[r-cran-mcmcpack] 65/90: Imported Upstream version 1.2-3
Andreas Tille
tille at debian.org
Fri Dec 16 09:07:49 UTC 2016
This is an automated email from the git hooks/post-receive script.
tille pushed a commit to branch master
in repository r-cran-mcmcpack.
commit bba73acb5b6ba488fe9bfbb69981347f84e2f215
Author: Andreas Tille <tille at debian.org>
Date: Fri Dec 16 08:07:22 2016 +0100
Imported Upstream version 1.2-3
---
DESCRIPTION | 8 ++++----
HISTORY | 7 +++++++
MD5 | 22 ++++++++++++----------
R/MCMCbinaryChange.R | 0
R/btsutil.R | 26 --------------------------
R/hidden.R | 4 ++--
R/make.breaklist.R | 26 ++++++++++++++++++++++++++
R/testpanelSubjectBreak.R | 11 +++--------
confdefs.h | 16 ++++++++++++++++
data/PErisk.rda | Bin 1966 -> 1967 bytes
data/SupremeCourt.rda | Bin 470 -> 470 bytes
man/HMMpanelFE.Rd | 6 +++---
man/testpanelSubjectBreak.Rd | 14 +++++---------
src/MCMCpoissonChange.cc | 8 +++++++-
14 files changed, 85 insertions(+), 63 deletions(-)
diff --git a/DESCRIPTION b/DESCRIPTION
index 1dc4999..8daa241 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
Package: MCMCpack
-Version: 1.2-2
-Date: 2012-3-2
+Version: 1.2-3
+Date: 2012-4-16
Title: Markov chain Monte Carlo (MCMC) Package
Author: Andrew D. Martin, Kevin M. Quinn, and Jong Hee Park
Maintainer: Jong Hee Park <jhp at uchicago.edu>
@@ -18,6 +18,6 @@ Description: This package contains functions to perform Bayesian
License: GPL-3
SystemRequirements: gcc (>= 4.0)
URL: http://mcmcpack.wustl.edu
-Packaged: 2012-03-02 21:27:17 UTC; adm
+Packaged: 2012-04-16 03:27:11 UTC; jongheepark
Repository: CRAN
-Date/Publication: 2012-03-03 06:49:19
+Date/Publication: 2012-04-16 06:09:44
diff --git a/HISTORY b/HISTORY
index 795db05..95948b2 100644
--- a/HISTORY
+++ b/HISTORY
@@ -1,6 +1,13 @@
//
// Changes and Bug Fixes
//
+
+1.2-2 to 1.2-3
+ * fixed extraneous warnings in hidden.R (thanks to Jeffrey Arnold)
+ * added make.breaklist.R
+ * fixed example codes in HMMpanelFE.Rd and testpanelSubjectBreak.Rd
+ * fixed testpanelSubjectBreak() to handle missing values in subject.id
+
1.2-1 to 1.2-2
* fixed bug in dwish() (thanks to Taráz E. Buck)
* fixed .onAttach() to no longer require packages and use packageStartupMessage()
diff --git a/MD5 b/MD5
index 130294a..e263329 100644
--- a/MD5
+++ b/MD5
@@ -1,5 +1,5 @@
-dd3a6642c422701839fb4c4b6fc003dc *DESCRIPTION
-ed15423ad364f55ae6092d7ab7db5683 *HISTORY
+a922886738229c9f9430e5da5a1d9273 *DESCRIPTION
+fae7df2629b8b9da9d1a7fb07deefa5a *HISTORY
45b74f9b9cce80e9215266776a1be85c *LICENSE
4851e0ecf23d44b58fcd53e597210689 *NAMESPACE
1eaf24e8d468ac0b624a9aa90cded306 *R/BayesFactors.R
@@ -40,29 +40,31 @@ e27dcf535f23c25018bb778b61640f22 *R/MCMCquantreg.R
3a98b4d919a265b46c29d14310bb90d4 *R/SSVSquantreg.R
555c87acdcc2f8572b14151e29022e20 *R/SSVSquantregsummary.R
fdfcfe9909cadaf15c47bbb604be2257 *R/automate.R
-9eb132545aaf170132109e198464b69d *R/btsutil.R
+ad65af39e4dc71e5a2facf7d5ad62317 *R/btsutil.R
56829ce5340d26f9488b8b9894ff5cd0 *R/distn.R
18daa9fee800c7d8d5827405738531e2 *R/hidden-hmodels.R
-0f904468b8ce8fc5c3869f87d11da905 *R/hidden.R
+4dc19c1cb7809d9be5e7353929358aa1 *R/hidden.R
+317675992da250e591e78cc96b10581b *R/make.breaklist.R
f799711e5d3fb7140cbc0995360339b3 *R/procrust.R
2e00b2b1593d22b5d1c579e92b294106 *R/scythe.R
4de1106e38b8ca3b12396ec0d84bba4c *R/testpanelGroupBreak.R
-cb8aeff19ebf8c85748bad1a57484e7a *R/testpanelSubjectBreak.R
+0907ed9bbd42212d50c6208dce57d640 *R/testpanelSubjectBreak.R
c67ac5238de55cdc8c49d017e0226ed9 *R/tomog.R
6377da3453982f578df4a886f673122c *R/utility.R
7e2021def8a88edb41fe78d0a0ae642c *R/zzz.R
e42f762c27e26e0ace372c03c47ac4e5 *README
46e955baeb55718a380a11fb34d2ea67 *cleanup
+a315f6c986020fa2b8a8190a77bd0015 *confdefs.h
e09d1c5ac6250f832af90566a9e565fb *configure
21e1a006fb388ed39daf16e6d2ec9549 *configure.ac
e17202d22ce5ab5ab7b01e3b248a4008 *data/Nethvote.rda
-d03a27617f7933fbbb47732af740b129 *data/PErisk.rda
+9926c591df7085545233c2cf11b6b56f *data/PErisk.rda
93e636ac58b966f9a74315f51be07d25 *data/Rehnquist.rda
accb7b5b46207a02e49de2702a6faff4 *data/Senate.rda
-2b95cdf5122fe0f86191ba668da542dc *data/SupremeCourt.rda
+f740448e1b0cff40e7c05a354ea69e59 *data/SupremeCourt.rda
a5cfcf73e21c03eeaf4f3baa5c492c14 *inst/CITATION
2b42855343b5f0c51c7e9411aef57509 *man/BayesFactor.Rd
-872365fa451a1717a950a8102453eb8e *man/HMMpanelFE.Rd
+c3e38d2c8a1adee2ff096056416ca6ec *man/HMMpanelFE.Rd
2e322e903df0d9c20513670710537995 *man/HMMpanelRE.Rd
293ff17f852d0e60dd1aac1f419353d2 *man/MCMCSVDreg.Rd
9af751257fcacc3a57543e4ae6a5e6cc *man/MCMCbinaryChange.Rd
@@ -119,7 +121,7 @@ c64abc923f2c6f56a167a815c881ffb1 *man/plotState.Rd
6a88d877e88b5abfd7a66d659c9b4e6a *man/procrust.Rd
41feb17dda3b00cffa2f6176ba584b74 *man/readscythe.Rd
f151473ede68e2cff21c06afd3060373 *man/testpanelGroupBreak.Rd
-8a3c03825fcdcdf22359faa43b330574 *man/testpanelSubjectBreak.Rd
+24a4ef81a3573532c757c0c1dad393eb *man/testpanelSubjectBreak.Rd
254168bb5258fd00a7b766fb7d3bbfd9 *man/tomog.Rd
f5ba8eb700c67f188505a4585dd39f6b *man/topmodels.Rd
a4c560360ba912bb342a429e434bcc96 *man/vech.Rd
@@ -156,7 +158,7 @@ a78375184abf7c97bf7fbf3f7614cad3 *src/MCMCoprobit.cc
387f517e5df0bde2b483c8986ce10983 *src/MCMCoprobitChange.cc
6addbe81efb51cb4b8401e9a6931f782 *src/MCMCordfactanal.cc
6bfed80bb7c109537461e6381837b6c4 *src/MCMCpoisson.cc
-628013b775615c38951a2d931e1c1fe4 *src/MCMCpoissonChange.cc
+73fb68d449004927f02a4bbf764258f6 *src/MCMCpoissonChange.cc
84b94f6e67b026a915a23fb85bdaf00b *src/MCMCprobit.cc
8a2f4a79c6bc951310074f5757041424 *src/MCMCprobitChange.cc
a6a01c3906a8b92b9f1a12cfa92a1a63 *src/MCMCprobitres.cc
diff --git a/R/MCMCbinaryChange.R b/R/MCMCbinaryChange.R
old mode 100644
new mode 100755
diff --git a/R/btsutil.R b/R/btsutil.R
index 872857a..03bd069 100644
--- a/R/btsutil.R
+++ b/R/btsutil.R
@@ -249,29 +249,3 @@
-## find the most reasonable model
-## by choosing a model with a larger break number
-## If the difference of natural log marginal likelihood is smaller than the threshold,
-## choose the model with a smaller break number
-## to avoid the over-identification problem
-"make.breaklist" <- function(BF, threshold=3){
- eBF <- exp(BF)
- N <- nrow(BF)
- out <- rep(NA, N)
- for (i in 1:N){
- order.i <- order(eBF[i,], decreasing=TRUE)
- if(eBF[i, order.i[1]] / eBF[i, order.i[2]] > threshold){
- out[i] <- order.i[1]
- }
- else if (eBF[i, order.i[1]] / eBF[i, order.i[2]] < threshold & order.i[1]<=order.i[2]){
- out[i] <- order.i[1]
- }
- else if (eBF[i, order.i[1]] / eBF[i, order.i[2]] < threshold & order.i[1]>order.i[2]){
- out[i] <- order.i[2]
- }
- else{
- cat("\n Error occurs at i = ", i)
- }
- }
- return(out-1)
-}
diff --git a/R/hidden.R b/R/hidden.R
index 118d134..53d8efc 100644
--- a/R/hidden.R
+++ b/R/hidden.R
@@ -230,7 +230,7 @@
Lambda.ineq.constraints){
Lambda <- matrix(0, K, factors)
- if (is.na(lambda.start)){# sets Lambda to equality constraints & 0s
+ if (any(is.na(lambda.start))){# sets Lambda to equality constraints & 0s
for (i in 1:K){
for (j in 1:factors){
if (Lambda.eq.constraints[i,j]==-999){
@@ -413,7 +413,7 @@
function(psi.start, X){
K <- ncol(X)
- if (is.na(psi.start)){
+ if (any(is.na(psi.start))){
Psi <- 0.5 * diag(diag(var(X)))
}
else if (is.double(psi.start) &&
diff --git a/R/make.breaklist.R b/R/make.breaklist.R
new file mode 100644
index 0000000..6ccd62a
--- /dev/null
+++ b/R/make.breaklist.R
@@ -0,0 +1,26 @@
+"make.breaklist" <- function(BF, threshold=3){
+ eBF <- exp(BF)
+ N <- nrow(BF)
+ out <- rep(NA, N)
+ for (i in 1:N){
+ order.i <- order(eBF[i,], decreasing=TRUE)
+ if(sum(is.na(eBF[i,]))>0){
+ out[i] <- 1
+ }
+ else{
+ if(eBF[i, order.i[1]] / eBF[i, order.i[2]] > threshold){
+ out[i] <- order.i[1]
+ }
+ else if (eBF[i, order.i[1]] / eBF[i, order.i[2]] < threshold & order.i[1]<=order.i[2]){
+ out[i] <- order.i[1]
+ }
+ else if (eBF[i, order.i[1]] / eBF[i, order.i[2]] < threshold & order.i[1]>order.i[2]){
+ out[i] <- order.i[2]
+ }
+ else{
+ cat("\n Error occurs at i = ", i)
+ }
+ }
+ }
+ return(out-1)
+}
diff --git a/R/testpanelSubjectBreak.R b/R/testpanelSubjectBreak.R
index 7128cf4..7e4766c 100644
--- a/R/testpanelSubjectBreak.R
+++ b/R/testpanelSubjectBreak.R
@@ -3,13 +3,14 @@
##
## written by Jong Hee Park 03/2009
## modified and integrated with other codes by JHP 07/2011
+## fixed a starting.id and ending.id
######################################################################
"testpanelSubjectBreak" <-
function(subject.id, time.id, resid, max.break=2, minimum = 10,
mcmc=1000, burnin=1000, thin=1, verbose=0,
b0, B0, c0, d0, a = NULL, b = NULL, seed = NA,
- Time = NULL, start.id = NULL, end.id = NULL, ps.out = FALSE){
+ Time = NULL, ps.out = FALSE){
## seeds
seeds <- form.seeds(seed)
lecuyer <- seeds[[1]]
@@ -30,12 +31,6 @@
if (is.null(Time)){
Time <- rep(N, 1)
}
- if (is.null(start.id)){
- start.id <- min(subject.id)
- }
- if (is.null(end.id)){
- end.id <- max(subject.id)
- }
NC <- length(unique(subject.id))
time.list <- as.numeric(table(subject.id))
@@ -55,7 +50,7 @@
else {
psout <- array(NA, c(max(time.list), sum(2:(max.break+1)), NC))
}
- for (i in start.id:end.id){
+ for (i in 1:NC){
residual <- resid.list[[i]]
nk <- length(residual)
out <- as.list(rep(NA, max.break))
diff --git a/confdefs.h b/confdefs.h
new file mode 100644
index 0000000..9e17b72
--- /dev/null
+++ b/confdefs.h
@@ -0,0 +1,16 @@
+#define PACKAGE_NAME ""
+#define PACKAGE_TARNAME ""
+#define PACKAGE_VERSION ""
+#define PACKAGE_STRING ""
+#define PACKAGE_BUGREPORT ""
+#define STDC_HEADERS 1
+#define HAVE_SYS_TYPES_H 1
+#define HAVE_SYS_STAT_H 1
+#define HAVE_STDLIB_H 1
+#define HAVE_STRING_H 1
+#define HAVE_MEMORY_H 1
+#define HAVE_STRINGS_H 1
+#define HAVE_INTTYPES_H 1
+#define HAVE_STDINT_H 1
+#define HAVE_UNISTD_H 1
+#define HAVE_TRUNC 1
diff --git a/data/PErisk.rda b/data/PErisk.rda
index aaf01b7..cb0f7de 100644
Binary files a/data/PErisk.rda and b/data/PErisk.rda differ
diff --git a/data/SupremeCourt.rda b/data/SupremeCourt.rda
index 9ffadab..31d84ca 100644
Binary files a/data/SupremeCourt.rda and b/data/SupremeCourt.rda differ
diff --git a/man/HMMpanelFE.Rd b/man/HMMpanelFE.Rd
index c7e4b3a..8c6913a 100644
--- a/man/HMMpanelFE.Rd
+++ b/man/HMMpanelFE.Rd
@@ -162,10 +162,10 @@
## draw time-varying individual effects and sample y
j = 1
true.sigma.alpha <- 30
- true.alpha <- true.alpha1 <- true.alpha2 <- rep(NA, N)
+ true.alpha1 <- true.alpha2 <- rep(NA, N)
for (i in 1:N){
Xi <- X[j:(j+T-1), ]
- true.mean <- Xi%*%true.beta
+ true.mean <- Xi \%*\% true.beta
weight <- Weight[j:(j+T-1)]
true.alpha1[i] <- rnorm(1, 0, true.sigma.alpha)
true.alpha2[i] <- -1*true.alpha1[i]
@@ -187,7 +187,7 @@
b0=0, B0=1/100, c0=2, d0=2, Time = time.id)
## get the estimated break numbers
- estimated.breaks <- make.breaklist(BF, thresh=3)
+ estimated.breaks <- make.breaklist(BF, threshold=3)
## model fitting
out <- HMMpanelFE(subject.id = id, y, X=X, m = estimated.breaks,
diff --git a/man/testpanelSubjectBreak.Rd b/man/testpanelSubjectBreak.Rd
index 1365fd0..98cb96e 100644
--- a/man/testpanelSubjectBreak.Rd
+++ b/man/testpanelSubjectBreak.Rd
@@ -10,7 +10,7 @@
\usage{testpanelSubjectBreak(subject.id, time.id, resid, max.break=2,
minimum = 10, mcmc=1000, burnin=1000, thin=1, verbose=0,
b0, B0, c0, d0, a = NULL, b = NULL, seed = NA,
- Time = NULL, start.id = NULL, end.id = NULL, ps.out = FALSE)}
+ Time = NULL, ps.out = FALSE)}
\arguments{
\item{subject.id}{A numeric vector indicating the group number. It should start from 1.}
@@ -57,11 +57,6 @@
\item{Time}{Times of the observations. This will be used to find the time of the first observations in panel residuals. }
- \item{start.id}{If users want to check the residual breaks within a subset of panel subjects, users can specify the starting subject id and
- the ending subject id.}
-
- \item{end.id}{See start.id.}
-
\item{ps.out}{If ps.out == TRUE, state probabilities are exported. If the number of panel subjects is huge, users can turn it off to save memory.}
\item{...}{further arguments to be passed}
@@ -145,10 +140,10 @@
## draw time-varying individual effects and sample y
j = 1
true.sigma.alpha <- 30
- true.alpha <- true.alpha1 <- true.alpha2 <- rep(NA, N)
+ true.alpha1 <- true.alpha2 <- rep(NA, N)
for (i in 1:N){
Xi <- X[j:(j+T-1), ]
- true.mean <- Xi%*%true.beta
+ true.mean <- Xi \%*\% true.beta
weight <- Weight[j:(j+T-1)]
true.alpha1[i] <- rnorm(1, 0, true.sigma.alpha)
true.alpha2[i] <- -1*true.alpha1[i]
@@ -170,7 +165,8 @@
b0=0, B0=1/100, c0=2, d0=2, Time = time.id)
## estimated break numbers
- estimated.breaks <- make.breaklist(BF, thresh=3)
+ ## thresho
+ estimated.breaks <- make.breaklist(BF, threshold=3)
## print all posterior model probabilities
print(attr(BF, "model.prob"))
diff --git a/src/MCMCpoissonChange.cc b/src/MCMCpoissonChange.cc
index 791d891..795bfd8 100644
--- a/src/MCMCpoissonChange.cc
+++ b/src/MCMCpoissonChange.cc
@@ -7,6 +7,7 @@
//
// 07/06/2007
// 12/20/2007 included in MCMCpack
+// 03/09/2012 fixed a bug in a random uniform draw (thanks to Matt Blackwell)
#ifndef MCMCPOISSONCHANGE_CC
#define MCMCPOISSONCHANGE_CC
@@ -85,7 +86,12 @@ Matrix<> tau_comp_sampler(rng<RNGTYPE>& stream,
}
else {
- Matrix<> ut = stream.runif(yt, 1);
+ Matrix<> ut = stream.runif(yt, 1);
+ // redraw the uniform if there are any repeats
+ // thanks to Matt Blackwell
+ while (::unique(ut).size() != ut.size()) {
+ ut = stream.runif(yt, 1);
+ }
Matrix<> sort_ut = ::sort(ut);
Matrix<> tau_tj(yt, 1);
for(int i=1; i<yt; ++i){
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/r-cran-mcmcpack.git
More information about the debian-science-commits
mailing list