2015-07-19 warnes * [r2058] R/ci.R, man/ci.Rd: ci.binom() was using an incorrect method for calculating binomial confidence interval. The revised code calculates the Clopper-Pearson 'exect' interval, which is *conservative* due to the discrete nature of the binomial distribution. 2015-05-02 warnes * [r2018] Rename 'trunk' to 'pkg' for compatibility with R-forge 2015-04-06 warnes * [r1916] Add ChangeLog files to repository 2014-07-24 warnes * [r1869] Update NEWS for gmodels 2.16.0 * [r1868] - Estimable now adds the class 'estimable' to returned objects. - New ci() method for estimable objects. - Minor improvemets to man page formatting. 2013-07-18 warnes * [r1710] Looks like Brian Ripley repackaged for R 3.0.0 and bumped version number, so change it to 2.15.5 * [r1709] Update for gmodels 2.15.4 * [r1708] Update to current Rd syntax * [r1707] Correct bug in estimable.mlm 2013-07-15 warnes * [r1706] Remove unused argument to ci.mer 2012-06-28 warnes * [r1577] Update for gmodels version 2.15.3. * [r1576] Move percentile() function to a separate file. * [r1575] Update est.mer() to support new S4 "mer" class. * [r1574] Make lme4 example executable. 2012-06-27 warnes * [r1573] Add test code submitted by Ariel.Muldoon@oregonstate.edu. 2012-04-19 warnes * [r1528] Update for release 2.15.2 * [r1527] Update version and date. * [r1526] The 'Design' package has been replaced my 'rms', so update man page references. * [r1525] More fixes for support of S4 'mer' class from lme4 package. * [r1524] Split long line. * [r1523] Changes to pass R CMD check 2011-12-14 warnes * [r1521] Improve formatting of ci.mer(). * [r1520] Modify est.mer to work with recent lme4 'mer' S4 objects. 2011-01-16 warnes * [r1466] Fix warnings reported by R CMD check. Update version number to 2.15.1. 2009-05-09 warnes * [r1337] Add tests for lme4 'mer' objects * [r1336] Update for 2.15.0 * [r1335] Update description for 2.15.0 * [r1334] Add support for lme4's 'mer' objects * [r1333] Add support for lme4's 'mer' objects * [r1332] Fix .Rd syntax error * [r1331] Add softlinks for ChangeLog and NEWS to top level dir for convenience * [r1330] Move ChangeLog and NEWS files into inst directory * [r1329] Update Greg's email address 2008-04-10 warnes * [r1255] Improve languages a bit 2008-01-02 warnes * [r1236] Update Marc's email address 2007-12-12 warnes * [r1233] Move copyright notice for Randall's contributions from License section to Author section of the DESCRIPTION file. 2007-12-07 warnes * [r1232] Update DESCRIPTION and NEWS for release 2.14.1 * [r1231] Correct minor typos in man page for estimable() * [r1230] Add support for lme models to estimable() * [r1229] Replace non-ascii characters in Soren's name with (equivalent?) ascii character to avoid character encoding issues. 2007-10-22 warnes * [r1196] Clarify GPL version 2007-07-26 warnes * [r1105] Add support for mlm to estimable(). * [r1104] Add estimable method for mlm objects 2007-03-09 warnes * [r1079] Remove stray character * [r1078] Update NEWS file. * [r1077] Update version number * [r1076] Minor code formatting changes * [r1075] Flip lower and upper interval in ci.lmer(). Add example to man page. * [r1074] Fix some old email addressses that got missed 2006-11-29 warnes * [r1029] Update for 2.13.1 * [r1028] Correct declartion of S3 methods for estimable() * [r1027] Add additional suggested packages * [r1026] - Add generic - Fix code vs. doc inconsistiencies 2006-11-28 warnes * [r1025] Remove extraneous comma that causes errors in R 2.5.0 2006-11-27 warnes * [r1016] Update for 2.13.1 * [r1015] Add missing export of methods for estimable() 2006-11-14 ggorjan * [r1012] Removed executable property 2006-08-02 warnes * [r977] Update my email address 2006-06-06 nj7w * [r966] Updated ci, estimable and fit.contrast as per Randall Johnson 2006-06-05 nj7w * [r965] Additions as per Randall C Johnson * [r964] Additions as per Randall C Johnson * [r963] - New function to estimate CI's and p-values using mcmcsamp() from the Matrix package 2006-05-05 nj7w * [r959] Fixed an error: According to Marc Schwartz - there was an error when a matrix without dimnames(or names(dimnames)) was passed as x argument 2005-12-13 nj7w * [r808] Removed ChangeLog * [r807] Updated NEWS 2005-12-12 nj7w * [r796] Updated version number for CRAN 2005-12-04 warnes * [r781] Update for 2.11.0 * [r780] Integration of code changes suggested by Randall C Johnson to add support for lmer (lme version 4) objects to ci(), estimable(), and fit.contrast(). Addition of simplified coefficient specificaiton for estimable() based on a function provided by Randall C Johnson. It is now possible to do things like: estimable(reg, c("xB"=1,"xD"=-1) ) instead of: estimable(reg, c( 0, 1, 0, -1) ) which should make estimable much easier to use for large models. 2005-12-01 nj7w * [r776] Updated Greg's email address 2005-10-27 warnes * [r709] Update version number. Bump minor version since we added functionality. * [r708] Add ci.binom() to NAMESPACE, bump version 2005-10-26 warnes * [r707] Add ci.binom 2005-10-25 warnes * [r706] Add gdata::nobs to import list. Needed by ci() 2005-09-12 nj7w * [r671] Updated Greg's email 2005-09-07 nj7w * [r667] Fixed man page 2005-09-06 nj7w * [r664] Updated DESCRIPTION * [r663] Added NEWS * [r662] Fixed the Package name 2005-09-02 nj7w * [r655] Added ChangeLog 2005-08-31 nj7w * [r644] Added DESCRIPTION file * [r643] removed DESCRIPTION.in 2005-07-11 nj7w * [r627] Revision based on Marc Schwartz's suggestions: 1) Added 'dnn' argument to enable specification of dimnames as per table() 2) Corrected bug in SPSS output for 1d table, where proportions were being printed and not percentages ('%' output) 2005-06-09 nj7w * [r625] Updating the version number, and various help files to synchronize splitting of gregmisc bundle in 4 individual components. * [r623] Updates by Marc Schwartz: CrossTable: # Revision 2.0 2005/04/27 # Added 'format = "d"' to all table count output # so that large integers do not print in # scientific notation 2005-05-13 nj7w * [r621] 1) Using dQuote.ascii function in read.xls as the new version of dQuote doesn't work proprly with UTF-8 locale. 2) Modified CrossTable.Rd usage in gmodels 3) Modified heatmap.2 usage in gplots. 2005-05-11 warnes * [r620] Add dependency on gdata::frameApply. 2005-03-31 warnes * [r593] Add ceofFrame function to NAMESPACE * [r592] coefFrame example needs to properly load ELISA data from gtools package * [r588] Ensure that each file has $Id$ header, and no $Log$ * [r587] Add coefFrame() function contributed by Jim Rogers 2005-01-18 warnes * [r521] Removed Windows Line Endings 2005-01-14 nj7w * [r518] Updated the manual to reflect prop.chisq change in its R file. 2005-01-14 warnes * [r517] Nitin added display of the Chisquare contribution of each cell, as suggested by Greg Snow. 2005-01-12 warnes * [r515] Add dependency on R 1.9.0+ to prevent poeple from installing on old versions of R which don't support namespaces. 2004-12-23 nj7w * [r507] Split the function print.CrossTable.vector in two parts - for SAS behaiour and SPSS behaviour. Also put the code of printing statistics in a function 'print.statistics' 2004-12-21 warnes * [r502] Added & extended changes made by Nitin to implement 'SPSS' format, as suggested by Dirk Enzmann . 2004-09-30 warneg * [r464] Fix typos. 2004-09-27 warneg * [r461] Updated to pass R CMD check. 2004-09-03 warneg * [r450] Add explicit package to call to quantcut in example. * [r446] initial bundle checkin 2004-09-02 warneg * [r442] Initial revision 2004-05-25 warnes * [r327] Updates from Mark Schwartz. 2004-04-13 warnes * [r314] Fix latex warning: it doesn't like double subscripts. 2004-03-26 warnes * [r306] Reflect movement of code from 'mva' package to 'stats' in R 1.9.0. 2004-03-25 warnes * [r296] - Estimable was reporting sqrt(X^2) rather than X^2 in the output. - Provide latex math markup for linear algebra expressions in help text. - Other clarifications in help text * [r295] Add enhancements to estimable() provided by S�ren H�jsgaard \email{sorenh@agrsci.dk}: I have made a modified version of the function [..] which 1) also works on geese and gee objects and 2) can test hypotheses af the forb L * beta = beta0 both as a single Wald test and row-wise for each row in L. 2003-11-17 warnes * [r221] - Fix incorrect handling of glm objects by fit.contrast, as reported by Ulrich Halekoh, Phd . - Add regression test code to for this bug. 2003-08-07 warnes * [r217] - Fixed incorrect denominator in standard error for mean in ci.default. 2003-04-22 warnes * [r190] - the variable 'df' was used within the lme code section overwriting the argument 'df'. 2003-03-12 warnes * [r173] - Fixed a typo in the example - Added to lme example 2003-03-07 warnes * [r168] - Minor changes to code to allow the package to be provided as an S-Plus chapter. 2003-01-30 warnes * [r160] - Renamed 'contrast.lm' to 'fit.contrast'. This new name is more descriptive and makes it easier to create and use methods for other classes, eg lme. - Enabled fit.contrast for lme object now that Doug Bates has provided the necessary support for contrasts in the nlme package. - New contrast.lm function which generates a 'depreciated' warning and calls fit.contrast - Updated help text to match changes. * [r158] - Removed argument 'correct' and now print separate corrected values for 2 x 2 tables. - Added arguments 'prop.r', 'prop.c' and 'prop.t' to toggle printing of row, col and table percentages. Default is TRUE. - Added argument 'fisher' to toggle fisher exact test. Default is FALSE. - Added McNemar test to statistics and argument 'mcnemar' to toggle test. Default is FALSE. - Added code to generate an invisible return list containing table counts, proportions and the results of the appropriate statistical tests. * [r157] - Added explicit check to ensure that the number of specified contrasts is less than or equal to the ncol - 1. Previously, this failed with an obtuse error message when the contrast matrix had row names, and silently dropped contrasts over ncol-1. 2002-11-04 warnes * [r142] - Moved fisher.test() to after table is printed, so that table is still printed in the event that fisher.test() results in errors. 2002-10-29 warnes * [r138] - Fixes to fast.svd to make it actually work. - Updates to man page to fix mistmatches between code and docs and to fix warnings. * [r137] - Moved make.contrasts to a separate file. - Enhanced make contrasts to better label contrast matrix, to give how.many a default value, and to coerce vectors into row matrixes. - Added help page for make.contrasts. - Added link from contrasts.lm seealso to make.contrasts. * [r136] Initial checkin for fast.prcomp() and fast.svd(). 2002-09-26 warnes * [r127] - Added note and example code to illustrate how to properly compute contrasts for the first factor in the model. 2002-09-24 warnes * [r124] - Fixed a typo. 2002-09-23 warnes * [r119] - Fixed syntax errors in barplot2.Rd and CrossTable.Rd - Fixed incorrect translation of 'F' (distribution) to 'FALSE' in glh.test.Rd * [r117] - Modified all files to include CVS Id and Log tags. * [r116] - Added CrossTable() and barplot2() code and docs contributed by Marc Schwartz. - Permit combinations() to be used when r>n provided repeat.allowed=TRUE - Bumped up version number 2002-08-01 warnes * [r114] - Corrected documentation mismatch for ci, ci.default. - Replaced all occurences of '_' for assignment with '<-'. - Replaced all occurences of 'T' or 'F' for 'TRUE' and 'FALSE' with the spelled out version. - Updaded version number and date. 2002-04-09 warneg * [r109] Checkin for version 0.5.3 2002-03-26 warneg * [r104] - Changed methods to include '...' to match the generic. - Updated for version 0.5.1 * [r99] Removed incorrect link to 'contrast' from seealso. 2002-02-20 warneg * [r81] Minor changes, typo and formatting fixes. 2002-01-17 warneg * [r70] - Fixed errror in last example by adding 'conf.int' parameter to 'estimable' call. * [r69] - Fixed typo in code that resulted in an syntax error. 2002-01-10 warneg * [r68] - print.glh.test() was using cat() to printing the call. This didn't work and generated an error. 2001-12-19 warneg * [r66] - Fixed display of formulae. - Added description of return value * [r65] - Removed extra element of return object. 2001-12-18 warneg * [r64] - Updated documentation to reflect change of parameters from 'alpha' to 'conf.int', including the new optional status of the confidence intervals. * [r63] - Modified to make confidence intervals optional. Changed 'alpha' parameter giving significance level to 'conf.int' giving confidence level. * [r62] - Added summary.glh.test to alias, usage, and example sections. * [r61] - Modified to work correctly when obj is of class 'aov' by specifying summary.lm instead of summary. This ensures that the summary object has the fields we need. - Moved detailed reporting of results from 'print' to 'summary' function and added a simpler report to 'print' * [r60] - Modified to work correctly when obj is of class 'aov' by specifying summary.lm instead of summary. This ensures that the summary object has the fields we need. * [r59] Initial checkin. 2001-12-17 warneg * [r56] - Fixed spelling errors. * [r55] - Fixed the link to contrasts.lm. - Rephrased title/description to be more clear. 2001-12-10 warneg * [r49] Renamed 'contrsts.coeff.Rd' to 'estimable.Rd' corresponding to function rename. * [r48] renamed from contrast.coeff.R to estimable.R (incorrectly via contrast.lm.R) 2001-12-07 warneg * [r37] - Added text noting that lme is now supported. * [r36] - Fixed typo: DF column was being filled in with p-value. * [r35] - Added ci.lme method to handle lme objects. 2001-10-16 warneg * [r27] Fixed unbalanced brace. 2001-08-25 warneg * [r12] - Added CVS header. - Added my email address. 2001-05-30 warneg * [r2] Initial revision