from pestrickland/master
       new  1aa0ec2   STYLE missing whitespace in isomap comment
       new  83d1cf4   MAINT speed up slow test_iforest_sparse
       new  cfa498c   FIX in randomized_svd flip sign
       new  bf99bc8   DOC: Use SVG badge for Travis CI. [ci skip]
       new  23a0895   Remove redundant and
       new  2d6f68a   Removes X= after raise.
       new  7475bea   MAINT Upgrade pip in appveyor and display version
       new  cb0f216   MAINT fix / speedup travis on 0.17.X
       new  538e650   Merge pull request #6311 from jakirkham/badge_svg_travis
       new  2311bab   Merge pull request #6315 from ogrisel/ci-0.17.X
       new  706b45e   Merge pull request #6314 from tshauck/remove-superfluous-declaration
       new  ebe3ff5   fix math in soft VotingClassifier example
       new  71e25e6   Merge pull request #6319 from gjreda/fix-soft-voting-docs
       new  f7f82b9   Merge pull request #6293 from yenchenlin1994/fix-doc-typo-in-binary-tree
       new  b7998c2   Merge pull request #6292 from yenchenlin1994/fix-error-message-in-check-cv
       new  a8e7e48   Merge pull request #6285 from yenchenlin1994/update-DictVectorizer-doc-about-one-hot-encoding
       new  117cf5e   Merge pull request #6058 from davidthaler/carousel_fix
       new  7f70390   typo fixes in the current changelog
       new  64dfddf   Merge pull request #6294 from rasbt/whats_new_typos
       new  856424a   fix joblib error in LatentDirichletAllocation
       new  f67e52d   Merge pull request #5945 from mth4saurabh/improve-mlp-docs
       new  268af69   Merge pull request #6324 from chyikwei/fix-lda-joblib-error
       new  89c918a   DOC whats_new entry for #6258
       new  f6ba9ff   fix joblib error in LatentDirichletAllocation
       new  ac7450a   DOC whats_new entry for #6258
       new  9c511d0    Changes to be committed: 	modified:   examples/applications/svm_gui.py
       new  f393f59   Normalize LDA transform's return value
       new  2dc389c   doc:added multilabel y support for all splitters
       new  36ec92a   Fixed typo in one word.
       new  71ec574   DOC: Fix redundant doc for f_regression.
       new  55f04f9   Merge pull request #6331 from joerobmunoz/fix-model-evaluation-typo
       new  b7b6ac1   Merge pull request #6040 from amueller/circle_texlive
       new  fcfaa5f   Updated examples and tests that use scipy's lena
       new  d6ebe74   Merge pull request #5880 from jakevdp/spectral-doc
       new  88b96a7   Merge pull request #5675 from AlexanderFabisch/gmm_clustering
       new  1784bf3   Merge pull request #5863 from trevorstephens/cw_util_fix
       new  3280f88   Merge pull request #5425 from amueller/update_doc_readme
       new  6c62763   Fix test_label_binarizer on Windows
       new  c4e0ce6   FIX/NRT signature expects the init to not be a c-defined method (Py2.7)
       new  885c512   Add sklearn/preprocessing/tests to source distribution
       new  4efec8f   DOC: incorrect phrasing of inputs to roc
       new  c761ba7   Merge pull request #6260 from ogrisel/pr-5920-followup
       new  a186014   Merge pull request #6328 from rvraghav93/model_selection_repr_fix
       new  8a326f8   Updated examples and tests that use scipy's lena
       new  dd85a67    Add parallelism to multiple algorithms
       new  a6604c3   inverse_transformation for NMF decomposition
       new  630541e   Remove redundant that in test_ranking.py
       new  f7bcfc5   Merge pull request #5872 from lucasdavid/parallel-nn-in-isomap
       new  9522916   Merge pull request #6339 from yenchenlin1994/fix-typo-in-test_ranking
       new  80bb29b   Fix memory leak in Barnes-Hut SNE
       new  56d9c75   MAINT simplify freeing logic for Barnes-Hut SNE memory leak fix
       new  6b1d351   Merge pull request #6125 from AnishShah/issue6118
       new  04f2857   Remove redundant assignment of parameter shuffle
       new  7cd0419   Merge pull request #6327 from yenchenlin1994/fix-lda-transform-output-probability
       new  97015de   Merge pull request #6336 from GaelVaroquaux/fix_roc_auc_doc_phrasing
       new  6ffc85e   BLD: fix build issue introduced by gh-5492.
       new  d2b6c7a   Fixed legend text and pep8 warnings
       new  a1e546b   DummyClassifier: Added doc information about default strategy
       new  ff97d74   Merge pull request #6362 from hlin117/dummy
       new  5404d5d   Merge pull request #6360 from alyssaq/patch
       new  ce2a03e   Add numpy dev wheel to travis build matrix.
       new  b5286c3   weird work-around for spearman test
       new  9bff7d6   add test for auto increasing
       new  e6cc861   fix floating point indexing in generic univariate
       new  de38d1b   make X_stride be np.npy_intp
       new  1e5820b   Semantics of ``assert_almost_equal`` now consistent on floats and arrays. see https://github.com/numpy/numpy/issues/5200
       new  2270af2   Merge pull request #6354 from rgommers/fix-build-error
       new  aa2f659   DOC: very minor typo
       new  e98a11b   Merge pull request #6366 from yarikoptic/master
       new  a43a9f5   MAINT clean up install script
       new  c36c4ca   MAINT more informative output to circle/check_build_doc.py
       new  0e06be5   FIX check_build_doc.py false positive detections
       new  b35345c   Merge pull request #6343 from lesteve/fix-tsne-memory-leak
       new  765f77e   Read web pages as UTF-8 by default. Use with-blocks to open files.
       new  e2e773d   Add test for init errors
       new  69e4c14   FIX in randomized_svd flip sign
       new  53713c9   Fix for issue #6352
       new  ffafc80   Merge pull request #6333 from lesteve/fix-test-label-binarizer-on-windows
       new  2d8eb00   Fix broken link of Wainwright's paper
       new  4694813   Merge pull request #6384 from yenchenlin1994/fix-broken-link-in-sparse-example
       new  140a5ac   MAINT depr of center_data, normalize in linear_model
       new  faa63df   Remove redundant assignment of variable tol
       new  1b27536   Change comment formula of enet_coordinate_descent_gram
       new  234e4ab   Merge pull request #6391 from yenchenlin1994/remove-redundant-assignment-in-test-locally-linear
       new  a7038a4   Merge pull request #6363 from ogrisel/pr-6332-followup
       new  f679095   Merge pull request #6375 from jakirkham/backport_fix_svd_sign_flip
       new  10f5db0   FIX fetch_california_housing
       new  dd3d721   MAINT more informative output to circle/check_build_doc.py
       new  58ba2b8   FIX check_build_doc.py false positive detections
       new  8007d91   Fix memory leak in Barnes-Hut SNE
       new  f3e6609   MAINT simplify freeing logic for Barnes-Hut SNE memory leak fix
       new  7477df2   FIX cythonize TSNE
       new  807796e   MAINT remove non-existing cache folder in 0.17.X branch
       new  368a98f   ENH/MAINT Check for changes in pxd files too. Cleanup code
       new  b93ccb9   ENH/MAINT Save hashes once for every file
       new  ea042f1   Release 0.17.1
       new  65a2b8f   Fixed codestyle
       new  bf9cd88   Merge pull request #6254 from rvraghav93/cythonize_when_pxd_changes
       new  4b7a0e0   enh: reorganize build files
       new  d963b2f   Add test for local linear embedding
       new  281ac0c   Merge pull request #6393 from yenchenlin1994/add-test-for-local-linear-embedding
       new  c2b1edd   ENH use n_jobs in brute-force radius_neighbors
       new  55dfe49   Merge tag '0.17.1' into releases
       new  1452704   Merge branch 'releases' into dfsg
       new  82c41fb   Merge branch 'dfsg' into debian
       new  baf20b4   changelog entry + adjusted years in copyright
       new  69b8aa1   few patches picked up from master for safe passage on sid
       new  2141869   Merge pull request #6396 from nelson-liu/reorganize_build_files
       new  4a5a7e1   Merge pull request #6349 from yenchenlin1994/remove-redundant-shuffle-assignment
       new  3fa684d   add get_feature_names to PolynomialFeatures
       new  ddc1740   fix PolynomialFeatures.powers_ in python 0.16.1
       new  22f468a   Update svm.rst
       new  1601dae   Fixes the 'make cython' by adjusting the path to cythonize.py
       new  c9d66db   Merge pull request #6414 from ssaeger/issue_6413
       new  897f8b6   don't do ^1
       new  f723338   Update svm.rst
       new  c543633   Update cross_validation.rst
       new  46fc1be   Remove redundant words in sklearn
       new  9f4b53e   Update svm.rst
       new  e631f2c   Merge pull request #6412 from Naereen/patch-3
       new  3f88b98   doc: fix capitalization in svm.rst citation to match paper.
       new  613e8be   Merge pull request #6427 from nelson-liu/fix_svm_doc_citation
       new  323bcea   Merge pull request #6426 from yenchenlin1994/remove-redundant-typo
       new  b2799ca   DOC: Added memory workaround to dbscan doc
       new  f45f374   Fixed typo
       new  cef770e   Made seperate topic for mem consumption for large sample sizes
       new  d7d0ae6   Merge pull request #6404 from dsquareindia/dbscan_doc_enh
       new  28b58b0   Merge pull request #6442 from radarhere/patch-1
       new  8fb928d   fixed doc for powers, added test
       new  f5025c3   Correct PCA and RandomizedPCA docs related to whitening
       new  b6adb92   Merge pull request #6447 from MechCoder/whiten_pca
       new  7895d38   Merge pull request #6372 from amueller/poly_feature_names
       new  7e79adc   Huber regressor
       new  540c7c6   Merge pull request #5291 from MechCoder/huber_loss
       new  ca4c727   Fix misuse of it's
       new  90f2c18   Merge pull request #6453 from yenchenlin1994/fix-its-typo
       new  07728d9   FIX StratifiedShuffleSplit train and test overlap
       new  bff21be   Add doc of attributes for VotingClassifier
       new  f300632   DOC: Align the code blocks in CONTRIBUTING.md with tab
       new  8d42471   MAINT: Fix typo in comment for travis/install.sh
       new  bba6826   Merge pull request #6461 from scottclowe/fix-typo
       new  0de2cb5   Merge pull request #6462 from scottclowe/doc-align-contrubting-code-indents
       new  68de3b2   Merge pull request #6458 from yenchenlin1994/add-doc-for-voting-classifier
       new  1049642   Merge pull request #6371 from yenchenlin1994/add-test-for-vote-classifier
       new  f0f0aff   DOC: add issue and pull request template
       new  150afe6   Merge pull request #6379 from lesteve/fix-stratified-shuffle-split-train-test-overlap
       new  c60c3f0   Merge pull request #6411 from nelson-liu/create_issue_pr_template
       new  7872126   Corrected typos in cross_validation.KFold docstring
       new  d3d8490   Merge pull request #6469 from ralexx/docstring-typos
       new  6d3e87a   DOC: Clarify the scoring in Logistic Regression
       new  94e2768   Merge pull request #6481 from maniteja123/logistic-doc
       new  2e4aafd   MAINT remove my email address
       new  6a841a1   doc: add link to user guide in isolation forest docstring
       new  03a1d45   Merge pull request #6490 from nelson-liu/fix_isolation_forest_docstring
       new  d6e7f6b   Add Python and PyPi version badges
       new  3d001f3   Fix MLP batch_size test warning for default value
       new  1d50534   Merge pull request #5699 from bhargav/mlp-test-warning
       new  f9b4e7e   don't import train_test_split from deprecated cross_validation module in mlp
       new  c0bc2f8   Merge pull request #6503 from rhiever/patch-1
       new  2d19895   MAINT temporary disable MKL in Travis
       new  f27c3b2   Merge pull request #6508 from lesteve/travis-nomkl
       new  3f39676   MAINT bump up numpy and scipy versions for travis latest
       new  9f25765   Merge pull request #6494 from ogrisel/fix-travis-conda-numpy
       new  07cede7   Multitarget regression meta estimator
       new  328ebfa   ENH RandomizedPCA collapsed into PCA
       new  25c931a   Merge pull request #5299 from giorgiop/pca+randomizedpca
       new  a1ee6eb   Update Python version in README
       new  eb242c2   ENH: FeatureHasher now accepts string values.
       new  d4a8d73   Merge pull request #6521 from giorgiop/patch-2
       new  876f123   Do not ignore files starting with _ in nose
       new  2e7d9ad   FIX: improve docs of randomized lasso
       new  bf81451   Fix consistency in docs and docstring
       new  a754e09   Added ref to Bach and improved docs
       new  f81e5aa   Try to fix link to pdf
       new  3a83071   fix x and y order
       new  77c17bc   Correcting Typo
       new  5e41327   Merge pull request #6530 from Arafatk/master
       new  ce11cbc   added MultiOneVsRest classifier & testing suite
       new  edd0eab   change "lasso_path" to use "return_n_iter" argument
       new  aadaf0f   Merge pull request #6531 from gasagna/patch-1
       new  4eca0c9   updated info for cross_val_score
       new  26c6621   Update copyright year in the doc footer (2015 → 2016)
       new  7345351   Merge pull request #6543 from ohld/update-footer-year
       new  b286b85   Add and print a total_n_estimators information
       new  06bf797   DOC update docstring in LogisticRegression
       new  aa31fa6   Added additional references for ARDRegression
       new  e8a8a67   Two typo on my change, sorry
       new  195a5eb   Merge pull request #6546 from Naereen/patch-1
       new  681fd2f   DOC: Fix order in ARD references
       new  113ee40   Merge pull request #6549 from FlorianWilhelm/ardregression_doc_addition
       new  c2eaf75   Merge pull request #6173 from dsquareindia/featurehasher_fix
       new  b64e992   Merge pull request #6542 from ohld/make_scorer-link
       new  e2e6bde   Merge pull request #6498 from clamus/rand-lasso-fix-6493
       new  9691824   Merge pull request #6466 from lesteve/nose-ignore-files-tweak
       new  7580746   Fix broken link in ABOUT
       new  bd6b313   Merge pull request #6565 from bryandeng/doc-link
       new  549474d   [gardening] Fix NameError ("estimator" not defined). Remove unused var "ind".
       new  e228581   Merge pull request #6566 from practicalswift/fix-nameerror-and-remove-unused-variable-in-multiclass
       new  e9492b7   LabelBinarizer single label case now works for sparse and dense case
       new  528533d   MAINT: Simplify n_features_to_select in RFECV
       new  945cb7e   Merge pull request #6221 from dsquareindia/LabelBinarizer_fix
       new  54af09e   Fixing typos in logistic regression docs
       new  146f461   Merge pull request #6575 from hlin117/logregdocs
       new  07a6433   Fix typo in html target
       new  b3c2219   Add the possibility to add prior to Gaussian Naive Bayes
       new  5a046c7   Update whatsnew
       new  5d92bd5   Merge pull request #6579 from nlathia/issue-6541
       new  65b570b   Update scorer.py
       new  56d625f   Merge pull request #6569 from MechCoder/minor
       new  22d7cd5   Make dump_svmlight_file support sparse y
       new  eed5fc5   Merge pull request #6395 from yenchenlin1994/make-dump_svmlight_file-support-sparse-y
       new  9b7176d   [DOC] Fix broken links
       new  afc058f   Merge pull request #6376 from tracer0tong/issue_6352
       new  42d49c8   Make assign_rows_csr support Cython fused types
       new  e5c366f   Merge PR #6037: copy_X in KernelPCA
       new  d7cf4b0   Merge pull request #6588 from yenchenlin1994/use-fused-types-in-assign_rows_csr
       new  e34bbc1   Use fused type in inplace normalize
       new  08e1db4   Test normalize function in data.py
       new  ea9896e   Merge pull request #6539 from yenchenlin1994/use-fused-type-in-inplace-normalize
       new  d574a03   type checking before comparison, to avoid warning messages If np.array is pssed  as a parameter 'n_values', warning of multiple comparison is printed. This patch avoid this problem.
       new  0fd64c2   DOC: mark weights as optional
       new  3183895   Merge pull request #6608 from tacaswell/doc_knc_docstring
       new  24198ff   Merge pull request #6419 from nelson-liu/fix_linkcheck
       new  b454293   DOC: Correct some errors in chebyshev distance closes #6576
       new  acdaeaf   Merge pull request #6606 from tkamishima/tkamishima_one_hot_encoder_parameter_checking
       new  d2aaf0f   doc: fix spellings of 'license'
       new  29ee54a   Meta estimator for multi output classification
       new  3078d7d   Merge pull request #6127 from hugobowne/MultiOneVsRestClassifier
       new  6e43d6e   What's new?
       new  33396b1   Merge pull request #6611 from maniteja123/chebyshev_doc
       new  6ce198c   [MRG] #6581 n_samples of utils.resample can be more when replace is True
       new  45ad9f4   [MRG] fix documentation typo
       new  c20c76e   FIX use the model_selection
       new  6225512   [MRG] Misprint in documentaion fixed
       new  3971c00   #6609: related_projects.rst, one commit - @rvraghav93, @jnothman
       new  689a91e   [MRG+3] Add weighting option to cohen_kappa_score
       new  589fcad   [DOC] fix reference to examples
       new  7efeb5a   ENH added elkan k_means algorithm
       new  273e91b   change default order in check_array and update variable names in _k_means_elkan.pyx
       new  f460330   docstring elaboration and typo in fixes.py
       new  d7c4409   doc typo
       new  0a094b9   Fix grammar bug in docstring of SelectFromModel
       new  56ce722   Merge pull request #6632 from mszep/patch-1
       new  2efa626   [MRG] Fix strict tests
       new  e012b7b   FIX assert_array_almost_equal in test_image
       new  7f57a65   Merge pull request #6636 from ogrisel/fix-strict-tests-2
       new  dfa90ec   add booking.com testimonial
       new  7e8f8b7   Update plots_digits_classification.py
       new  c3b212a   NMF Bug fix when init='custom'
       new  bc8d088   Small fix on ensemble.rst
       new  ef7d046   Merge pull request #6638 from braunmagrin/patch-1
       new  1c044b3   fixed doc formatting
       new  039aff7   debian/rules: skip test_dump tests on arm64 + robustify use of xargs to not fail if no matches
       new  a71d31c   Attribute explained_variance ratio is now available for LDA with the SVD solver. See issue #5216. However, this attribute is corrected from the previous commit which, might not return the expected values.
       new  355ab0d   Corrected the dimension of the array explained_variance_ratio_ for LDA, solver svd.
       new  9bf24fd   Cleaned the calculation of explained_variance_ratio_ in LinearDiscriminantAnalysis, SVD solver.
       new  5d33974   Regression test for explained_variance_ratio_, LDA, SVD solver.
       new  d019a53   Added 4 spaces to conform to PEP8.
       new  7531b30   Modified the length of the slice: now uses a variable, not a hardcoded number.
       new  0476784   Modified the What's new file to mention the bug fix about the wrong results returned with discriminant_analysis.LinearDiscriminantAnalysis's attribute explained_variance_ratio.
       new  74c29e0   misc
       new  262e9e4   fix test warning
       new  45ff64b   type echecking before comparison, to avoid warning of multiple comparison (#6615)
       new  1d487fb   [MRG+1] issue #6532 Add `inverse_transform` function (#6570)
       new  e6bb6b8   Removed redundant Classification metrics entry: metrics.brier_score_loss.
       new  b2002dc   Merge pull request #6653 from davidwarshaw/doc-remove-redundant-brier_score_loss
       new  c9b082a   [DOC] Moves comma to make sentence more clear
       new  14175c7   Merge pull request #6658 from ClimbsRocks/patch-5
       new  18396be   ENH: Add indicator features to imputer output
       new  d837998   Merge pull request #6607 from maniteja123/imputer_indicator
       new  f254ce2   ENH move the raise statement alone inside a gil block (#6644)
       new  566a0e7   Make cython compiler directives global
       new  b094a59   Merge pull request #6664 from yenchenlin1994/make-cython-compiler-directives-global
       new  8b6010d   [MRG] Fixed issue #6635: incorrect docstring for Birch
       new  28758cc   [MRG+2] Use fused types in sparse mean variance functions (#6593)
       new  c59900c   add badge for doi into readme (#6672)
       new  2c0cbdc   DOC: refactor issue and PR template (#6470)
       new  1fa52e4   Disable checks in _centers_sparse
       new  30fcef4   Fixed minor typo in multitasklasso docs
       new  5876f88   Merge pull request #6682 from hlin117/multilasso
       new  bd99858   [MRG +1] DOC - add example to cross_val_predict and cross_val_score (#6446)
       new  a0b8c52   DOC Make the PR/Issue headers smaller (#6685)
       new  0625e03   Integration of the new GaussianMixture class.
       new  6fb6c63   Modification of the ignore_warning function and _IgnoreWarning class.
       new  6bc346b   Merge pull request #6666 from tguillemot/GSoC-GMM-new
       new  0f65de8   typing mistake
       new  97ad719   STYLE minor flake8 fixes in nmf.py
       new  c0cb2bf   [DOC] clarify distance mode in graph creation functions.
       new  475a6b7   Revert definition of check_array
       new  ae51a55   revert doc string of check_array
       new  251fc91   small grammar fix
       new  88e7e77   make centers C contig
       new  97b90fb   Merge pull request #6698 from ixjlyons/fix-neighbor-docs
       new  e6b0e21   Update decomposition.rst
       new  bf1b753   STYLE use one-liner
       new  981df96   Merge remote-tracking branch 'scikit-learn/master'
       new  c5fd5cb   DOC: fixed typos and one style issue in plot examples
       new  bb102d8   Merge pull request #6707 from saiwing-yeung/my-fixes
       new  1328c01   Merge pull request #6705 from mathcass/decomposition_typo
       new  427179b   [DOC] update and rename readme to readme.md (#6669)
       new  78a6748   Correct the deprecation of the random_integers numpy function. (#6712)
       new  6d7dfdb   Merge pull request #5414 from vighneshbirodkar/el
       new  e733b28   model_seclection._validation.cross_val_predict_apply_func
       new  d9da20d   Merge pull request #6671 from zivori/model_seclection._validation.cross_val_predict_apply_func
       new  37880af   COSMIT better test name
       new  1a5be6d   Fix IsotonicRegression __setstate__ (#6729)
       new  80230d6   DOC add reference to multinomial logistic example
       new  93ffce3   [DOC] Added return values to _intercept_dot docstring in linear.logistic
       new  97e4232   [MRG+1] Bug fix for OneVsOneClassifier - converted input to numpy array (#6626)
       new  3ab37f7   Add TPOT to related projects (#6732)
       new  8182f4b   Fixing default order docs of sklearn.utils.check_array
       new  7d2d927   importing VotingClassifier in the example
       new  22cf46e   [MRG+1] Reenable MKL for python 3.5 in Travis (#6661)
       new  c5c2b37   Update OAS estimator description (#6758)
       new  e8d331f   DOC: Improve docs for average precision score
       new  7c0ebbc   FIX does not draw the ellipse! #6640 (#6757)
       new  65c4df1   Update plot_roc_crossval.py (#6765)
       new  294ee9d   Merge pull request #6677 from MechCoder/remove_checks_sparse
       new  ec8299a   DOC: fix typo in check_arrays
       new  4d6f977   Merge pull request #6770 from nelson-liu/fix_check_array_typo
       new  97e7245   Merge pull request #6760 from maniteja123/average_precision
       new  6a82a2e   Merge pull request #6755 from hnykda/hnykda-fix-votingclassifier-example
       new  a4d070e   [MRG] Disabled randomized_svd warning when number of iterations is not specified #6746 (#6749)
       new  7ae7ea5   pep8
       new  7ce4f5c   FIX incorrect gradient computation for SquaredEpsilonInsensitive loss
       new  f73e9dd   DOC do not mention private class in changelog
       new  89247fc   FIX SquaredHinge is a classification loss (#6771)
       new  a249821   DOC cross-reference preprocessing estimators and functions in their docstrings
       new  0c879ba   [MRG+1] Fix: Replace pylab with matplotlib.pyplot #6754 (#6762)
       new  2dd5a9a   change last pl to plt
       new  8f199fe   Modification of the GaussianMixture class to use cholesky precision matrices. (#6743)
       new  e743cf7   MAINT: re-organize setup.py & make sure all tests are discoverable by nose
       new  fdcb5b4   MAINT: fix setup for metrics/cluster
       new  1cc4e9a   TST: fix relative import in model_selection tests
       new  304f4b5   TST: make model_selection test an absolute import
       new  a89f754   FIX bug where expected_mutual_information may miscalculate
       new  3d9463d   Merge pull request #6724 from jnothman/emi_bug
       new  417788c   DOC minor whats_new tweaks
       new  c95a0cb   Typo in `plot_cv example`
       new  2f28ebe   Merge pull request #6790 from arokem/patch-1
       new  65e91f7   Make csr_row_norms support fused types
       new  dc42bde   FIX Don't warn when scoring param is passed + PEP8 (#6798)
       new  a769b25   ENH remove excessive verbosity when no parameters set for CV
       new  cb8ce73   make Parallel be one less verbose in LDA (#6772)
       new  82d9b48   TYPO Update decomposition.rst (#6803)
       new  e42a370   Merge pull request #6274 from jakevdp/setup_py_fix
       new  3f40bf4   Replace print statement in fastica() with warning (#6807)
       new  5676ed2   Add a check whether stripping accents is needed (#6805)
       new  2cc6160   plot decision function histogram on bench datasets
       new  3877d32   binary predict in IsolationForest + test added
       new  c20c9b7   cosmit
       new  6bdcba6   update doc + bench
       new  37b1594   add an example + change predict to respect convention
       new  f89fe80   cosmit
       new  a374fa3   fuse anomaly examples
       new  67fa7e0   fuse benchmarks
       new  5fc1364   cosmit
       new  7e8631a   cleanup bench script
       new  f4a513a   cosmit
       new  ce45d36   take into account raghav comments
       new  165392e   misc
       new  ea947b8   better figsize in iforest bench
       new  62f467e   Merge pull request #6610 from jeff1evesque/feature-6609
       new  e17d207   doc: update mailing list info (#6812)
       new  20f89ef   Remove add_row_csr (#6676)
       new  c7d6f9f   Test row_norms for float32 data
       new  1107f22   use class_weight through sample_weight in LogisticRegression with liblinear
       new  af171b8   Merge pull request #6817 from TomDLT/logistic_class_weight
       new  a6a6ff6   Merge pull request #6785 from yenchenlin1994/make-csr_row_norms-support-fused-types
       new  81559ee   Update tree.rst (#6826)
       new  204472a   FIX segmentation fault in libsvm (#6708)
       new  b13acb1   and related minor documentation fixes
       new  df80c4c   Merge pull request #6801 from jnothman/noparams
       new  4854454   [MRG+1] Parameter check in NMF and IsolationForest (#6814)
       new  d161bfa   Add missing double quote (#6831)
       new  1002de2   Fix small typos in docstring and comment (#6840)
       new  ed7af66   Typo documentation fixes in Advanced Installation (#6753)
       new  c7465f2   DOC - improved doc on ridges alpha parameter (#6844)
       new  1f76574   Add description about default score_func
       new  5a333cd   fixed error message for invalid class_weight values
       new  d12bc46   Merge pull request #6850 from kmike/class_weight_error
       new  814223c   [MRG+1] FIX support memmap scalars as CV scores (#6789)
       new  18092fd   fix balance
       new  e669f87   Modification of the GaussianMixture class. (#6824)
       new  0f76f18   fix Indentation and redundant info
       new  6064327   fix Indentation and redundant info
       new  9795916   Doc: Fix typo (#6854)
       new  a5176e5   Merge pull request #6853 from krishnakalyan3/class_weight_balance
       new  95c4172   [MRG+1] Add a set recording pushed elements in DBSCAN (#6799)
       new  70d7fec   FIX workaround bug in numpy 1.9 (#6856)
       new  b280432   DOC insert missing function description
       new  8ab2e3c   DOC fix broken image path
       new  a2e0bd2   PEP8
       new  c625256   Fixed pep8 errors
       new  e047551   pep8
       new  7c886e4   fix pep8
       new  d9b8ece   Lasso docs fixed (#6880)
       new  f9b1365   TST ensure tests are importable
       new  234dc6d   Change GMM -> GaussianMixture. (#6886)
       new  069bfba   link to gpr guide (#6885)
       new  9d444c9   add doctest extension (#6774)
       new  a91e044   DOC add betaworks testimonial (#6825)
       new  19364bf   Update ISSUE_TEMPLATE.md
       new  1837224   Merge pull request #6613 from nelson-liu/correct_license_spelling
       new  afd5d18   [MRG] ENH Restructure grid_scores_ into a dict of 1D arrays
       new  c6a457f   Add Brier_score_loss to model validation doc
       new  7fea3ef   CI Install psutil (#6897)
       new  1f86b1d   [MRG] ENH Add Calinsky-Harabaz and Fowkes-Mallows clustering metrics (#6823)
       new  2242b1b   Make KernelCenterer a _pairwise operation
       new  3b4ae28   Fix CircleCI failure (#6903)
       new  cc9dbac   Adding test for PR #6900
       new  13f68c9   Simplifying imports and test
       new  f91b4e5   updating changelog links on homepage (#6901)
       new  1bec8e3   first commit
       new  488f7d6   changed binary average back to macro
       new  7d6c629   changed binomialNB to multinomialNB
       new  6301b1f   emphasis on "higher return values are better..." (#6909)
       new  725ff4a   fix typo in comment of hierarchical clustering (#6912)
       new  6bea11b   [MRG] Allows KMeans/MiniBatchKMeans to use float32 internally by using cython fused types (#6846)
       new  ccefc2e   Fix sklearn.base.clone for all scipy.sparse formats (#6910)
       new  1870d6d   DOC If git is not installed, need to catch OSError
       new  10a6c33   DOC add what's new for clone fix
       new  f485fce   fix a typo in ridge.py (#6917)
       new  055bc4c   pep8
       new  94faf0e   Merge pull request #6908 from HashCode55/master
       new  d41b706   Merge pull request #6900 from fishcorn/fishcorn-patch-1
       new  1dbc069   TST: Speed up: cv=2
       new  5820246   DOC Minor grammar fix (#6925)
       new  4a2bc34   DOC Fix text data tutorial (#6923)
       new  459e8bf   COSMIT fix typo in author name
       new  3cc7fea   [MRG+1] Added support for sample_weight in linearSVR, including tests and documentation. Fixes #6862 (#6907)
       new  91746b3   Fix minor typo in grid_search.py
       new  b4ca1fb   Merge pull request #6935 from proinsias/proinsias-gs-typo
       new  9cef05f   documentation: fixing typo (#6938)
       new  1220ca3   fix pipeline docs typos (#6940)
       new  aec7f25   Doc: spectral.py: Add default value to gamma (#6941)
       new  e2a2b4d   Fix typos (#6942)
       new  0e59633   DOC Add missing parameter defaults for some ensembles
       new  5ea317b   Merge pull request #6944 from proinsias/proinsias-rfc-doc-defaults
       new  fe03879   Fix doc refs to StratifiedKFold and KFold (#6936)
       new  a0d6d53   DOC Add missing parameter defaults for Naive Bayes classes
       new  3f37cb9   [MRG+2] convert to boolean arrays for boolean distances (#6932)
       new  a546e1a   Properly handle bounds in isotonic regression with `np.clip` (#6949)
       new  7889054   Merge pull request #6946 from proinsias/proinsias-doc-nb
       new  4d33cd6   rename function to func (#6700)
       new  a0ccea9   TST avoid imports from deprecated module
       new  03ec6c9   [MRG] Added sample weight parameter to LinearSVC.fit (#6939)
       new  69e1df8   DOC what's new entries for LinearSVR/C weight support
       new  b661d1e   Minor grammar fix
       new  d6c479f   Merge pull request #6979 from ClimbsRocks/patch-8
       new  b242539   Fix minor typo
       new  2ea5666   Merge pull request #6982 from fisache/master
       new  4c32aa2   update for issue #6523 (add logistic_regression_path to classes.rst (#6970)
       new  71c6872   avoid creating a diagonal matrix (#6989)
       new  7b7093c   DOC make confusion_matrix_plot.py exmaple more copy-pastable
       new  97c47d9   Merge pull request #6966 from doshyt/doc-update-confusion
       new  a2dac46   Fix minor typo in doc (#7001)
       new  fc229a1   [MRG + 1] docs: Change example of KDTree to prevent DeprecationWarning (#6962)
       new  f6bfe6a   [MRG+1] test against single-class y_true values for #6980 (#6987)
       new  dfb9217   removing confusing comment (#7007)
       new  2857acd   Updated README, removed git clone with write reference (#7009)
       new  c67f091   Added deprecation cycle and modified examples (#7013)
       new  46bd3e3   Edit CONTRIBUTING doc (#7017)
       new  b0455b1   DOC Fixed check_estimator example (#7025)
       new  f9d26fe   adding user guide link  to LinearDiscriminantAnalysis issue #6829 (#7031)
       new  e0d501f   Joblib 0.10.0 (#6985)
       new  89e0735   [MRG+1] Deprecated GMM functions (#7012)
       new  536b26a   [MRG + 1] update test_common.py deprecation warning (#7024)
       new  1a8ddbe   [MRG +1] adds DeprecationWarning to DPGMM and VBGMM (#7020)
       new  7449dc2   DOC: fix spelling of Schütze in tfidf citations
       new  1fbd0b8   DOC fix misconception in DBSCAN documentation
       new  48d27c5   DOC: fix typo (#7044)
       new  6b863e5   DOC: fix typo in sprint title in about (#7043)
       new  f95069d   DOC update refs for MiniBatchKmeans
       new  d92413b   typofix in modules/grid_search.rst (#7045)
       new  d6a1230   DOC fix MiniBatchKmeans See Also and colon
       new  faa6765   DOC harmonize KMeans docstring with doc standards
       new  aa4f756   remove redundant NotFittedError import from voting classifier (#7060)
       new  f40b36d   Typo
       new  dd0681f   Merge pull request #7062 from neo01124/master
       new  d9c9c01   DOC: fix typo
       new  6b84597   Merge pull request #7063 from samuelstjean/patch-1
       new  cbd3bdb   DOC fix typo in KMeans docstring
       new  8bc8aed   Merge pull request #7057 from doshyt/doc-update-confusion
       new  94246be   [MRG+1] RadiusNeighborsClassifier edit (#6918)
       new  7a7e809   pep8
       new  c84ff5e   [MRG+3] Add mean absolute error splitting criterion to DecisionTreeRegressor (#6667)
       new  70bbb82   DummyClassifier and DummyRegressor should raise NotFittedError (#7069)
       new  1cf192b   fixes #6829 - user guide links to DiscriminantAnalysis (#6838)
       new  1f1f7da   Merge pull request #7041 from jnothman/dbscan_doc
       new  92853f9   DOC Reword description to remove ambiguity (#7066)
       new  0316279   DOC joblib default start method is not forkserver (#7077)
       new  4f4a580   Merge pull request #7033 from nelson-liu/fix_tfidf_intro_to_ir_citation
       new  2ace75b   COSMIT Reduce duplicated code (#7053)
       new  ab6dd89   MNT use model_selection (#7082)
       new  8972d82   DOCS / COSMIT duplicate word typos
       new  ac8863e   Merge pull request #7083 from amueller/typos
       new  1ad6fe3   DOC revert removal of duplicate word in doctest
       new  e2bc2a7   add __dir__ to bunch for better autocomplete (#7090)
       new  d829091   pep8
       new  376aa50   [MRG+1] feature: add beta-threshold early stopping for decision tree growth (#6954)
       new  9b25f5d   Fix docs links (#7005)
       new  a6b5ff2   add missing verb in whatsnew
       new  46f81f0   Updated docstring for bootstrap argument of IsolationForest (#7095)
       new  dcca31c   Modify v0.18 DeprecationWarning messages (#7067)
       new  5085a7f   add versionadded for MAE (#7105)
       new  cb4f4e1   DOC: fix spacing in forest mae versionadded (#7106)
       new  7de6e70   [MRG + 1] add "versionadded" tags to min_impurity_split docstrings (#7093)
       new  ce61232   [MRG+1] Check for bad labels in confusion_matrix. (#6958)
       new  7bde7e7   doc: fix early stopping indentation in what's new (#7108)
       new  7161755   [MRG+1] Added text and logo for ottogroup testimonial (#6802)
       new  331964e   [MRG+1] DOC Modified the navbar menu to be responsive with mobile. (#7098)
       new  b8be019   [MRG + 1] load_iris dataset: added return_X_y option (#7049)
       new  e4b837c   Removed unused imports
       new  9539c0c   [MRG + 1] ENH show no warning with chi2 of empty feature (#6978)
       new  d8db108   Fixed libsvm call with unicode kernel. Added corresponding test.
       new  9e913c0   Merge pull request #7064 from Erotemic/quickfix_py2_libsvm_kernel_unicode
       new  91129f3   Merge pull request #7119 from jrfiedler/remove-unused-imports
       new  d2dca4e   COSMIT fix python2 -> python3 in comment
       new  f2df3db   ENH TfidfTransformer.transform() performance bug fix (#7125)
       new  928f724   [MRG] fix #5269:  Overflow error with sklearn.datasets.load_svmlight (#7101)
       new  1e2123d   DOC rm obsolete DATASET_PROPOSAL.txt (#7133)
       new  2a9492b   [MRG] MAINT Removing more deprecated stuff for v0.18 (#5528)
       new  ca4e802   [MAINT] Install sphinx-gallery v0.1.0
       new  17885af   conf.py integrate sphinx-gallery
       new  04116ce   update image sources namespace
       new  bd5cecc   update references link namespace
       new  3d79f03   update reference namespace in gallery scripts
       new  7ec5447   Add carousel
       new  d6be865   References to examples
       new  8034b29   more examples references
       new  96f6162   [MAINT] Update to sphinx-gallery v0.1.1
       new  e2e17ed   [MAINT] update to sphinx-gallery v0.1.2
       new  4773dbb   Remove custom touch to backreferences files
       new  83b21a4   Sphinx-Gallery has to be loaded after autodoc
       new  8b46c61   Python 2 backwards compatibility in example
       new  e10a272   Fix example to use rst syntax in notebook comments
       new  90fe87f   Extra missing links complained by sphinx
       new  acef8cd   Sphinx-doc absolute path for auto_example images
       new  40169b9   Merge pull request #4658 from Titan-C/sphx-glr
       new  c6f6712   Contributing statement (#7097)
       new  0493d71   fix of typo (#7137)
       new  1b6c84d   fix indent in whats_new and fix indent in datasets/base.py (#7138)
       new  ba2bcde   Fix typo in example feature_stacker.py
       new  913967b   Fix typo (#7145)
       new  5bddfdb   add load_breast_cancer to API reference (#7149)
       new  05dfe6c   Correct Deprecation of DPGMM and VBGMM. (#7124)
       new  1bcd6c3   fix MultiOutputEstimator fit return docstring (#7150)
       new  89ff4cf   fix some sphinx warnings in whats_new.rst (#7151)
       new  c98adf7   [MRG] load_breast_cancer dataset: added return_X_y option (#7152)
       new  12d5f07   [MR] Speed issues in sample_without_replacement - bugfix (#7104)
       new  8994d0e   [MRG+1] added return_X_y option to toy datasets in sklearn.datasets (#7154)
       new  4f71f3b   [MRG+1] Dpgmm alpha (#7028)
       new  fc66544   cosmit
       new  4c7753c   fixed typo
       new  a453cd0   Merge pull request #7165 from tKammy/fixed_typo
       new  f50790c   DOC Add LabelKFold in _BaseKFold's docstring (#6344)
       new  8ad12ed   Use pgtol parameter in fmin_l_bfgs call (#7167)
       new  65b7d7a   [MRG+2] Modification of GaussianMixture class. (#7123)
       new  1f8eaac   TST make test_forest robust to platform variability
       new  c99aada   DOC Typo: consequective -> consecutive (#7171)
       new  8f8c3d2   [MRG] Check for NaN or Inf in metric input (#6976)
       new  c6136af   Add spherecluster related package
       new  916b31a   Address reviewer comments
       new  37dbb1c   [MRG] Make Silhouette_score support sparse X (#7170)
       new  252b356   FIX broken NMF test with numpy 1.7.1
       new  4327acb   STYLE trailing whitespace
       new  2714b04   FIX broken iforest test under numpy 1.7.1
       new  d96c05a   Merge pull request #7172 from jasonlaska/patch-1
       new  96f6ec0   [MRG] extending tfidf documentation (#7015)
       new  0cffdfa   BUG fixed a typo in ``sklearn.gmm`` (#7182)
       new  f893565   [MRG+2] ENH: (minor) add quantile_low/high parameters to robust scaler (#5929)
       new  1f182cc   Fix -np.infty in GaussianMixture. (#7180)
       new  02806e9   cosmit
       new  d2c04c8   Doc minor fix ensemble.rst (#7186)
       new  7b3c467   [MRG] Add whats_new about fixing example (#7184)
       new  72fd2f5   rephrasing
       new  617f0ca   DOC fix Olivier's misspelt name
       new  dc6706e   [MRG] edit citation indentation in about.rst and add gsoc info for '15 '16 (#7191)
       new  6683c61   Merge pull request #7143 from iver56/patch-1
       new  7e55d99   FIX broken test when run with arch -i386 on OSX
       new  42120e5   [MRG+1] Rename CV params n_{folds,iter} to n_splits (#7187)
       new  0e1fd12   DOC Correction: max_depth is effective when max_leaf_nodes is not None (#7183)
       new  55de9bb   DOC improve formatting for learning_rate docstrings
       new  c31ad7a   [MRG + 1] [MAINT] Remove non-trivial logic during kernel initialization (#6963)
       new  668b329   FIX make HuberRegressor tests stable under 32 bit Linux (#7196)
       new  974639f   Explicit warnings in model persistence documentation (#7157)
       new  9d0c2dd   [MRG] DOC updated normalize param docstring (#7161)
       new  5856a2c   clarify splitting sparse data always outputs csr matrix
       new  24af423   edit docstring for arrays argument as well
       new  8b6753d   remove accidentally added whitespace
       new  07e0dd3   removed versionadded for arrays param and fixed spacing
       new  19d6d92   [MRG+1] Fix for "_transform_selected" not copying when "selected='all'" (#7201)
       new  5f2297e   feature: add tests for sparse matrices
       new  da6c5c2   try to fix travis error
       new  6766aea   be more explicit about purpose of test
       new  a55fd92   Merge pull request #7199 from nelson-liu/edit_cv_sparse_docs
       new  70642c0   GaussianNB now support sample weighting
       new  a18344e   Merge pull request #7213 from zhengruifeng/doc_fix
       new  6e1418e   ValueError that is thrown by _check_param_grid method now includes parameters name when values of a parameter have incorrect type.
       new  454b0f2   minor wording fix
       new  a212b1a   Merge pull request #7217 from ClimbsRocks/patch-9
       new  444cf51   fixing ElasticNet sparse decision function to match output with dense output
       new  076d955   test case for ElasticNet multiple output matching with dense
       new  540e2a0   pushing Transpose into dot product
       new  02cd2fb   removing make_sparse_data call
       new  fe20981   removing extra fit call and cleanup code
       new  2d39122   droping extra y
       new  388df6b   fixing alignment
       new  fa8fe85   update what's new + simplify test
       new  5c6409c   fix issues with manifold tests throwing pep8 errors (#7209)
       new  214a482   ValueError that is thrown by _check_param_grid method now includes parameters name when values of a parameter have incorrect type.
       new  fa6b88c   Merge branch 'master' of https://github.com/b0noI/scikit-learn
       new  39a04f1   update setup.py PyPI trove classifier section (#7197)
       new  f7f1b96   "values" has been changed ot the "value" in the ValueError message.
       new  277b058   [MRG] Remove heavy memory footprint in BaseBagging due to OOB scoring (#7118)
       new  59bf211   Minor doc fixes in pairwise metrics (#7227)
       new  5e9e941   Text of the ValueError in _check_param_grid has been changed.
       new  abee34e   phrasing
       new  040a766   [MRG] Fix TSNE init as NumPy array (#7208)
       new  234d256   [MRG] Add homogeneous time series cross validation (#6586)
       new  114d8b1   ENH Better error message when refit=True. (#7234)
       new  084ef97   [MRG+2] ElasticNet and Lasso now support float32 dtype input (#6913)
       new  6028e42   fixed log_loss bug
       new  79406f4   fixed error message when y_pred and y_test labels don't match
       new  d97a25f   Wrap up changes to fix log_loss bug and clean up log_loss
       new  104e09a   Add labels argument to log_loss to provide labels explicitly when number of classes in y_true and y_pred differ
       new  d6299c0   DOC: minor docstring format fixed (#7249)
       new  a7d748b   Rename `TimeSeriesCV` to `TimeSeriesSplit` (#7245)
       new  b3e122a   [MRG+2] parallelized VotingClassifier and sample_weight support (#5805)
       new  50ab94d   MockDataFrame.__array__ should take an optional dtype
       new  45e9c2f   NotAnArray.__array__ should take an optional dtype
       new  68b4c68   Merge pull request #7252 from lesteve/fix-mock-dataframe-check-array
       new  52bfd25   fix error in boolean mask usage in GaussianNB. (#7258)
       new  acf8368   [MRG + 1] doc: specifically address reinforcement learning in faq (#6479)
       new  9b2e53e   Rename smaller-is-better scoring metrics
       new  7e079c0   Update tests to new scorer names
       new  3b95d5f   Fix deprecation warning for 1d matrix input (#7263)
       new  a9bf914   Fix a typing issue in expected mutual information (#7240)
       new  ef64969   Introduce deprecation warning and fix tests
       new  e8f3044   Add class example for KMeans
       new  dcb3165   TST remove print statement in gmm test, remove verbosity in mlp test.
       new  002ff12   fix some warnings in test outputs.
       new  9d695d7   Keep old metric names for deprecation period
       new  d464972   Keep old metric names in tests
       new  7f4d279   Minor docstring clarification (#6926)
       new  58b35d8   DOC Updates sklearn naming convention for consistency (#7268)
       new  b8b31dd   [DOC] Capitalizes "C/C++" from "c/c++" (#7273)
       new  0585462   Tests that are checking that _grid_search method throws correct ValueError have been added.
       new  401d2a1   Test method test_grid_search_incorrect_param_grid has been renamed to test_grid_search_param_grid_includes_sequence_of_a_zero_length.
       new  aaf203d   Merge pull request #7264 from MridulS/kmeansdoc
       new  42fe520   Do not show deprecated scorers in list of valid names
       new  3250f15   [MRG + 1] Rebased navbar fix (#7202)
       new  1e2b443   MAINT/DOC: pngmath deprecated for sphinx >= 1.4 (#7279)
       new  ebb0645   DOC reference CalibratedClassifierCV in SGDClassifier.predict_proba (#7281)
       new  5b20d48   [MRG] ENH enable setting pipeline components as parameters (#1769)
       new  845e702   [MRG+2] Make `enet_coordinate_descent_gram` support fused types (#7218)
       new  4566251   [MRG+3] Add flake8 linting of the diff in Travis (#7127)
       new  47eae84   MockClassifier has been added to the _grid_search tests.
       new  3d16c47   Do not remove log_loss scorer during deprecation period
       new  09cd4c7   Distinguish instances from functions in deprecated
       new  be94946   Update joblib to 0.10.2
       new  03e3012   Exclude sklearn/externals from diff for flake8
       new  0a0bf24   Revert "ENH: Add indicator features to imputer output (#6607)" (#7292)
       new  6a2b4f7   [MRG] identity activation function for MLPs (#7267)
       new  e2648b1   classes parameter renamed to labels in hamming_loss()
       new  c5f19a4   Fix ConvergenceWarning's in tests
       new  f0862f7   [MRG+1] Bayesian Gaussian Mixture (Integration of GSoC2015 -- second step) (#6651)
       new  82f913d   rename unveil_tree_structure example to enable CI on it
       new  f906b9f   Lines over 81 chracter length got reformatted.
       new  a4f632f   [MRG] Update joblib to 0.10.2 (#7290)
       new  e94c35d   ENH make validation message easier to understand (#7294)
       new  f7bdba2   Merge pull request #7300 from nelson-liu/rename_unveil_tree_structure
       new  bfbc0bc   Merge pull request #7282 from asnt/fix-convergence-warnings
       new  c931bf0   FIX #7293: StandardScaler.with_mean in preprocessing doc
       new  18055e8   ENH/MAINT/TST Complete deprecation of default average in P/R/F metrics
       new  367d73f   PEP8
       new  011472a   Go back to warning all callers
       new  db56cf4   Merge pull request #7216 from b0noI/master
       new  d9428de   Remove unused import
       new  f07dbd4   add directions for contact to AUTHORS.rst
       new  3ea1c63   gitter is not for issues
       new  9f65998   fix regressor var name
       new  61bfe8a   make more explicit which checks are run
       new  2ab7e34   clf = Perceptron
       new  6787da4   make it clear that linear_model.Perceptron is a classifier
       new  b4661b1   colon
       new  6297815   BLD: Add 'alldeps' extra for installing with pip (#6990)
       new  7fc4176   pep8
       new  88a395a   Link to mean_squared_error function
       new  ccad485   Merge pull request #7316 from totalgood/doc-clf-reg-patch
       new  0f2a00f   Remove dtype conversion warning (#7265)
       new  2b11962   Move comment on relationship between labels and pos_label to docstring
       new  49fb295   [MRG+2] Fix repr on isotropic kernels when a 1-D length scale is given (#7259)
       new  f4419a9   Small documentation fix
       new  fed401c   change decision tree example to use pydotplus for compatibility with Python 3
       new  f916449   Merge pull request #7317 from amueller/common_test_names
       new  ce3c84f   DOC fix code formatting
       new  a627a07   [MRG+1] Fix for test_neighbors_metrics() check kd-tree algorithm when possible (#7339)
       new  3dcfb56   [DOC] add release dates for each version on whats_new page (#7225)
       new  d201708   DOC move changelog heading to correct place
       new  287a329   [MRG] Add link to AUTHORS.rst in readme (#7320)
       new  2ba3478   Keep abstractmethod decorator
       new  b9d16b2   Merge pull request #7342 from themrmax/pydotplus-decision-tree-plotting
       new  927f00c   fix warning and behavior in randomized_svd wrt power iterations
       new  4a72792   change default iterated_power to auto.
       new  5daee96   DOC fix iterated_power='auto' in doctests
       new  6673f75   Update examples to new scorer names
       new  5305861   Fix flake8_diff.sh when Travis runs in a fork
       new  6c12514   [MRG] DOC removed redundant word from KNN docs (#7345)
       new  4dbab26   results_ --> cv_results_; "test_mean_score" --> "mean_test_score" et al.
       new  0f11b0c   @ogrisel's suggestions
       new  35bb1c6   Merge pull request #7324 from raghavrv/rename_results_
       new  a03db89   [MRG+1] Support of the collections.Sequence type has been added to the _check_param_grid method from model_selection. (#7323)
       new  680ab51   [MRG + 2] remove complicated equality checks in clone as __init__ shouldn't touch anything. (#5540)
       new  6972d6c   DOC minor doc fixes for sphinx. (#7357)
       new  22e4305   TST fix test to match altered deprecation msg
       new  2c4b43c   change installation description to not discourage people from installing scipy or numpy with pip. (#7313)
       new  62ad5fe   FIX Revert unintentionally committed code
       new  3935404   Test that warning is raised when using deprecated scorer
       new  4b2304f   Merge pull request #7261 from betatim/negative-scorers
       new  2bf96df   Merge pull request #7304 from jnothman/binary_prf_mine
       new  49d126f   [MRG + 1] FIX use high precision cumsum and check it is stable enough (#7331)
       new  31a4691   [MRG+1] fix sampling in stratified shuffle split (#6472)
       new  b4872fe   [MRG] add warning when importing old or new pickle. (#7248)
       new  105d033   [MRG+1] Add hyperlink and example for kappa (#7214)
       new  a86fd7f   Add Dirichlet process prior to BayesianGaussianMixture (#7386)
       new  ba7cf38   Removed isclose from _binary_clf_curve. (#7353)
       new  12a0125   DOC fix for svd_solver in PCA docstring
       new  8da5092   [MRG] Fail imputer early when number of features are not the same in fit and transform (#7374)
       new  9a12555   [MRG+1] ENH/MNT Rename labels --> groups in CV tools (#6660)
       new  802814e   make export_graphviz return string if out_file is None
       new  51679f9   remove unused StringIO.get_value
       new  95655ac   fix lint issues
       new  b966b44   DOC fix class.rst for Label / Group CV tools
       new  e1b5d60   MLP bug fixes with random state and loss, removed decision_function, and added test
       new  1e8ca7d   ENH more informative validation error message
       new  5a018a3   Merge pull request #7393 from ogrisel/mlp-multilabel
       new  3388a05   ENH raise error when scoring=sklearn.metrics.{classification,ranking,...}.*
       new  f02fce8   ENH rename l-bfgs to lbfgs in MLP for the sake of consistency
       new  c3b2115   MNT Remove unused constants from public API(?) (#6517)
       new  34c8227   link to scikit-learn tag on stackoverflow, minor formatting.
       new  b18b161   [MRG+1] TST Stronger test for _check_is_permutation (#7395)
       new  c726085   DOC mention true positives etc in confusion matrix docs (#7392)
       new  0b818d3   [MRG+1] edited the FAQ to fix a typo and include info regarding global random states (#5927)
       new  3dffa08   FIX 7155: GridSearchCV predict_proba delegation to SDGClassifier
       new  3e12412   ENH rename algorithm to solver in MLP for consistency
       new  069c892   Merge pull request #7309 from amueller/how_to_ask_questions
       new  c0a04a3   Fixing typo (#7402)
       new  f241f5f   Fixed issue #5830, capitalization in advanced installation documenation
       new  474483e   Added user guide link in function_tranformer docstring (#5785)
       new  b4b23a3   only return string if out_file is None
       new  284ae9c   fix indentation
       new  a31cc79   Merge pull request #6400 from scikit-learn/radneigh-njobs
       new  2a0bd7b   fix lint error
       new  940f194   Merge pull request #7396 from ogrisel/mlp-optimizer
       new  44f1ad4   Merge pull request #7401 from ogrisel/sgdclf-predict_proba-delegation
       new  3d8d0de   DOC add what's new entry for #7159
       new  8dba703   Merge pull request #7404 from amueller/capitalization_rebase
       new  1f381ae   Fixed doc for issue #7409 (#7410)
       new  3d1a2fd   COSMIT don't use docstrings in tests (gp kernel tests) (#7399)
       new  873f3eb   Fix #6994: MLP partial_fit with unseen classes (#7358)
       new  0d0b484   TST fix generated test names in case where test fails
       new  e5fede5   MAINT master is now 0.19.dev0
       new  4d9fab5   Release 0.18rc
       new  2e5c102   make pickle version test more stable (#7415)
       new  f75ecd7   make pickle version test more stable (#7415)
       new  b687ab3   Release 0.18rc1
       new  4d9a12d   [MRG] add whatsnew for #7248 (unpickle version warning) (#7400)
       new  1578edd   TST move test naming and use inspect
       new  ba5563f   retain default behaviour of out_file=tree.dot
       new  d9bd30e   FIX use == to avoid relying on string interning
       new  11d34ed   FIX use == to avoid relying on string interning
       new  ee0f9b1   DOC fix intercept_ and coef_ docs for LogisticRegression(CV) (#7420)
       new  c4552cf   DOC fix various syntax in whats_new (#7421)
       new  db4e244   ENH: display the original exception (#5906)
       new  268b6ea   use sparse contingency matrix for supervised cluster metrics Remove max_n_classes option
       new  2432451   DOC fix various syntax in whats_new (#7421)
       new  afddc75   ENH: display the original exception (#5906)
       new  d3fafa1   use sparse contingency matrix for supervised cluster metrics Remove max_n_classes option
       new  b1bd976   FIX missing import in setup.py
       new  86ca052   FIX missing import in setup.py
       new  cef2b62   Release 0.18rc2
       new  016f3d5   lda_qda doc: QDA vs Naive Bayes correction (#5997)
       new  4272811   added contributors to whats new for 0.16 and 0.17 (#5840)
       new  500b768   add depreciation warning for out_file default
       new  7dfed85   use double quotes
       new  19aebe2   use == instead of 'is' for string comparison (#7422)
       new  9864cd5   DOC add species distributions dataset to classes.rst (#7430)
       new  205b3a1   edit depreciation warning and add versionadded to docstring
       new  f921231   only show DepreciationWarning if out_file paramater is actually missing
       new  02332f5   use == instead of 'is' for string comparison (#7422)
       new  5e42ecf   DOC add species distributions dataset to classes.rst (#7430)
       new  da43d3e   added contributors to whats new for 0.16 and 0.17 (#5840)
       new  e9e5f18   lda_qda doc: QDA vs Naive Bayes correction (#5997)
       new  396dc39   Change flake8_diff.sh logic for branch in main repo (#7437)
       new  b679dd1   Change flake8_diff.sh logic for branch in main repo (#7437)
       new  941921f   [MRG] Text vectorizers memory usage improvement (#7272)
       new  e222058   Make named_check private
       new  cefd3ca   [MRG + 1] DOC Default parameter info for KernelPCA (#6035)
       new  1054e07   [MAINT] Upgrade to Sphinx-Gallery 0.1.4 (#7354)
       new  f3baca6   [MRG+2] Faster isotonic rebased (#7444)
       new  97c8099   [MRG+2] Faster isotonic rebased (#7444)
       new  c1cbfd1   ENH: display the original exception (#5906)
       new  610879f   fix typo in DBSCAN explanation
       new  cb462b5   remove unnecessary comma
       new  466387e   Merge pull request #7447 from stvn66/stvn66-dbscan-docs-typo
       new  2fce9ae     Provided a fix for issue #7140 - Minor discrepancy in hashing feature   transformation example.
       new  171b31d   DOC fix deprecation warning in faces decomposition example
       new  b69d5fa   remove lambda for flake8 linting
       new  93567dd   [MRG+2] DOC: update persistence documentation with Joblib new behavior (#7434)
       new  77d9c09   fixing failing faces decomposition example
       new  0c2264d   replace RandomizedPCA by new API in remaining documentation
       new  29992d9   add explicit comment in face decomposition example
       new  bd4e00b   [MRG] DOC Example: Fix minor standard output formatting (#7456)
       new  687615d   Merge pull request #7449 from elena-sharova/hashing_feature_transform_example_#7140
       new  28bcb43   [MRG + 1] DOC replace RandomizedPCA with PCA and svd_solver='randomized' in documentation (#7450)
       new  54b0e4b   Add OneVs{One,All}Classifier._pairwise: fix for #7306 (#7350)
       new  46d337b   Merge pull request #7412 from jnothman/fix_test_descriptions
       new  14666a8   add object inheritence to class definition
       new  070093e   [MRG] ENH: Allow specifying arbitrary kernel for semi-supervised learning. (#5762)
       new  3dcc8be   Merge pull request #7389 from jnothman/sniff_bad_scoring2
       new  1ea557e   Merge pull request #7390 from themrmax/export-graphviz-to-string
       new  8a7948a   MAINT remove unneeded import (#7465)
       new  9b56104   [MRG] Add Deprecation Warning to address making "batch" the default learning method in LDA in 0.19 release (#6999)
       new  3e7a7ca   [MRG+1] Fix regression in silhouette_score for clusters of size 1 (#7438)
       new  1dff0ae   clean up deprecation warning stuff in common tests
       new  a1eb062   Reorganize README
       new  1885297   Remove write clone
       new  154635e   Update install
       new  8746c1c   Fix changes re: @amuller
       new  b8641e6   Update mailing list url
       new  c9e227b   Merge pull request #7368 from amueller/deprecation_warnings_tests
       new  9d9aaaa   Add @raghavrv suggested changes
       new  a5feb12   added contributors for 0.18, minor formatting fixes.
       new  6e48ab7   s/Contributors/Code Contributors/g
       new  eda7479   Merge pull request #7018 from willingc/update-readme
       new  f31b946   Addressing issue #7468. (#7472)
       new  b1cc108   [MRG] bump master to 0.19-dev (#7413)
       new  c8096a0   add some more entries to the mailmap
       new  32d1236   [MRG+2] FIX adaboost estimators not randomising correctly (#7411)
       new  62612be   more mailmap and whatsnew additions
       new  bd5e85f   Fix typo in whats_new.rst
       new  98678ba   Avoid creating an ambiguous reference for usptream/master
       new  c9f42b6   Merge pull request #7428 from amueller/contributors018
       new  85f2900   Sync docstring and definition default argument in kneighbors (#7476)
       new  d8a1422   Minor k_means_.py cleanup.
       new  4ed0a6a   Minor test_k_means.py cleanup.
       new  536eb01   Revert PEP8 fixes.
       new  7f076d2   Merge pull request #7480 from omtcyfz/minor-k-means-cleanup
       new  5fcc7e5   DOC: Added Nested Cross Validation Example (#7111)
       new  8a95476   [MRG] Text vectorizers memory usage improvement (#7272)
       new  0a090f8   [MRG + 1] DOC Default parameter info for KernelPCA (#6035)
       new  ffd43ff   fix typo in DBSCAN explanation
       new  8e3ac46   remove unnecessary comma
       new  f873d1c   [MRG+2] DOC: update persistence documentation with Joblib new behavior (#7434)
       new  3521961     Provided a fix for issue #7140 - Minor discrepancy in hashing feature   transformation example.
       new  aaa01bf   DOC fix deprecation warning in faces decomposition example
       new  eb4f0f4   fixing failing faces decomposition example
       new  ae198c1   replace RandomizedPCA by new API in remaining documentation
       new  4357a6e   Add OneVs{One,All}Classifier._pairwise: fix for #7306 (#7350)
       new  2a84390   TST fix generated test names in case where test fails
       new  950cf4d   TST move test naming and use inspect
       new  fd11ca8   Make named_check private
       new  3505b03   ENH raise error when scoring=sklearn.metrics.{classification,ranking,...}.*
       new  6a78ba0   make export_graphviz return string if out_file is None
       new  c43bafa   MAINT remove unneeded import (#7465)
       new  2f43bf2   [MRG] Add Deprecation Warning to address making "batch" the default learning method in LDA in 0.19 release (#6999)
       new  6bc1257   [MRG+1] Fix regression in silhouette_score for clusters of size 1 (#7438)
       new  a27564a   clean up deprecation warning stuff in common tests
       new  eecfdda   Reorganize README
       new  aebf7f6   Addressing issue #7468. (#7472)
       new  d25579a   [MRG+2] FIX adaboost estimators not randomising correctly (#7411)
       new  5ac9202   Fix typo in whats_new.rst
       new  d50390f   added contributors for 0.18, minor formatting fixes.
       new  859776a   Sync docstring and definition default argument in kneighbors (#7476)
       new  e3db699   DOC: Added Nested Cross Validation Example (#7111)
       new  3a106fc   Fix _get_importances. (#7487)
       new  b444cc9   [MRG+2] Timing and training score in GridSearchCV (#7325)
       new  af41bfd   DOC insert missing whats_new heading
       new  01edccb   [MRG+2] Timing and training score in GridSearchCV (#7325)
       new  8e3f888   REL changes for release in 0.18.X branch (#7414)
       new  576eccb   no fixed width within bold
       new  61d2ad5   DOC Attempt to impose some order on What's New 0.18
       new  73ed6e6   DOC more what's new subdivision
       new  ce3a61d   fix rebase messup
       new  2705e1b   fix scoring function table
       new  4ae77df   minor fixes to whatsnew
       new  7901190   fix lbfgs rename (#7503)
       new  644dbcc   fix lbfgs rename (#7503)
       new  191a93d   minor doc fixes
       new  60a1356   [MRG + 1] More versionadded everywhere! (#7403)
       new  38030a0   [MRG + 1] More versionadded everywhere! (#7403)
       new  da118d0   [MRG+3] Fix min_weight_fraction_leaf to work when sample_weights are not provided (#7301)
       new  8207185   Fix typo in PCA docstring
       new  a545864   DOC: trivial rst fix to whats_new.rst (#7511)
       new  442d944   [MRG] DOC Attempt to impose some order on What's New 0.18 (#7496)
       new  c887653   [MRG+1] #7466 add pandas to CI jobs (#7499)
       new  5bb9830   minor doc fixes (#7514)
       new  45e3431   DOC: Tidy up mds module documentation. (#7453)
       new  4773044   DOC release date in what's new (#7518)
       new  3953a9a   add feature_name to diabetes dataset (#4477)
       new  32236ff   [MRG+1] minor link fix for SAG paper. (#7523)
       new  05e702c   MNT Tags built by sublime ctags may be of form (.)tags* (#7494)
       new  2f4b661   [MRG+1] correct condition in decision tree construction (#7441)
       new  69ca580   [MRG] CI build web site from branches (#7498)
       new  c336a43   [MRG+1] micro-optimize HashingVectorizer and FeatureHasher (#7470)
       new  ef8fa61   Merge tag '0.18' into releases
       new  e78c6e6   Merge branch 'releases' into dfsg
       new  e804609   Merge branch 'dfsg' into debian
       new  24dfefb   changelog + dropped changesets
       new  218f90e   refreshed patches
       new  595703c   a little patch for PY3_OR_LATER import
       new  20d0b00   DOC clearer what's new for min weight leaf optimisation
       new  69c8382   [MRG+1] Added 'average' option to passive aggressive classifier/regressor. (#4939)
       new  8442eea   [MRG+1] minor edits to multiclass documentation (#5837)
       new  fd2d1bb   add warning about python 2.6 being no longer supported (#7532)
       new  62c86f0   [MRG + 1] Raised value error in HashingVectorizer and CountVectorizer (#7507)
       new  ec32689   [MRG+1] Fixing error with step parameter #7467 (#7469)
       new  e108068   learning_curve takes error_score parameter and propagates it. (#7397)
       new  3619bd3   BUGFIX _store train_scores only if return_train_score is True (#7535)
       new  dee786a   DOC fix typo in NearestNeighbors docstring (#7545)
       new  b2cb502   debian/patches/deb_use_system_joblib sklearn started to use other submodules of joblib directly, so previous stub for import did not work.  Now update local namespace of the stub entirely with definitions from joblib.__dict__
       new  0100685   debian/rules - allow joblib to parallelize to 2 processes, to overcome recent joblib (0.10.2) bug with 0 value
       new  b86e6f3   extra word in code comment (#7561)
       new  3fbfb1b   fix typo in RandomizedSearchCV docs (#7564)
       new  d356027   DOC return values of sample() function in mixture/base.py (#7571)
       new  2b18a81   debian/rules - skip some minor failing tests on i386 and armel for now
       new  280c70b   add numpy versions for backports / fixes
       new  20912ed   FIX flake8_diff.sh git repo unshallowing (#7575)
       new  1dd5e26   Rangespan testimonial has misformatted link (#7576)
       new  2437535   DOC FIX: update warning about random seed with scipy (#7577)
       new  b32897f   [MRG+2]: Fix Ridge GCV with centered data (#6178)
       new  c4b47ff   DOC add what's new for #6178
       new  d5b66f9   [MRG+1] Docs: refer users to the other encoders to do one hot encoding for labels. (#7315)
       new  fa6fafc   [MRG + 1] fix test_20news_vectorized (#7431)
       new  1fa3bf7   [MRG] DOC Fixed the display of search results (#7517) (#7560)
       new  9551ec8   DOC add spaces before colons in docstrings (#7589)
       new  e17d5c9   [MRG] Fix GaussianProcess batch predict in Py3k  #7329 (#7330)
       new  92f4731   changed address for nose, additionally, should nose2 be considered for testing since the nose project web site suggests new projects should consider using Nose2, or py.test?
       new  3beec40   DOC add link in the contributing doc for rebasing on master (#7541)
       new  f1b57df   got rid of extra line
       new  0a1f6cd   [MRG] FIX Calling fit_transform instead of transform in Pipeline's fit_predict (#7585)
       new  45cb11d   [MRG+2] DOC add info to conventions for multi-label fitting (#7519)
       new  fd40cbc   add a test
       new  1e3da55   Update install instructions for Windows (#7550)
       new  21a3c19   hanged address for nose (#7592)
       new  6c8a1b1   fix error in test code
       new  3d558fa   Merge pull request #7525 from rickiepark/master
       new  9b2aac9   [MRG + 1] [TST] (half-cosmetic) use less nose.tools import to simplify future transition to py.test (#7384)
       new  c948319   BUG: handle integer inputs in LLE (fixes #6033) (#6282)
       new  36e67a9   Fix test suites (#6348)
       new  73bb593   fix an indentation issue in whatsnew (#7605)
       new  cee48cd   removed cv_ prefix (#6252)
       new  689a188   DOC numpydoc requires space between name and column (#7601)
       new  ea752cd   [MRG + 1] Added some documentation for loading external datasets (Issue 3808) (#7516)
       new  5c98812   [MRG + 2] updated iris svm example colormap so the three species can be distinguished (#7611)
       new  5f80525   [MRG] Documentation describes the Need Contributor label (#7599)
       new  baeafbb   DOC pos_label can be string (#7607)
       new  610461c   TST ensure relationship between predict_proba and predict_log_proba (#7612)
       new  33ed90d   [MRG] FIX min_samples_split and min_samples_leaf validation for float vs int (#7604)
       new  44acb33   COSMIT fix typo in docstring
       new  8f938ee   DOC fixed avoidable warning of test-doc (#7524)
       new  b01c8c4   DOC fix link format in reference for BayesianMixtureModel
       new  7955ce0   Add whats new entry for #6282 (#7629)
       new  868a58b   [MRG+1] FIX Make sure GridSearchCV and RandomizedSearchCV are pickle-able (#7594)
       new  5393386   [MRG] FIX regression: Pipeline.fit_predict uses fit_params as in 0.17.x (#7621)
       new  cbd3bca   [MRG + 1] FIX: remove checks from PyFunc distance metric (fixes #6287) (#6288)
       new  dd8a6fa   dev-7360: Explain the meaning of prefix [MRG + 1] and [MRG + 2] in the title of the pull request. (#7636)
       new  0ae762e   DOC paraphrase multiclass description
       new  0b352a7   DOC What's new for #6288
       new  2b0439c   Improve robustness of LFW dataset downloader (#6112)
       new  2221387   FIX n_iter_without_progress and min_grad_norm in TSNE
       new  e4a1d82   DOC what's new for #6497 and 0.18.1 section
       new  d8614a3   DOC better docstring for TruncatedSVD (#7651)
       new  cd0d46c   MAINT make appveyor fail on old builds when PR is update (#6365)
       new  7141dd4   DOC Removing deprecated DPGMM that was also not rendering correctly from classes. (#7606)
       new  ab18f68   DOC Fixed missing "Next" button (#7641)
       new  0e12b2f   DOC: fix the copybutton on the code blocks (#7634)
       new  329b7ee   dev-7369: Make common metric tests look nicer (#7620)
       new  5e4f524   [MRG] edit releases link from sf to pypi (#7637)
       new  53e6381   [MRG+1] Throw an error with explicit message if n_estimators is not an integer. (#7457)
       new  78dbcb2   [MRG] fix #6101 GradientBoosting decision_function for sparse inputs (#6116)
       new  65a4de5   DOC: fix return value doc for mixture.base._e_step (#7682)
       new  fa59873   [MRG+1] FIX unstable cumsum (#7376)
       new  0198e2c   warning for PCA with sparse input (#7649)
       new  0653aac   [MRG+1] CircleCI timeout extended (#7693)
       new  647b7c1   DOC Changed Contributor's Guide to Development Guide #7690 (#7691)
       new  dad35f5   DOC mention JOBLIB_START_METHOD in crash/freeze FAQ
       new  387f25c   DOC use combined unicode chars in author names (#7706)
       new  829efa5   [MRG+1] Learning curve: Add an option to randomly choose indices for different training sizes (#7506)
       new  74e4c42   FIX #6420: Cloning decision tree estimators breaks criterion objects (#7680)
       new  2caa144   [MGR + 2] fix selectFdr bug (#7490)
       new  cfc280d   switch to multinomial composition for mixture sampling
       new  55672f9   add shape assertions to test
       new  82a6740   Update joblib to 0.10.3 version (#7696)
       new  30b936d   Remove temporary work-around for 0.15 release
       new  707b6f9   [MRG+1] Added unit test for adding classes_ property to GridSearchCV, fixes #6298 (#7661)
       new  716819b   explain learning_curve(shuffle=True) test.
       new  61a9a9b   add missing links for users to whatsnew
       new  568c002   [MRG + 1] Move n_iter and get_params invariance tests to common estimator_checks (#7677)
       new  4e1c101   Use n_components=3 to test actual regression
       new  ee3e617   [MRG+2] adding multilabel support for score_func (#7676)
       new  cd714b1   [MRG] Added warning on keyboard interrupt during MLP fit (#7614)
       new  ad6f094   [MRG+2] switch to multinomial composition for mixture sampling (#7702)
       new  edcb513   removed parameter that was documented as attribute (#7711)
       new  f122efa   Fix typo in OMP author name.
       new  5c60f1f   [MRG+1] Reorder EllipticEnvelope docstring. (#7734)
       new  177ac84   [MRG + 1] Printing the total time in cross_validation (#7640)
       new  74a9756   [MRG+2] Norm inconsistency between RFE and SelectFromModel (was _LearntSelectorMixin) #2121 (#6181)
       new  7892edd   Address #7733 - MultiTaskElasticNet user guide links to MultiTaskLasso (#7741)
       new  8f4ebb5   r2_score - add more doctest examples (#7727)
       new  4ddb744   [MRG + 1] Clarified error msg in plot_partial_dependence (#7673)
       new  5adc832   BF: avoid importing from inside joblib (#7731)
       new  0dfc9a5   [MRG + 1] ElasticNetCV: raise ValueError if l1_ratio=0 (#7591)
       new  3c18735   [MRG+2] logit -> logistic in plot_logistic.py and minor visual improvements to the plot. (#7730)
       new  0d87370   BF: do full pre-cythonization for older systems (borrowed from pandas)
       new  5c2df21   Pre-cythonized
       new  3290fab   changelog entry
       new  bfcab0a   skip test_non_meta_estimators on ppc64el to avoid FTBFS
       new  f260898   [MRG] Correcting length of explained_variance_ratio_, eigen solver (#7632)
       new  3f4524e   [MRG] DOC :issue: role to simplify what's news (#7657)
       new  4da44c8   [MRG+1] replaced some assert_true(np.allclose(x, y)) with assert_almost_equal (#7742)
       new  6b381ae   DOC use target_names over named categories in 20newsgroups example (#7423)
       new  ff5c36e   DOC Correct linking of TruncatedSVD (#7749)
       new  73d3f03   [MRG + 1] FIX raise an error message when n_groups > number of groups (#7681) (#7683)
       new  788a458   [MRG+2] LOF algorithm (Anomaly Detection) (#5279)
       new  9d535ad   DOC: Provide link to LDA and NMF in the example tutorial closes #5876 (#5984)
       new  0ea8e8b   [MRG + 1] fix bug with negative values in cosine_distances (#7732)
       new  014dfbe   BF: correctly depend on versioned _uncythonize, otherwise make is confused and does not see python-install2.7
       new  379f54b   [MRG+2] DOC correct docstring for sample_gaussian (#6957)
       new  3ff22f5   Added citation for sklearn (#7759)
       new  94c2094   [MRG+1] BUG: MultiLabelBinarizer.fit_transform sometimes returns an invalid CSR matrix (#7750)
       new  061803c   [MRG+2] Fixed n**2 memory blowup in _labels_inertia_precompute_dense (#7721)
       new  edc9e7f   [MRG] ENH faster sample_without_replacement for recent numpy (#7703)
       new  e841996   [MRG] CI build pdf docs when master or version branch (#7747)
       new  4c8f12d   DOC parameter is called "sample_weight", not "sample_weights" (#7782)
       new  10323f5   [MRG + 2] ENH add sample_weight support to Pipeline.score (#7723)
       new  4ae1013   [MRG+1] ENH PCA variants now expose singular values (#7685)
       new  6e50c8f   [MRG+2] DOC framework for keeping API refs for deprecated classes/funcs (#7725)
       new  38f6a91   [MRG + 2] FIX Be robust to non re-entrant/ non deterministic cv.split calls (#7660)
       new  1e746e2   DOC Minor changes to README (#7791)
       new  c77003c   [MRG] DOC Isolation Forest decision function doc - Clarify that it computes normality not abnormality. (#7792)
       new  aa57569   Change fit() docstrings (#7783)
       new  2ddf0ce   [MRG + 1] rfecv: verbosity: Set verbose threshold low bound from 2 to 1 (#7644)
       new  99342b6   fix rst errors for pypi (#7800)
       new  b0ac236   DOC remove duplicate versionadded
       new  304c67c   COSMIT delete untrue and obsolete comment
       new  f907aef   DOC add missing heading in what's new (#7780)
       new  6a2d8d5   [MRG+2] BUILD Rewrite setup.py files to handle cython dependencies (#7719)
       new  87cfad3   Fix typo in testimonials.rst
       new  9dc9470   DOC More uses of :issue: in whats_new
       new  bd911ec   [MRG] MAINT: Return self for fit in Spectral Biclustering and CoClustering (#7814)
       new  db3c095   [MRG+1] Warning raised by MultiTaskElasticNet.fit is now ConvergenceWarning. (#7806)
       new  d7c956a   [MRG] FIX Validate and convert X, y and groups to ndarray before splitting (#7593)
       new  02cc6f5   [MRG] TST Ensure __dict__ is unmodified by predict, transform, etc (#7553)
       new  9fd70a8   [MRG+2] Fix for OvR partial_fit bug (#7786)
       new  2e4b589   DOC Added bug fix #7786 to whats_new.rst (#7830)
       new  46001a0   [MRG + 2] FIX bug in svm's decision values when `decision_function_shape` is `'ovr'` (#7724)
       new  3b1541b   [MRG + 3] OneVsRestClassifier: don't expose predict_proba and decision_function if base estimator doesn't support them (#7812)
       new  cd4806a   Disable git pager in flake8_diff.sh
       new  4502520   [MRG + 1] Update funding and website. (#7836)
       new  8f1141b   DOC markup typo in regression.py (#7837)
       new  7272414   DOC : fix docstring of AIC/BIC in GMM
       new  41e1b8f   [MRG+1] label binarizer not used consistently in CalibratedClassifierCV (#7799)
       new  c56925a   [MRG + 1] DOC adding :user: role to whats_new (#7818)
       new  f1ec430   Fix docstring typo (#7844) n_features --> n_components
       new  f32d257   [MRG+2] DOC adding separate `fit()` methods (and docstrings) for DecisionTreeClassifier and DecisionTreeRegressor (#7824)
       new  9667ff2   [MRG + 2] Fixed parameter setting in SelectFromModel (#7764)
       new  31ee1a8   [MRG + 2] FIX LogisticRegressionCV to correctly handle string labels (#5874)
       new  d3e1d4f   We need to touch PKG-INFO when our pre-cythonized files are copied so setup.py does not try to regenerate
       new  5a16dd7   DOC use the class_with_call template for kernels (#7813)
       new  5230382   [MRG + 1] Fix numerical instability in LassoLars when alpha=0 (#7778) (#7849)
       new  9d57e75   DOC fix typo on rescaled image sizes (#7864)
       new  fdb726d   BF: use setup.py build_src for "make cython" (Closes: #7873)
       new  56e3ba3   New upstream version 0.18
       new  d5d42ea   Updated version 0.18 from 'upstream/0.18'
       new  fd94d7f   Re-add debian/ dir
       new  7391fb9   Moved packaging to Debian Science team
       new  80fc9e7   Build-Depends: python-pytest, python3-pytest

The 22492 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails.  The revisions
listed as "adds" were already present in the repository and have only
been added to this reference.


-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/scikit-learn.git



More information about the debian-science-commits mailing list