[mlpack] annotated tag debian/2.2.0-1 created (now 86ce27d)
Barak A. Pearlmutter
barak+git at pearlmutter.net
Thu Mar 23 17:54:04 UTC 2017
This is an automated email from the git hooks/post-receive script.
bap pushed a change to annotated tag debian/2.2.0-1
in repository mlpack.
at 86ce27d (tag)
tagging 24c4cc3f07b09b50444312e894f2af890a7345a6 (commit)
replaces debian/2.1.1-1
tagged by Barak A. Pearlmutter
on Thu Mar 23 17:10:39 2017 +0000
- Log -----------------------------------------------------------------
mlpack Debian release 2.2.0-1
-----BEGIN PGP SIGNATURE-----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=hUeK
-----END PGP SIGNATURE-----
Barak A. Pearlmutter (2):
Merge tag 'upstream/2.2.0'
log changes
Dirk Eddelbuettel (1):
minor documentation update
Govind Gopakumar (3):
Set logging to true, working on reset to previous value now
Added reset flag, but test fails on something else entirely?
Reverted to without any output guards
Kirill Mishchenko (4):
Add Classify in SoftmaxRegression
Change remaining references of Predict to Classify
Change Predict() to wrap Classify()
Add minor changes
Lakshya Agrawal (9):
Added my name to contributors list
Added Test For Gaussian::Distribution
Added Test for Gaussian::Train
Added TestCase for GaussianDistribution::Train(observation, probabilities)
Improved Load to load column vectors
Improved Load to Load col and row arma vectors
Added Test for col and row load
Added test case for load and save
Minor fixes in load_save_test.cpp
Marcus Edel (8):
Remove unsued debug messages.
Minor style fix; use the correct filename.
Update tutorials link.
Decrease the number of parallel builds.
Increase the number of parallel builds (travis).
Minor spelling fix.
Add mlpack logo and use unified badges.
Update link to Keon's excellent tutorial to build mlpack on windows.
Mike Izbicki (2):
Replaced all instances of `#include <mlpack/core.hpp>` with `#include <mlpack/prereqs.hpp>`. This doesn't affect build times when building the whole project, but it greatly reduces build times after modifying only a few files.
moved cmath include for appveyor
Praveen Ch (8):
Case insensitive condition added
Added necessary header files for independant invocation
Added a test file to check the ARFF Loader
Simple test for ARFF Loader
Sample test file for ARFF Loader
Added testcase in load_save_test.cpp
Removed the test file
Added name
Rishabh Kumar (1):
Initialise "success" variable to fix warnings
Ryan Curtin (107):
Backport Load() and Save() functionality into prereqs.hpp.
Style fixes.
More style fixes.
Use mlpack random generator objects.
Increase the number of points to reduce variance in tests.
Add Mike as a contributor.
Update to latest version number.
Use slightly safer setting for LD_LIBRARY_PATH that preserves the old value.
Minor spelling and grammatical fixes.
Fix minor doxygen issue in order to trigger benchmark build.
Transpose is one word, no need for camelcase.
Fix inconsistency.
Fix style and add some comments.
Fix a space I missed.
Refactor DecisionStump to hold children when it is a stump.
Update tests for slightly changed (and not final) API.
Add function to get the number of children.
Fix implementation for move constructor to avoid infinite recursion.
Add decision tree to compilation.
Add decision tree first implementation.
Add tests for first part of decision trees (the gains).
Fix minor bugs in splitting strategies.
Refactor auxiliary infos to be part of the class.
Add some tests for the decision tree.
Add tolerance for floating point errors.
Ensure splits when only two points are in the set.
Test the probabilities also.
Add simple generalization test for numeric features.
Add test for categorical data and stumps.
Add convenience typedef for stumps.
Fix -Wuninitialized and handle NoRecursion correctly.
Fix bug: it was right the first time.
Add first pass at CLI program.
Build main program.
Fix minor bugs with Classify() versions.
Fix bugs in main program.
Document addition of decision tree.
Add documentation.
Revert changes to decision stumps that I didn't mean to push.
Remove unnecessary class names to fix Windows build.
Backport CLI interface to 2.2.x.
Add public overload for weighted learning.
A first pass at DBSCAN. But it doesn't pass the tests.
Don't increment currentCluster when we don't create a cluster.
Initialize bitset correctly, to all 1s.
Set cluster of a point when ProcessPoint() is called.
Fix selection to not return all 0s.
Smaller datasets mean faster tests.
Test for outliers.
Add a test of actual clustering quality.
First pass on command-line program.
Add copy and move constructors.
Fix some DBSCAN implementation bugs.
Declare copy and move constructors.
Add untested implementation for move and copy constructors.
Fix header guards.
Fix minor documentation issues.
Minor style fixes.
Add program documentation.
Transpose on save.
Fix merges.
I love when the error is in the test, not the algorithm.
Adapt decision tree test to new DiscreteDistribution API.
Fix serialization.
Fix syntax and style issues.
Fix Random().
Switch to arma::Col<size_t>.
Invert loops to better use memory.
Fix support for multidimensional discrete distributions.
Update HISTORY.
Use "traditional" initialization to help fix Windows build (hopefully).
Add nice Javascript search box (thanks Dirk for the tip).
Update some contribution dates.
Move around includes to make compile work.
Minor style fixes.
Remove stray random seed.
Fix typo (#914).
Minor style fixes (tabs to spaces, formatting).
Minor formatting fixes and clarifications.
Add forward declarations of new Load() overloads.
Oops, remove extra forward declarations.
Tiny pedantic style changes.
Minor include fix.
Fix Armadillo warning.
Use extern templates to compile Load() overloads, so that spirit doesn't get included.
Add new file.
Set random seed when splitting dataset.
Fix ambiguous namespace.
Avoid warning in a different way.
mlpack_deprecated is lowercase... why did this compile before this change?
Refactor MissingPolicy to handle when the string is not a part of missingTokens.
Refactor MapPolicy classes to also allow a first pass.
Refactor DatasetMapper to allow a first pass over data.
Refactor for new DatasetMapper API.
Significant refactoring to fix ImputationTest.
Fix test compilation.
Remove unreleased features.
Update HISTORY.
Update copyright dates.
Strip Github display from README for release and update dependencies.
Update version.
Update versions for release.
Remove unused file.
Add license.
Add license to DBSCAN and decision tree files.
Remove ann directory from CMake config.
Re-add random_init.hpp since it is used by PCA.
Sudhanshu Ranjan (4):
Update COPYRIGHT.txt
Update core.hpp
Update dbscan.hpp
Update dbscan_impl.hpp
Tiramisu 1993 (1):
Implement multidimensional discrete distribution
dinesh Raj (3):
Refactored range search model to include boost variant and visitor paradigm
Refactored range search model to include boost variant and visitor paradigm
Fixed coding style in Range Search model classes and added my name to copyright.txt and core.hpp
piyush-jaiswal (1):
Added name and email to COPYRIGHT.txt and src/mlpack/core.hpp
s1998 (1):
Added description of function in dbscan.hpp and dbscan_impl.hpp
shikhar (10):
Fix support for printing arma objects with custom precision
Fix include order to include armadillo before mlpack
Fix element alignment with high precision output from prefixedoutstream
Unwrap arma::Op objects for printing in prefixedoutstream
Extract proxy object from abs operation in prefixedoutstream_impl.hpp
Split prefixedoustream tests and fix includes in prefixedoustream.hpp
Remove default type void from std::enable_if in prefixedoutstream.hpp
Avoid use of internal Armadillo structures in prefixedoustream_impl.hpp
Use CheckMatrices in gmm_test.hpp
Update CheckMatrices to support tolerance values
stereomatchingkiss (37):
add overload, able to move string
fix bug--infinite recursive call
first commit
1 : fix bug, did not consider case like "210DM, 1~200"
fix bug--category conversion should based on columns but not rows
remove useless codes
support tsv
use LoadCSV to implement csv/tsv/txt loader
fix bug--LoadCSV should parse txt parse file too
can specify fatal or not if file cannot open
fix bug--should not use empty string to open file
treat \t and space as same category
refine string construct
refine comments
refine comments
refine comments and parsers
add new test cases for strings like "200-DM"
simplify parser by phrase_parse
simnplify parser and refine format
add forward declaration for DatasetInfo, wihtout it the vc2015 compiler cannot find the declaration, weird
add load_csv.hpp and load_csv.cpp
split part of the implementation details into cpp, this may reduce some compile times
remove forward declaration
include mlpack/core.hpp before arma_extend.hpp to prevent some weird bugs
add header load_csv.hpp
change order of sources
remove useless include file
change order of header file
move implementation details from cpp back to hpp
move part of the implementation details to cpp
1 : use extern template to export part of the implementation of Load function
use std::string to replace raw buffer, cpp11 guarantee memory layout is contiguous
1 : fix bug, wrong constructor
fix format
remove useless file
add license
use preprocessor to omit extra instant of Load function under windows
yashu-seth (5):
adds GammaDistribution::Train(observations, probabilities)
adds tests for GammaDistribution::Train(observations, probabilities)
adds my name to core.hpp and COPYRIGHT.txt
fixed typo in perceptron documentation
added tests for SimpleWeightUpdate
-----------------------------------------------------------------------
This annotated tag includes the following new commits:
new 3c0a56c adds GammaDistribution::Train(observations, probabilities)
new c66a2ff adds tests for GammaDistribution::Train(observations, probabilities)
new c226c21 adds my name to core.hpp and COPYRIGHT.txt
new 77a029b Replaced all instances of `#include <mlpack/core.hpp>` with `#include <mlpack/prereqs.hpp>`. This doesn't affect build times when building the whole project, but it greatly reduces build times after modifying only a few files.
new 1f9718f moved cmath include for appveyor
new 6f2601b Backport Load() and Save() functionality into prereqs.hpp.
new d9db269 Style fixes.
new 44a61d8 More style fixes.
new 391fb9b Use mlpack random generator objects.
new 2cff809 Increase the number of points to reduce variance in tests.
new 8f9caaf Add Mike as a contributor.
new 733e806 Update to latest version number.
new 2470574 Use slightly safer setting for LD_LIBRARY_PATH that preserves the old value.
new d2ea6a7 Remove unsued debug messages.
new e73da37 Minor style fix; use the correct filename.
new 484f3a9 fixed typo in perceptron documentation
new 9f0e7ab added tests for SimpleWeightUpdate
new 5aa9b01 Minor spelling and grammatical fixes.
new ceb91e2 Fix minor doxygen issue in order to trigger benchmark build.
new 8db686a Update COPYRIGHT.txt
new d14022c Update core.hpp
new ddb54a2 Added name and email to COPYRIGHT.txt and src/mlpack/core.hpp
new 71cc647 Transpose is one word, no need for camelcase.
new d63943c Update tutorials link.
new bf340bc Decrease the number of parallel builds.
new 0ca3e68 Increase the number of parallel builds (travis).
new a8ec9f5 Fix inconsistency.
new 29992bb Initialise "success" variable to fix warnings
new 13e8c1b Refactored range search model to include boost variant and visitor paradigm
new 7d6f24a Refactored range search model to include boost variant and visitor paradigm
new 378bbc6 Fixed coding style in Range Search model classes and added my name to copyright.txt and core.hpp
new 4d3daa3 Fix style and add some comments.
new 9925e3a Fix a space I missed.
new 7b19f0b Minor spelling fix.
new d895826 Refactor DecisionStump to hold children when it is a stump.
new b4fb1d9 Update tests for slightly changed (and not final) API.
new 41fa7af Add function to get the number of children.
new ab28ece Fix implementation for move constructor to avoid infinite recursion.
new 495098a Add decision tree to compilation.
new 4d3f1fa Add decision tree first implementation.
new d25083f Add tests for first part of decision trees (the gains).
new f08b674 Fix minor bugs in splitting strategies.
new 4a82bbc Refactor auxiliary infos to be part of the class.
new 4483792 Add some tests for the decision tree.
new 7d7ef76 Add tolerance for floating point errors.
new 0141d5f Ensure splits when only two points are in the set.
new 02d8fb5 Test the probabilities also.
new 6f9083d Add simple generalization test for numeric features.
new d62eaf2 Add test for categorical data and stumps.
new f6bf8ba Add convenience typedef for stumps.
new 41f57e3 Fix -Wuninitialized and handle NoRecursion correctly.
new d0a5f54 Fix bug: it was right the first time.
new c37dcad Add first pass at CLI program.
new 968cff3 Build main program.
new b6f29d7 Fix minor bugs with Classify() versions.
new c3b82d4 Fix bugs in main program.
new add2e1b Document addition of decision tree.
new 6069db9 Add documentation.
new a2679d6 Revert changes to decision stumps that I didn't mean to push.
new 9445e21 Remove unnecessary class names to fix Windows build.
new 682e938 Backport CLI interface to 2.2.x.
new b880154 Add public overload for weighted learning.
new 9ff5c13 Added my name to contributors list
new 12ab06d Added Test For Gaussian::Distribution
new 3cb7d9b Added Test for Gaussian::Train
new 5a47b85 Added TestCase for GaussianDistribution::Train(observation, probabilities)
new 22db8b7 A first pass at DBSCAN. But it doesn't pass the tests.
new 94d19e3 Don't increment currentCluster when we don't create a cluster.
new 63f3eb9 Initialize bitset correctly, to all 1s.
new 061ee78 Set cluster of a point when ProcessPoint() is called.
new 1a7ef69 Fix selection to not return all 0s.
new a45e65f Smaller datasets mean faster tests.
new 95143cf Test for outliers.
new 87d85a3 Add a test of actual clustering quality.
new e1e6af8 First pass on command-line program.
new 2f74df3 Add copy and move constructors.
new 5889848 Fix some DBSCAN implementation bugs.
new 163c7a8 Declare copy and move constructors.
new 233d832 Add untested implementation for move and copy constructors.
new 6ed3156 Update dbscan.hpp
new 43ad99c Update dbscan_impl.hpp
new 4d6d10b Added description of function in dbscan.hpp and dbscan_impl.hpp
new 155033e Fix header guards.
new 81b069f Fix minor documentation issues.
new 9fdf710 Minor style fixes.
new d0b151a Add program documentation.
new 63ab288 Transpose on save.
new 2c4052d Fix merges.
new f7eea76 minor documentation update
new 4b1afb3 I love when the error is in the test, not the algorithm.
new e179397 Implement multidimensional discrete distribution
new d27edce Adapt decision tree test to new DiscreteDistribution API.
new effe15d Fix serialization.
new 960194d Fix syntax and style issues.
new f698dac Fix Random().
new 6fa43e0 Switch to arma::Col<size_t>.
new 0217359 Invert loops to better use memory.
new aabb5cf Fix support for multidimensional discrete distributions.
new 0f730e5 Update HISTORY.
new a69e1ba Use "traditional" initialization to help fix Windows build (hopefully).
new 50c24b4 Add nice Javascript search box (thanks Dirk for the tip).
new cf0ad66 Update some contribution dates.
new 951a660 Add mlpack logo and use unified badges.
new 18942ed Fix support for printing arma objects with custom precision
new d6d75a7 Fix include order to include armadillo before mlpack
new ab13cc1 Fix element alignment with high precision output from prefixedoutstream
new d8f8891 Unwrap arma::Op objects for printing in prefixedoutstream
new 7d7cf93 Extract proxy object from abs operation in prefixedoutstream_impl.hpp
new f70b47f Split prefixedoustream tests and fix includes in prefixedoustream.hpp
new ea2f548 Remove default type void from std::enable_if in prefixedoutstream.hpp
new f616691 Move around includes to make compile work.
new e70270d Minor style fixes.
new 979d3db Set logging to true, working on reset to previous value now
new 9ef1a8d Added reset flag, but test fails on something else entirely?
new 55ec7af Reverted to without any output guards
new 11906d9 Remove stray random seed.
new 2677006 Fix typo (#914).
new 80c4504 Case insensitive condition added
new 9fe06bb Added necessary header files for independant invocation
new 400e541 Added a test file to check the ARFF Loader
new 140131b Simple test for ARFF Loader
new 095e1e6 Sample test file for ARFF Loader
new 010a88e Added testcase in load_save_test.cpp
new 10f404c Removed the test file
new bffa0a5 Added name
new 64a3bb2 Minor style fixes (tabs to spaces, formatting).
new c62f254 Improved Load to load column vectors
new 5bd0989 Improved Load to Load col and row arma vectors
new 48a460f Added Test for col and row load
new 839f55a Added test case for load and save
new 1cd3567 Minor fixes in load_save_test.cpp
new 820f7a0 Minor formatting fixes and clarifications.
new 75d9916 Add forward declarations of new Load() overloads.
new 5694552 Oops, remove extra forward declarations.
new dfbbbe0 Update link to Keon's excellent tutorial to build mlpack on windows.
new 69c90e2 Avoid use of internal Armadillo structures in prefixedoustream_impl.hpp
new 6fe8a3b Tiny pedantic style changes.
new 9601a41 Use CheckMatrices in gmm_test.hpp
new 1e101a7 Update CheckMatrices to support tolerance values
new b7e60e9 Minor include fix.
new 3aa3823 add overload, able to move string
new eafbf33 fix bug--infinite recursive call
new c1e1649 first commit
new 47b726e 1 : fix bug, did not consider case like "210DM, 1~200" 2 : fix bug, cannot parse transpose file with correct result
new 3f88f91 fix bug--category conversion should based on columns but not rows
new 94f3642 remove useless codes
new d9c76da support tsv
new 2b5b84f use LoadCSV to implement csv/tsv/txt loader
new e06dc6f fix bug--LoadCSV should parse txt parse file too
new d548cff can specify fatal or not if file cannot open
new 2c5af4d fix bug--should not use empty string to open file
new f5ecb50 treat \t and space as same category
new 7a3b13b refine string construct
new 7981888 refine comments
new ff58df8 refine comments
new 17fe89c refine comments and parsers
new b292fad add new test cases for strings like "200-DM"
new 82093f0 simplify parser by phrase_parse
new d2fceb5 simnplify parser and refine format
new bc83001 add forward declaration for DatasetInfo, wihtout it the vc2015 compiler cannot find the declaration, weird
new 74a1486 add load_csv.hpp and load_csv.cpp
new c9da0c9 split part of the implementation details into cpp, this may reduce some compile times
new 4948ddb remove forward declaration
new 4c1a846 include mlpack/core.hpp before arma_extend.hpp to prevent some weird bugs
new 960806a add header load_csv.hpp
new bcbf263 change order of sources
new e64d1e9 remove useless include file
new 112f73a change order of header file
new 820519f move implementation details from cpp back to hpp
new 6a47f98 move part of the implementation details to cpp
new 4e3ed90 1 : use extern template to export part of the implementation of Load function 2 : use LoadCSV to relpace implementation of the part of csv loading
new 92e5b15 Fix Armadillo warning.
new 40fda23 Use extern templates to compile Load() overloads, so that spirit doesn't get included.
new d092de8 Add new file.
new c72cb2b use std::string to replace raw buffer, cpp11 guarantee memory layout is contiguous
new c9f67a4 1 : fix bug, wrong constructor 2 : use std::string to replace raw buffer
new 347c8d6 fix format
new d512789 remove useless file
new 97711d8 add license
new 9a8ca1a use preprocessor to omit extra instant of Load function under windows
new 11fc169 Add Classify in SoftmaxRegression
new 1e54ed1 Change remaining references of Predict to Classify
new 0b49d05 Change Predict() to wrap Classify()
new 41cd00c Add minor changes
new 52be857 Set random seed when splitting dataset.
new 7af28c7 Fix ambiguous namespace.
new b1e6cb1 Avoid warning in a different way.
new 1547077 mlpack_deprecated is lowercase... why did this compile before this change?
new e4483b3 Refactor MissingPolicy to handle when the string is not a part of missingTokens.
new 2f22721 Refactor MapPolicy classes to also allow a first pass.
new 22c3192 Refactor DatasetMapper to allow a first pass over data.
new 4121399 Refactor for new DatasetMapper API.
new 0f14f37 Significant refactoring to fix ImputationTest.
new e4551bf Fix test compilation.
new 4d6c2a4 Remove unreleased features.
new ef20415 Update HISTORY.
new b8be8f5 Update copyright dates.
new 9f7bac8 Strip Github display from README for release and update dependencies.
new 38ada5a Update version.
new c35183c Update versions for release.
new 21a1663 Remove unused file.
new 98477ea Add license.
new d9164ab Add license to DBSCAN and decision tree files.
new 95867f7 Remove ann directory from CMake config.
new 40bf682 Re-add random_init.hpp since it is used by PCA.
new 101149b Merge tag 'upstream/2.2.0'
new 24c4cc3 log changes
The 207 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/mlpack.git
More information about the debian-science-commits
mailing list