Logfile for ergm version 2.2-3

DATE: 2010-Jan-25 
FILE: many
CHANGES: Deleted headers from all .R, .Rd, .c, and .h files.  (These will be replaced later with updated versions)

DATE: 2010-Feb-2
FILE: src/ SAN.c, SAN.h, wtedgetree.c, wtedgetree.h
CHANGES: Replaced code accidentally deleted when deleting "headers" last time

DATE: 2010-Feb-2
FILE: R/ control.san.R, ergm.llik.miss.R, ergm.san.R
CHANGES: Replaced code accidentally deleted when deleting "headers" last time

DATE: 2010-Feb-2
FILE: R/is.inCH.R
CHANGES: Added function to see whether a point in R^d is inside the interior of
         the convex hull of a given set of points.  NB:  Uses Rglpk package

DATE: 2010-Feb-2
FILE: man/edgelist.ergm.Rd
CHANGES: Replaced code accidentally deleted when deleting "headers" last time

DATE: 2010-Feb-3
FILE: R/simulate.ergm.R
CHANGES: Major changes to both simulate.ergm and simulate.formula to reflect
         new "ergm.getMCMCsample" function

DATE: 2010-Feb-3
FILE: R/ergm.getMCMCsample.R
CHANGES: Major changes:  ergm.getMCMCsample is now only a wrapper to MCMC_wrapper
         and mcmcslave is now deprecated

DATE: 2010-Feb-4
FILE: R/ergm.R, ergm.estimate.R, ergm.mainfitloop.R, control.ergm.R
CHANGES: Several small changes that will eventually lead to implementation of
         the steplength algorithm Ruth has been working on 

DATE: 2010-Feb-4
FILE: DESCRIPTION, R/is.inCH.R 
CHANGES: Added Rglpk to "suggests" and added line so that Rglpk 
         (required by that function) is loaded when is.inCH is called

DATE: 2010-Feb-4
FILE: R/ergm.stepping.R
CHANGES: Added function implementing steplength algorithm

DATE: 2010-Feb-6
FILE: R/ergm.stocapprox.R, ergm.robmon.R, ergm.mainfitloop.R
CHANGES: Modified old calls to "ergm.getMCMCsample" in view of new updates

DATE: 2010-Feb-6
FILE: R/ergm.getMCMCsample.R
CHANGES: Added check of when the simulation has filled its allotment of
         memory for newedges (if so, then re-call function with 10x more memory)

DATE: 2010-Feb-6
FILE: R/simulate.ergm.R
CHANGES: Removed old check from here because it's now in ergm.getMCMCsample
         
DATE: 2010-Feb-6
FILE: R/ergm.getMCMCsample.R
CHANGES: Removed deprecated function ergm.mcmcslave after checking that no 
         existing function calls it (to retrieve the old function, see svn
         version 10805).

DATE: 2010-Feb-6
FILE: R/ergm.R
CHANGES: Restored defaults to original values

DATE: 2010-Feb-6
FILE: R/ergm.stocapprox.R, ergm.robmon.R, ergm.mainfitloop.R, ergm.phase12.R
CHANGES: Various small debugging tasks so that known examples work with new changes

DATE: 2010-Feb-6
FILE: R/getMCMCsample.R
CHANGES: Deleted lengthy verbose output used earlier for debugging

DATE: 2010-Feb-6
FILE: R/ergm.mple.R, ergm.pl.R
CHANGES: Made these versions look nearly the same as trunk versions, where a
         former bug had been fixed

DATE: 2010-Feb-6
FILE: R/simulate.ergm.R, control.simulate.R
CHANGES: fixed problems with simulate.ergm and control.simulate.ergm defaults

DATE: 2010-Feb-6
FILE: man/simulate.ergm.Rd, control.simulate.Rd
CHANGES: made doc files match changes to argument defaults

DATE: 2010-Feb-10
FILE: R/ergm.stepping.R
CHANGES: fixed bugs so this function can properly call ergm.estimate

DATE: 2010-Feb-10
FILE: R/ergm.estimate.R
CHANGES: Added comments, reorganized function, made subtle changes

DATE: 2010-Feb-10
FILE: R/ergm.degeneracy.R, ergm.llik.R, ergm.llik.miss.R
CHANGES: Changed "penalty" argument to "varweight"; changed ergm.llik function
         by adding comments

DATE: 2010-Feb-11
FILE: R/ergm.stepping.R
CHANGES: Added mle.lik to output produced; changed format of output

DATE: 2010-Feb-19
FILE: R/control.ergm.R, ergm.estimate.R
CHANGES: Changed default value of nr.maxit to 1000

DATE: 2010-Feb-19
FILE: R/ergm.llik.R
CHANGES: added "trustregion" argument to functions without one, since all are 
         assumed to have one.

DATE: 2010-Feb-19
FILE: R/ergm.stepping.R
CHANGES: multiple bug fixes

DATE: 2010-Feb-21
FILE: data/ecoli.RData
CHANGES: Added five different versions of the ecoli dataset (due to Salgado et 
         al, 2001) that is analyzed in Saul and Filkov, 2007

DATE: 2010-Feb-22
FILE: R/ergm.estimate.R, ergm.stepping.R
CHANGES: Eliminated experimental "xobs" argument of ergm.estimate

DATE: 2010-Feb-22
FILE: R/ergm.estimate.R, ergm.llik.R, ergm.llik.miss.R
CHANGES: brought these files up to date with trunk versions (and vice versa)

DATE: 2010-Feb-22
FILE: R/ergm.mainfitloop.R, ergm.robmon.R, ergm.stepping.R, ergm.stocapprox.R
CHANGES: Used "hessianflag" instead of "hessian" when calling ergm.estimate

DATE: 2010-Feb-22
FILE: R/ergm.llik.R, 
CHANGES: Added default values to all "trustregion" args; fixed bug in hessian
         by adding "drop=FALSE"

DATE: 2010-Feb-22
FILE: R/ergm.llik.miss.R, 
CHANGES: fixed bugs in hessian by adding "drop=FALSE"

DATE: 2010-Feb-22
FILE: R/ergm.R, control.ergm.R, man/simulate.ergm.Rd, man/control.ergm.Rd
CHANGES: fixed codoc agreement problems found by R CMD check

DATE: 2010-Feb-25
FILE: R/ergm.pl.R
CHANGES: Reverted to old code in one section that was not working

DATE: 2010-Feb-26
FILE: R/ergm.estimate.R, ergm.llik.R
CHANGES: Added closed-form lognormal approximation estimation capability;
         started rewriting Hessian code in ergm.llik.R

DATE: 2010-Feb-26
FILE: R/ergm.estimate.R
CHANGES: Added message outputting two distinct std error estimates (temporary)

DATE: 2010-Mar-8
FILE: R/ergm.estimate.R
CHANGES: Fixed error in closed-form lognormal calculation of Lout$value

DATE: 2010-Mar-8
FILE: R/ergm.stepping.R
CHANGES: Implemented several changes in preparation to submit article

DATE: 2010-Mar-9
FILE: src/edgetree.c, wtedgetree.c
CHANGES: Added PutRNGstate() function to match GetRNGstate().

DATE: 2010-Mar-16
FILE: R/ergm.estimate.R, ergm.stepping.R 
CHANGES: Implemented several small changes in preparation to submit article

DATE: 2010-Mar-16
FILE: R/ergm.llik.R
CHANGES: Fixed hessian functions, which had been returning negative hessian

DATE: 2010-Mar-16
FILE: data/ecoli.RData
CHANGES: Removed three of the five networks; now only ecoli1 and ecoli2 exist

DATE: 2010-Mar-16
FILE: man/ecoli.Rd
CHANGES: Began writing documentation for ecoli data.  Way not finished.

DATE: 2010-Mar-24
FILE: R/is.inCH.R, man/is.inCH.Rd
CHANGES: Changed arguments to R function, updated man page

DATE: 2010-Mar-24
FILE: R/ergm.R
CHANGES: Fixed code-documentation mismatch in order of arguments

DATE: 2010-Mar-26
FILE: R/InitMHP.R
CHANGES: Added a couple additional proposals from trunk version

DATE: 2010-Mar-26
FILE: R/ergm.MCMCse.lognormal.R, wtd.median.R, man/wtd.median.Rd
CHANGES: Added some trunk functions

DATE: 2010-Mar-26
FILE: DESCRIPTION
CHANGES: Added "Matrix" to list of suggested packages

DATE: 2010-Mar-26
FILE: man/ergm-internal.Rd
CHANGES: Added new R functions without documentation files to the list

DATE: 2010-Mar-28
FILE: man/ecoli.Rd
CHANGES: Updated documentation for ecoli datasets

DATE: 2010-Mar-29
FILE: R/ergm.mple.R
CHANGES: Removed "start" argument from glm call

DATE: 2010-Mar-30
FILE: R/ergm.estmate.R, ergm.stepping.R, ergm.llik.R
CHANGES: Fixed bugs in llik.fun2 and llik.gradient2; modified some of the syntax
         of the ergm.estimate function as well

DATE: 2010-Mar-30
FILE: R/ergm.mainfitloop.R, control.ergm.R, man/control.ergm.Rd
CHANGES: "Likelihood" metric is now called "lognormal"

DATE: 2010-Mar-31
FILE: src/edgetree.c
CHANGES: Added NetworkCopy function from trunk

DATE: 2010-Mar-31
FILE: R/ergm.utility.R
CHANGES: Deleted "safeupdate.formula" function, added "ergm.update.formula" and 
         "term.list.formula"

DATE: 2010-Mar-31
FILE: R/network.update.R, man/network.update.Rd
CHANGES: Added "output" argument to function

DATE: 2010-Apr-4
FILE: src/MHproposals.c
CHANGES: Fixed rarely occurring bug in Hastings ratio calculation

DATE: 2010-Apr-4
FILE: tests/ALLTERMTESTS
CHANGES: Removed this file; it's not needed

DATE: 2010-Apr-4
FILE: R/ergmMPLE.R, man/ergmMPLE.Rd, tests/ergmMPLEtest.R
CHANGES: Added a new function for returning change statistic matrix, used for
         calculating MPLE.  Also added test and documentation files.

DATE: 2010-Apr-6
FILE: src/MHproposals.c, MHproposals.h, MHproposals_bipartite.c
CHANGES: Fixed rarely occurring bug in Hastings ratio calculation; renamed a few
         proposal functions

DATE: 2010-Apr-9
FILE: R/ergm.san.R, control.san.R, man/san.Rd, control.san.Rd
CHANGES: Major changes to san.ergm; deleted control.san.ergm; added docs for san
         and control.san

DATE: 2010-Apr-19
FILE: man/control.ergm.Rd, ergm.Rd
CHANGES: Eliminated Boer et al reference, replaced with Snijders (2002)

DATE: 2010-Apr-21
FILE: man/san.Rd, control.san.Rd
CHANGES: Fixed codoc mismatches to get ready to upload to CRAN


