2020-01-12 Package version 1.2-23 submitted to CRAN.
* Changes in RSiena and RSienaTest:
Option projname=NULL in sienaAlgorithmCreate.
This option is used for all examples where siena07 is called,
to restrict file writing for examples to the temporary directory.
2020-01-10 Package version 1.2-22.
* Changes in RSiena and RSienaTest:
Changed extension names of output files from .out to .txt.
2020-01-04 R-Forge Revision 345, package version 1.2-22.
Changes in RSiena and RSienaTest:
* Tested effects reported in print of Multipar.RSiena and score.Test.
* Reordered effects object so that reciAct comes after instead of
before inAct and inAct.c.
* Operation of siena07 in Test 14 silenced (message Brian Ripley about
apparently minor changes depending on the platform).
2019-12-18 packages version 1.2-21.
Changes in RSiena and RSienaTest:
* Tests ended with deleting file Siena.out by unlink instead of file.remove.
2019-12-16 packages version 1.2-20.
Changes in RSiena:
* dumpChain (Chain) and getChainProbabilities (siena07models) dropped;
kept in RSienaTest.
* Configure line 2274 changed to RBIN="${R_HOME}/bin"
Changes in RSiena and RSienaTest:
* Tests ended with deleting file Siena.out.
2019-12-16 R-Forge Revision 344, package version 1.2-19.
Changes in RSiena and RSienaTest:
* SienaTimeTest: some corrections with consequences probably only if
one effect is being tested; or is left after automatic exclusion
of tested effects.
* Setting-related effects corrected.
* Some further changes in Settings model.
* New effects avGroupEgoX, outIso.
* CovariateDependentNetworkEffect: + lSimulatedOffset taken out of
definition missings for ChangingCovariate (seems wrong).
* Alphabetically ordered sources.list.
Changes in RSienaTest:
* In sienaBayes: more precise check that
prevAns has same specification as effects;
priorSigEta added to sienaBayesFit object.
dimnames(ThinParameters)[[3]] defined as effect names;
ridge for prior variance of rate parameters if priorRatesFromData=1 or 2
decreased from 0.5 to 0.01.
* sienaBayes: zm and zsmall are removed at the end.
* plotPostMeansMDS also shows a title, and coordinates show the dimensions
of the MDS solution.
* The value returned by plotPostMeansMDS contains not only the
plot coordinates, but also the similarities (correlations)
between the groups.
* glueBayes: added z$varyingInEstimated, z$varyingNonRateInEstimated,
z$ratesInVarying to the object produced.
(Else extract.sienaBayes is not going to work properly.)
* glueBayes: if one of the sampling parameters is different,
give a warning instead of a stop.
2019-10-16 R-Forge Revision 341, package version 1.2-18.
In this R-forge revision, only RSienaTest was updated.
Changes in RSiena and RSienaTest:
* Continuous dependent behavior variables implemented (Nynke Niezink).
This implies new effect types continuousFeedback, continuousIntercept,
continuousOneModeObjective, continuousRate, continuousWiener,
unspecifiedContinuousInteraction.
* imputationValues allowed in sienaDependent (Nynke Niezink)
* New effect outMore.
* component startingDate added to sienaFit object; this is reported in
siena.table(..., type='tex', ...).
* Speeded up calculation of IndegreeDistribution and OutdegreeDistribution
for sienaGOF if there are no missings or structurals.
* regrCoef and regrCor added to the sienaFit object also when !dolby.
* Some "warnings" changed back to "cat".
* EpochSimulation->totalRate renamed to grandTotalRate,
to avoid confusion with DependentVariable->totalRate.
* stop if useCluster and returnChains (in this case, no chains would
be returned anyway).
* sienaDataCreate: more informative message in case of constraints.
* Further explanation in help page for setEffect, and small extensions of
help pages for getEffects, includeEffects, and includeInteraction.
* small clarification in help page for sienaDependent.
* small clarifications about node sets in help pages for coCovar,
varCovar, coDyadCovar, varDyadCovar, and sienaDataCreate.
* small addition to help page for sienaTimeTest.
* Object names are given in sienaFit.print if simOnly.
* Settings model: corrected scores for rate parameters;
stepType in NetworkCache; new class settingNetworkEffects;
effect group nonSymmetricSymmetricObjective split in this and
nonSymmetricSymmetricSObjective (also operating for primary setting effects);
new effects settingSizeAct, settingSizeActSqrt, settingSizeActLog,
settingOppAct, settingOppActSqrt, settingOppActLog,
settingLogCreationAct, settingOppActD, settingOppActSqrtD,
settingOppActLogD, settingLogCreationActD.
These new effects are not yet operational (target statistics not calculated).
* inPopIntn and outActIntn dropped from effect group
nonSymmetricSymmetricObjective.
Changes in RSienaTest:
* Corrected error in names of array returned by extract.posteriorMeans.
* New parameter excludeRates in extract.posteriorMeans, plotPostMeansMDS.
* Use parameter pmonly also in plotPostMeansMDS.
2019-05-20 R-Forge Revision 340, packages version 1.2-17.
Changes in RSiena and RSienaTest:
* New effects outAct.c, inAct.c, outPop.c., inPop.c, degPlus.c.
* Effects antiInIso and antiInIso2 implemented for non-directed networks.
* Effects outTrunc, outTrunc2, outSqInv, isolateNet, and degPlus got
endowment effects.
* For inPopIntn, outPopIntn, inActIntn, and outActIntn, added "centered"
to the statistic name.
* In plot.sienaGOF new parameter fontsize; ... will also accept parameters
cex, cex.main, cex.lab, cex.axis.
* estMeans.sem added to object produced by siena07, if x$simOnly.
* Components regrCoef, regrCor, estMeans, and estMeans.sem mentioned in
help page for siena07.
* Hidden options in siena07 with x$targets (initializeFRAN.r).
* Extra identifying information for "total probability non-positive" error
(NetworkVariable.cpp).
* On various places replaced "cat" by "message" or "warning".
* Replaced && by any(.. & ..)
to comply with R 3.6.0 new standard (sienaTimeTest).
Changes in RSienaTest:
* sienaBayes: do not allow prevAns and prevBayes objects simultaneously;
omit partial averaging at the end of warming phase;
"prevAns" now is allowed to have a different specification of
random effects than "effects";
warming up lowest level reduced to 3 iterations of MCMCCycle,
but applied also if newProposalFromPrev.
2019-02-26 R-Forge Revision 339, packages version 1.2-16.
Changes in RSiena and RSienaTest:
* New effects maxXAlt, minXAlt (thanks to Per Block, Marion Hoffman,
Isabel Raabe, and Timon Elmer), outIsolate.
* Effect name of isolate effect (not shortName) changed to in-isolate.
* New parameter thetaValues in siena07, allowing simulations in Phase 3
with varying parameters according to matrix input.
Corresponding changes in print.sienaFit.
* sienaRI: earlier check for bipartite dependent variables; if so, stop.
* Argument "verbose" added to includeEffects and setEffect.
(It existed already for includeInteraction.)
* Wald.RSiena, Multipar.RSiena and score.Test also produce one-sided
test statistic if df=1.
* objects produced by Wald.RSiena, Multipar.RSiena and score.Test
have class "sienaTest".
* print method added for objects of class sienaTest.
* Changed/added some keywords and concepts in .Rd files.
Changes in RSienaTest:
* Allow arbitrary number of interactions in sienaBayes (via getEffects).
* simpleBayesTest and multipleBayesTest corrected; they were wrong
for models with user-defined interactions.
* glueBayes extended (to allow good use for models with fixed parameters)
and check for prior rates relaxed (to allow priorRatesFromData=2
leading to different prior rates.)
* New default (prevBayes$nwarm >= 1) for newProposalFromPrev in sienaBayes.
* siena.table for sienaBayes results extended with between-groups s.d.
* Argument "verbose" added to extract.posteriorMeans.
* sienaBayes: check that prevAns is a sienaFit object;
totally skip initial multigroup estimation if prevAns is given with
prevAns$n3 >= 500 and (!usePrevOnly).
2019-01-02 R-Forge Revision 338, packages version 1.2-15.
Changes in RSiena and RSienaTest:
* New effects sameXReciAct and diffXReciAct.
* Increased length of MLSimulation::preburnin.
* Changes in sienaFitThetaTable, averageTheta.last, sdTheta.last in
sienaprint.r to enable change in extract.posteriorMeans (see below).
* siena.table adapted for sienaBayes results.
Changes in RSienaTest:
* New functions shortBayesResults and plotPostMeansMDS.
* extract.posteriorMeans works also for an object saved as PartialBayesResult
(which has NA results for as yet unfinished runs)
(this was achieved by changes in sienaprint.r, see above).
* Examples added to help page for print.sienaBayesFit (in dontrun...).
2018-12-05 R-Forge Revision 337, packages version 1.2-14.
Changes in RSiena and RSienaTest:
* Check for length of parameter mult for ML estimation (initializeFRAN),
extend explanation of mult in help page for sienaAlgorithmCreate.
Changes in RSienaTest:
* extract.sienaBayes corrected.
* sienaBayes: subphases for initial global parameter estimation decreased
to 2;
prior variances for fixed parameters introduced;
after initialization by MoM estimation of multi-group data set,
a precision weighted mean (called "Kelley") of this estimate and the
prior mean is used for initial parameter values;
prewarming phase introduced before improveMH;
in case prevBayes is used, also parameter nImproveMH in the function
call of sienaBayes supersedes this parameter in the prevBayes object;
checks for dimensions of prior mean and covariance matrix
put earlier in the initialization phase.
2018-10-29 R-Forge Revision 336, packages version 1.2-13.
Changes in RSiena and RSienaTest:
* Correct error in print01Report that occurred for changing dyadic covariates
given as lists of sparse matrices.
* Correct error in sienaGroupCreate for non-centered actor covariates.
* Drop dependence on utils (the packages did not depend on it).
* Also get simulated dependent behavior variables for
siena07(,,, returnDeps=TRUE, ...) for ML estimation.
* siena.table corrected for data sets with several dependent variables.
* For siena08: new parameters which and useBound in plot.sienaMeta;
new parameter reportEstimates in siena08, allowing to reduce
the output produced.
* updateSpecification: also update randomEffects column.
* Change help page for sienaDependent.
* Some changes in sienaprint.r depending on sienaBayes, to keep the two
packages the same when possible.
Changes in RSienaTest:
* sienaBayes: new parameters proposalFromPrev, which allows
taking proposal distributions from prevBayes object;
incidentalBasicRates resuscitated;
allow fixed rate parameters, with priorRatesFromData=-1;
changed initial values of scale factors for proposal distributions;
new parameters target and usePrevOnly;
in case prevBayes is used, parameters nrunMHBatches, nSampVarying,
nSampConst, and nSampRates in the function call of sienaBayes supersede
those in the prevBayes object;
correct bug for three-effect interactions (projectEffects)
copy parameters modelType, behModelType, MaxDegree, Offset, initML,
from parameter algo to the algorithms created within sienaBayes;
more extensive checking of smallest eigenvalue of covariance matrix
(function correctMatrix);
For priorRatesFromData = 1 or 2, the resulting matrix for priorKappa != 1
was incorrect; this was corrected.
reported timing changed to elapsed system time;
some reordering of parameters.
This also required some corresponding changes in sienaprint.r.
* DESCRIPTION: Added 'Encoding: UTF-8' because sienaGOF.r complained about
non-ASCII.
* Data.cpp: Add a new 'SimNetworkData' type: This is for variables that
behave like dependent variable for effects but are not changed by the
simulation. Currently used for the primary setting network.
* PrimaryLayer: A network layer holding the full primary setting as network.
2018-05-13 CRAN for RSiena, version 1.2-12.
2018-05-07 R-Forge Revision 336, packages version 1.2-12.
Changes in RSiena and RSienaTest:
* Change default parameter of reciAct to 1 (implies changes only in
output, not computationally).
* Internal effect parameter carried through in sienaTimeFix.
Changes in RSiena:
* Shorter examples for sienaGOF and sienaGOF-auxiliary help pages.
2018-05-06 R-Forge Revision 335, packages version 1.2-11.
Changes in RSiena and RSienaTest:
* New effects gwdspFF and gwdspFB.
* Corrected effects simEgoInDist2 and simEgoInDist2W for two-mode networks.
Perhaps simEgoDist2 and simEgoDist2W were broken in a previous version;
if so, that was now corrected.
* Added sqrt version for reciAct, obtained for parameter = 2.
* Allowed the value parameter=NULL for setEffect and includeInteraction,
meaning that no change is made for the internal effect parameter.
For setEffect this is the new default, implying that when starting the
default values from allEffects.csv are used, just like in includeEffects
(where no parameter can be given).
* New auxiliary functions for sienaGOF: Triad Census and mixedTriadCensus
(contribution by Christoph Stadtfeld).
* triad census from igraph added to help page of sienaGOF-auxiliary.
* sienatable output for type="html": added rules=none, frame = void to
general options; changed minus to – ;
added column for asterisks to have better alignment for estimates.
* Stop cluster also if (!z$OK || !z$Phase3Interrupt) (robmon.r).
* Extension of help page for siena07 by mentioning functions for accessing
simulated networks for ML.
Changes in RSienaTest:
* extract.sienaBayes: error corrected that occurred if called with
extracted="all" but there are no varying, or no non-varying parameters.
2018-03-24 R-Forge Revision 334, packages version 1.2-10.
Changes in RSiena and RSienaTest:
* Example in help page siena07 for accessing generated networks for ML.
2018-03-21 R-Forge Revision 332, packages version 1.2-9.
Changes in RSiena and RSienaTest:
* model/effects/generic: Remove const from some of the value methods (as
it was in rev316).
* new effect group covarBehaviorSymmetricRate (effects.r,
effectsDocumentation.r).
* Added effects avExposure, totExposure, infectDeg,
susceptAvCovar, infectCovar for symmetric networks.
* New effects degAbsDiffX, degPosDiffX, degNegDiffX, degAbsContrX,
degPosContrX, degNegContrX,
* New effect group dyadBipartiteObjective (effects.r).
* Effects XWX, XWX1, and XWX2 enabled for bipartite networks.
* Added parameter=2 for FFDeg, BBDegm FBDeg, FRDeg, BRDeg.
* Corrected altInDist2, totInDist2 for two-mode networks.
* Dropped some duplications in effectsDocumentation.
* Modification of effect outTrunc2 (TruncatedOutdegreeEffect2).
Perhaps this was superfluous.
* Export of last simulated state from ML simulations if
returnDeps: new function gotoLastState() in MLSimulation, adaptation of
mlPeriod() in siena07models.cpp, and some other small changes.
* Added endowment and creation effects for maxAlt and minAlt.
* Error messages for non-character nodeSets in sienaDependent, coCovar,
varCovar.
* Improved way of handling missings for distance-2 network effects.
* Added components requestedEffects, theta, and se to siena08
(theta and se are the ML estimates).
* Error in summary of sienaMeta object corrected:
for the Estimated mean parameter the two-sided p was announced but a
one-sided p was given. Also the ML results under normality assumptions
were copied from print.sienaMeta to print.summary.sienaMeta.
* Drop transTriads<-TRUE in effects.r, so that initial model
for non-directed networks contains only basic rates and degree effect.
* Added reference to DESCRIPTION and citation info in \inst.
* The dropping of the exclusion in effects.r of effects if the variance of a
covariate is 0, or a covariate has only two values, of version 1.1-306,
is taken further (was incomplete).
* Message if attributes higher, disjoint, or atLeastOne are TRUE
at the end of sienaDataCreate.
* Incomplete changes toward export of observed information matrix for MoM
(derivative of score function): extensions to getStatistics in
siena07internals, implying also changes in siena07models.cpp,
Networkvariable.cpp, BehaviorVariable.cpp.
* Added altInDist2 test to parallel.R (test12, test13).
Changes in RSiena:
* siena07internals.cpp, siena07model.cpp, and siena07setup.cpp brought
in line with RSienaTest.
Changes in RSienaTest:
* new function extract.posteriorMeans for sienaBayes results.
* Add z$fix to sienaBayes object, and use !z$fix in check of maximum
estimated parameter value after initialization.
* Correct construction of groupwise effects object in sienaBayes so that this
will work also when evaluation effects for density and/or reciprocity are
not included; and when there are interaction effects for which the main
effects are not included.
* Corrections for print and summary of sienaBayes objects.
In print.sienaBayesFit:
include fixed parameters and give credibility intervals
for rate parameters;
also give posterior means and credibility intervals for standard deviations
of randomly varying parameters;
allow shorter ThinParameters;
print objects returned through partialBayesResult.RData.
* multipleBayesTest corrected (there was an error for testing 2 or more
linear combinations simultaneously) and adapted for cases with fixed
parameters;
adapted help page text.
* All remaining parts of rsiena01gui removed.
* As a compensation of this, sienaDataCreateFromSession exported,
with a more informative help page.
2017-09-18 R-Forge Revision 331
Changes in RSienaTest:
* simplify/harden configure.ac
Compiling examples and trying to load them should prove more robust against
the ongoing changes in the R-Forge build setup.
2017-09-18 R-Forge Revisions 328-330
Changes in RSienaTest:
* Added \R\formula.R and \tests\formula.R.
2017-09-18 R-Forge Revision 325
Changes in RSienaTest:
* Added \doc\tutorial-effect.tex
2017-09-18 R-Forge Revision 323
Changes in RSienaTest:
* StatisticCalculator.cpp: delete Network clones not owned by State.
2017-09-18 R-Forge Revision 322
Changes in RSiena:
* StatisticCalculator.cpp: delete Network clones not owned by State.
2017-09-10 R-Forge Revision 321
Changes in RSiena:
* Merge RSienaTest changes r319-r320. Renamed init.c to init.cpp.
2017-09-10 R-Forge Revision 320
Changes in RSienaTest:
* siena07*.h: Update method declarations.
* getTargets.R: getTheTargets renamed to getTargets to match with RSiena.
* R/algorithms.r: Uncomment doCreateChains (C side is missing). See change
log for "2017-09-03 version 1.2-1" where it is mentioned as dropped.
* utils/Utils.cpp: Make log call unambiguous.
* R/*, man/*: Whitespace changes to ease the merging.
2017-09-10 R-Forge Revision 319
Changes in RSienaTest:
* init.cpp: Register native routines.
* *.R: Apdapt .Call statements.
2017-09-09 R-Forge Revision 318, packages version 1.2-4.
Changes in RSiena and RSienaTest:
* Longer Description field in DESCRIPTION, as per CRAN suggestions.
* Correction of print.sienaFit, repairing the option include=FALSE
in setEffect and includeEffects.
* Included test of includeEffects in parallel.r.
Changes in RSienaTest:
* Extra line in example for sienaBayes.Rd.
2017-09-09 R-Forge Revision 317
Changes in RSienaTest:
* Effects:
Move 'using namespace std' to cpp files.
Make value() consistently 'const'.
Use std::abs for integers.
Remove unused fields.
* UniversalSetting.cpp: Do not hide GeneralSetting::terminateSetting().
* siena07models.cpp: Reduce state complexity (rchk test).
* siena07*.cpp: Fix unPROTECTed vars (rchk test).
2017-09-08 R-Forge Revision 316, cran package version 1.2-3
Changes in RSiena and RSienaTest:
* totAlt also for non-directed networks
* Added cran-comments.md, .Rbuildignore for comments about cran submission.
2017-09-07 R-Forge Revision 315
Changes in RSiena:
* Merge RSienaTest changes r311-r314.
2017-09-07 R-Forge Revision 314
Changes in RSienaTest:
* StatisticCalculator.cpp: Fix memory leak in deconstructor.
* Move some of the 'using namespace' from header to cpp files.
* Lots of whitespace changes to ease the merging process.
2017-09-06 R-Forge Revision 313
Changes in RSienaTest:
* DependentVariable.cpp: Fix null pointer in accumulateRateScores when
called on BehaviorVariable.
2017-09-04 R-Forge Revision 312
Changes in RSienaTest:
* configure.win: adapt windows configuration script
2017-09-04 R-Forge Revision 311
Changes in RSienaTest:
* configure.ac: add openmp test for macOS El Capitan compatibility
2017-09-03 R-Forge Revision 310
Changes in RSiena:
* effects.r: remove hard coded number of types of interactions
* NetworkDependentBehaviorEffect.cpp: fix uninitialized network pointer
2017-09-03 cran proposal package version 1.2-1 (not accepted)
Changes in RSiena and RSienaTest:
* Score test and sienaGOF corrected for the case that some parameters are
fixed and not tested.
* New function score.Test, and its output added to print.sienaFit
(if any parameters fit the bill).
* Argument varName added to updateTheta.
* In sienaAlgorithmCreate, for parameters Offset, MaxDegree, modelType,
and behModelType, require that these are named vectors with the names
of the dependent variables, or NULL; this is checked in initializeFRAN.
* For networkModelType 3 (Initiative, AAGREE), an offset is added to the
confirmation model; this is taken from Offset in the algorithm object.
* For effects inPopX, outPopX, inActX, outActX, sameXInPop, sameXOutAct,
diffXInPop, diffXOutAct, homXInPop, homXOutAct, altXOutAct, diffXOutAct,
changed interaction2 to '' (was '1')
and default parameter to 1 (AllEffects.csv).
* New effects sameXInPopIntn, sameXOutPopIntn, sameXInActIntn,
sameXOutActIntn,
* Additional parameter dropRates in print.sienaEffects (useful for
sienaBayes with many groups).
* In print.sienaEffects, omit last remark about random effects if only
one line is printed.
* Corrected use of modeltype (initializeFRAN.r, CInterface.r),
* Change in print.sienaAlgorithm for modelType.
* In CalculateDerivative (phase1.r) changed assignment for
diag(dfra)[a][b] (which is wrong) to diag(dfra)[which(a)[b]]
(avoids crash of sienaBayes in rare cases).
* Added stopCluster to robmon.r.
* Took out verbose reporting for sienaGOF with !is.null(cluster).
* Added an example for multiple processes in sienaGOF.Rd.
* Took out 'dontrun' for the example in sienaGOF.Rd and the first example
in sienaTimeTest.Rd.
* Reduced running time of some of the algorithms used in examples.
* LF line endings for files configure, configure.ac, cleanup, also in
Tortoise properties.
CRLF line endings for configure.win, cleanup.win, also in Tortoise
properties.
* To avoid confusion with the C++ function, renamed getActorStatistics() in
getActorStatistics.R to getTheActorStatistics, and renamed getTargets()
in getTargets.R to getTheTargets.
For similarity between RSiena and RSienaTest, added getActorStatistics.R
to RSienaTest, even if it is not used.
* Completed heading for getTargets() in siena07setup.h,
and used these additional parameters in getTheTargets.
Changes in RSienaTest:
* Error fixed in sienaBayes for definition of z$rateParameterPosition,
which could lead to errors in the case of more than one dependent
variable with an interaction effect specified for some other than
the last dependent variable.
* Parameter UniversalOffset of sienaAlgorithmCreate renamed to Offset.
* Added posterior variances to print.sienaBayesFit.
* Added stopCluster to sienaBayes.r.
* Dropped doCreateChains from algorithms.r.
* Terminated open connection at the end of sienacpp (CInterface.r).
Changes in RSiena:
* New parameter OffSet in sienaAlgorithmCreate.
* Registration of native routines (init.c,
using tools::package_native_routine_registration_skeleton,
\fixes in NAMESPACE, changed specifications of arguments of .Call,
and with help from Brian and Ruth Ripley).
2017-09-01 R-Forge Revision 309
Changes in RSienaTest:
* configure.ac: don't compile with openmpi if orted is missing
2017-08-30 R-Forge Revision 308
Changes in RSienaTest:
* effects.r: remove hard coded number of types of interactions
* Model.cpp: allow gmm type interaction effects
* NetworkDependentBehaviorEffect.cpp: fix uninitialized network pointer
Some forking here:
RSienaTest and RSiena went their own ways, reunited in 1.2-2.
2017-05-12 R-Forge Revision 307
Changes in RSiena and RSienaTest:
* Operation of option 'absorb' (behModelType=2) corrected.
Changes in RSienaTest:
* BothDegreesEffect.h and .cpp added to repository.
2017-05-10 R-Forge Revision 306
Changes in RSiena and RSienaTest:
* Added updateSpecification (effectsMethods)
* New effects inPopX, outPopX, inActX, outActX, sameWXClosure, degPlus,
absDiffX, avAltPop, totAltPop, egoPlusAltX, egoPlusAltSqX,
egoRThresholdX, egoLThresholdX, altRThresholdX, altLThresholdX.
* New effect group covarABipNetObjective.
* outActIntn added to effect group nonSymmetricSymmetricObjective.
* outOutAss dropped for symmetric networks; only outInAss remains.
* egoX effect has interactionType='ego' also for symmetric networks.
* The exclusion in effects.r of effects if the variance of a covariate is 0,
or a covariate has only two values, is dropped (these effects have no
meaning, but their exclusion was a potential nuisance for meta-analyses).
* Description of gwespFB and gwespBF corrected (allEffects.csv and manual).
* includeInteraction has an additional parameter random.
* siena08 now also accepts a list for ...
* somewhat extended help page for siena08.
* Longer example in sienaTimeTest.Rd.
* Indication of effect parameters dropped in names for altInDist2 and
totInDist2 (they have no effect parameters).
* ModelType now is specific to the dependent variable:
given as a named integer vector in sienaAlgorithmCreate, defined in
Dependentvariable.h and NetworkVariable.h as networkModelType.
* ModelTypeStrings now is a function instead of a string vector.
* behModelType added
(sienaAlgorithmCreate, siena07setup.cpp, BehaviorLongitudinalData,
BehaviorVariable),
and enum type BehaviorModelType specified in BehaviorVariable.cpp.
* New option 'absorb' for behModelType=2.
* sienaAlgorithmCreate, siena07: new option lessMem, reducing storage in
siena07 by leaving out z$ssc and z$sf2; but these are used by
sienaTimeTest and sienaGOF, so running those functions will be impossible
for sienaFit object obtained with lessMem=TRUE.
* modified print.sienaAlgorithm: added behModelType, n2start.
* small cosmetic change print.siena.
* Modified check for singular covariance matrix (phase3.2 in phase3.r).
* Warning if includeEffects is used with parameter random.
* Warning for impossible or zero changes if maximum likelihood.
* siena07: z$phase1devs, z$phase1dfra, z$phase1sdf, z$phase1sdf2,
z$phase1scores, z$phase1accepts, z$phase1rejects, z$phase1aborts,
z$sfl, z$msfinv, z$sf.invcov dropped from object produced.
* Checks for named vectors in initializeFRAN made more systematic.
* Check of effects object in mlInit (maxlike.r) does not use
is.data.frame any more.
Changes in RSiena:
* test13 dropped from parallel.R (using clusters undesirable
for basic testing).
Changes in RSienaTest:
* sienaBayes:
various changes to save memory (thanks to Ruth Ripley);
lessMem=TRUE in operation of siena07;
z$candidates removed, place of storing z$acceptances changed;
improved report of groups with no changes;
warning for impossible changes;
if priorRatesFromData=2, change to different robust covariance matrix
estimator when this is necessary (i.e., for small number of groups);
in print.summary, also report nImproveMH;
a few lines added to help page.
* test14 dropped from parallel.R (using clusters undesirable for basic testing)
and replaced by a test using maxlike.
2016-10-20 R-Forge Revision 305
Changes in RSienaTest:
* BehaviorEffect, NetworkDependentBehaviorEffect, StatisticCalculator:
initialize() with simulated state for GMM.
* Other effects: Fix [-Wreorder] warings.
2016-10-14 R-Forge Revision 304
Changes in RSiena and RSienaTest (George Vega Yon):
* tests/parallel.R: Changing makeForkCluster to makeCluster.
* man/siena07.Rd: Adding a description of the -cl- option + examples.
2016-10-13 R-Forge Revision 303
Changes in RSiena and RSienaTest (George Vega Yon):
* R/robmon.r: New argument -cl- allows users passing their own cluster,
creating a new cluster only if length(cl) == 0.
* R/siena07.r: Same as in robmon. Now, if the user passes -cl-, the cluster
is closed only if length(cl) == 0.
* man/siena07.Rd: Including the new -cl- option in the manual.
* tests/parallel.R: Adding new test using the new -cl- option.
2016-10-10 R-Forge Revision 302
Changes in RSiena and RSienaTest:
* Capitalization error corrected in CovariateDegreeFunction.cpp.
2016-10-09 R-Forge Revision 301
Changes in RSiena and RSienaTest:
* Warning if includeEffects is used with parameter initialValue - corrected.
* New effect: sameXCycle4, homCovNetNet, contrastCovNetNet, covNetNetIn,
homCovNetNetIn, contrastCovNetNetIn,
inPopIntnX, inActIntnX, outPopIntnX, outActIntnX
* Changed type of lcounters and counters in FourCyclesEffect to long int,
permitting the 4-cycles effects for larger and denser networks.
* Dropped cl.XWX effect from two-mode - one-mode coevolution
(did not belong).
* egoSqX is an ego effect (CovariateEgoSquared.cpp and .h).
* Added cycle4 for one-mode networks.
* Added outAct, outInAss for symmetric networks.
* SienaRI: Structural zeros and ones are excluded from the calculations;
added option getChangeStatistics;
row names given to matrices that have rows corresponding to effects;
adapted so that it runs for models with only 1 parameter;
adapted so that for a bipartite dependent variable it does not crash.
* Warning if includeEffects is used with parameter 'parameter'.
* Small additions to print.sienaAlgorithm.
* Clearer output for MaxDegree in print.sienaFit.
* Indication '^(1/2)' for outInAss for 2-mode networks changed to '^(1/#)',
where # is the effect parameter (allEffects.csv).
Changes RSienaTest:
* simulatedOffset changed from int to double (NetworkLongitudinalData)
2016-09-12 R-Forge Revision 300
Changes RSienaTest:
* doc/RSienaDeveloper.tex: Build system documentation.
2016-08-30 R-Forge Revision 299
Changes RSienaTest:
* network/Simmelian: Cleaned dependencies.
2016-08-28 R-Forge Revision 297
Changes in RSiena:
* buildsystem: added Makefile and src/sources.list
2016-08-17 R-Forge Revision 296
Changes in RSiena and RSienaTest:
* Warning if includeEffects is used with parameter initialValue - did not work.
* Warning if includeInteraction is used for more interactions
than available given parameters nintn and behNintn.
* Deleted session parameter from print01Report.
* Corrected cycle4 effect for parameter=2 (sqrt version).
* Additional auxiliary function CliqueCensus in help page sienaGOF-auxiliary.
* Error corrected in DyadicCovariateAvAltEffect.cpp.
2016-07-23 R-Forge Revision 295
Changes in RSienaTest (Felix Schoenenberger):
* buildsystem: added Makefile and src/sources.list
unitTests: removed sink('/dev/null')
2016-05-28 R-Forge Revision 294
Changes in RSiena and RSienaTest:
* In SimilarityEffect.cpp and SimilarityWEffect.cpp, changed
abs to std::abs to avoid ambiguity, dropping #include .
Changes in RSienaTest:
* Mac line endings given to configure file.
2016-05-23 R-Forge Revision 293
Changes in RSiena and RSienaTest:
* Removed manual and bibliography from installation;
still available in \RSienaTest\doc.
* Added functions initializeStatisticCalculation() and
cleanupStatisticCalculation() to class BehaviorEffect.
* New effects totAltEgoX, totAltAltX, egoSqX, diffX, diffSqX, egoDiffX,
avAltW, totAltW, avSimW, totSimW, jumpFrom, jumpSharedIn, mixedInXW, mixedInWX,
avWalt, totWAlt.
* New effect class DyadicCovariateDependentBehaviorEffect corresponding
to group dyadBehaviorNetObjective (effects.r, effectsDocumentation.r).
* added endowment and creation effects for inAct, inActSqrt, outAct, outActSqrt.
* inActIntn also implemented for two-mode dependent networks.
* New argument 'matrices' in print.summary.sienaFit.
* Changes to NAMESPACE and DESCRIPTION files to satisfy R3.3.0 for
external functions from base packages.
Changes in RSiena:
* siena01Gui and associated functions dropped.
* sienaDataCreateFromSession dropped.
Changes in RSienaTest:.
* Change in endowment effect estimation for avAlt effect.
* New effects simmelian, simmelianAltX, avSimmelianAlt, totSimmelianAlt.
2016-02-22 R-Forge Revision 292
Changes in RSienaTest:
* Fixed permission filter for 2-mode networks (NetworkVariable.cpp).
2016-02-03 R-Forge Revision 291
Changes in RSienaTest:
* Fixed include order.
* Removed `using namespace` from all headers.
2016-01-31 R-Forge Revision 290
Changes in RSiena and RSienaTest:
* New effects FBDeg, FRDeg, BRDeg (RFDeg was mentioned earlier
but was not implemented; its place is now taken by FRDeg),
gwespFFMix, gwespBBMix, gwespBFMix, gwespFBMix, gwespRRMix,
gwespMix.
* Corrected check for positive derivative matrix at the end of phase 1
omitted for effects with fixed parameters
(phase1.r, CalculateDerivative(); phase3.r).
* Added parameters fix, test, parameter to includeInteraction().
* More helpful error message for incorrect nodesets in sienaDataCreate;
extended help pages sienaDataCreate.Rd and sienaDependent.Rd.
* Correction of diag(diag(..)) to allow estimation with p=1
(phase1.r, initializeFRAN.r)
* fromBayes bug corrected.
Changes in RSienaTest:
* sienacpp added (programmed by Felix Schoenenberger); this includes
gmm estimation (Amati - Snijders);
and implies a lot of changes in other functions to enable this.
* settings model added (Snijders - Preciado; programmed by Mark Ortmann);
not yet documented. Also implies a lot of changes.
* sienaBayes(): option initML added; check for zero distances.
corrected function improveMH() in initialization;
check for dimensions priorMu and priorSigma done earlier.
* print.multipleBayesTest(): option descriptives added.
* glueBayes(): add p1 and p2 to created object.
2015-09-10 R-Forge Revision 289
Changes in RSiena and RSienaTest:
* New defaults for siena07() in sienaAlgorithmCreate():
doubleAveraging=0, diagonalize=0.2 (for MoM).
* Improved one-step approximations to expected Mahalanobis
distances in sienaGOF() (control variates for score function).
* Permit 3-way interactions with one ego and two dyadic effects
(initializeFRAN.r) (this was erroneously not allowed).
* New effects Jin, Jout, JinMix, JoutMix, altXOutAct,
doubleInPop, doubleOutAct.
* print01Report() now reports in-degrees also for two-mode networks.
* Better error handling for sienaTimeTest and scoreTest
(EvaluateTestStatistic() in Sienatest.r).
* inOutAss is dyadic.
* Check for positive derivative matrix at the end of phase 1
omitted for effects with fixed parameters
(phase1.r, CalculateDerivative()).
* Corrected effectName and functionName of inPopIntn, outPopIntn,
inActIntn, and outActIntn ('in' and 'out' were missing).
* Extended error message for "invalid effect" in initializeFRAN().
* Console message when allowOnly=TRUE is active in sienaDependent().
* Some change in includeInteraction.Rd.
Changes in RSiena:
* New effects homXOutAct, FFDeg, BBDeg, RFDeg, diffXTransTrip
(ported from RSienaTest).
* sameXInPop and diffXInPop also added for two-mode networks;
but they are not dyadic!
* In names of behavior effects and statistics dropped the
(redundant) parts "behavior" and "beh.".
Changes in RSienaTest:
* New function extract.sienaBayes() (bayesTest.r).
* sienaBayes(): options diagonalize=0.2, doubleAveraging=0 for estimation
of initial models in initialization phase; save initial results in case of
divergence during initialization phase;
check for initialEstimates done only for non-rate parameters.
2015-07-18 R-Forge Revision 288
Changes in RSiena and RSienaTest:
* plot.sienaRI: new parameter "actors";
proportions with piechart improved; function of parameter radius changed.
* fra2 constructed only if findiff (phase3.r).
* siena.table does no more produce the double minus sign in html output.
Changes in RSiena:
* Correction of error for two-mode networks in sienaGOF.
Changes in RSienaTest:
* New effects homXOutAct, FFDeg, BBDeg, RFDeg, diffXTransTrip.
* sameXInPop and diffXInPop also added for two-mode networks;
but they are not dyadic!
* In names of behavior effects and statistics dropped the
(redundant) parts "behavior" and "beh.".
* sienaBayes: new parameter nSampRates;
correction in use of prevBayes (with extra createStores());
more efficient calculation of multivariate normal density.
* Small changes in HowToCommit.tex.
2015-06-02 R-Forge Revision 286
(Version number was kept the same...)
Changes in RSienaTest:
* Correction of error for two-mode networks in sienaGOF.
2015-05-21 R-Forge Revision 286
Changes in RSiena and RSienaTest:
* test11 dropped from \tests\parallel.R.
* print.xtable.sienaFit exported as S3method in Namespace.
Changes in RSienaTest:
* Larger dontrun part in examples for SienaRIDynamics
(there still is an error).
Changes in RSiena:
* SienaRIDynamics not exported from RSiena only (there still is an error),
SienaRIDynamics.Rd dropped.
2015-05-20 R-Forge Revision 285
Changes in RSiena and RSienaTest:
* tmax added to sienaFit objects and tconv.max mentioned in print.sienaFit().
* sienaAlgorithmCreate() has new arguments n2start,
truncation, doubleAveraging, standardizeVar;
this leads to various changes in phase2.r.
* Diagonalization corrected (matrix transpose: for defining changestep,
fra %*% z$dinvv replaced by z$dinvv %*% fra in phase2.r).
* Truncation based on multivariate deviation if
diagonalize < 1 (phase2.r).
* When there are missings in constant or changing monadic covariates,
and centered=FALSE for their creation by coCovar() or
varCovar(), the mean will be imputed (used to be 0, which was an error).
For changing covariates, this is the global mean.
* In coCovar() and varCovar() there is a new argument imputationValues,
which are used (if given) for imputation of missing values.
Like all missings, they are not used for the calculation
of the target statistics in the Method of Moments,
but only for the simulations.
* For non-centered covariates, averageAlter and averageInAlter
values for zero degrees were replaced by the observed covariate mean.
* New effects: outOutActIntn, sharedTo, from.w.ind
* New effectGroup tripleNetworkObjective (allEffects.csv, effects.r)
(see the manual or effects.r for its characteristics).
* In the target statistic for the 'higher' effect, contributions for
value(ego)=value(alter) are now set appropriately at 0.5 (was 0).
* Decent error message when there are (almost) all NA in the dependent
behavioural variable (effects.r, function getBehaviorStartingVals()).
* The centering within effects for similarity variables at distance 2
now is done by the same similarity means just as for the simX effect
(attr(mydata$cCovars$mycov, "simMean"), etc.).
(Background: this was done earlier by "simMeans",
the implementation of which contained an error;
but the differentiation between "simMean" and "simMeans" is not important,
so "simMeans" was dropped for simplicity.)
The object "simMeans" was not yet dropped from the data structure
transmitted to C++ (siena07internals.cpp. initializeFRAN.r)
* z$positivized changed from matrix to vector (not important enough
for so much memory use).
* test7 renamed test 5, test6 and new test7 added in \tests\parallel.R
Changes in RSienaTest:
* Small changes in HowToCommit.tex.
2015-04-02 R-Forge Revision 284
Changes in RSiena and RSienaTest:
* New effects: simEgoDist2, simEgoInDist2, simEgoDist2W, simEgoInDist2W,
sameXOutAct, diffXInPop, diffXOutAct.
* New effectGroups covarABNetNetObjective,
covarANetNetObjective, and covarBNetNetObjective (allEffects.csv, effects.r).
* Bug corrected that occurred in print01Report for a sienaGroup object
where the component objects have constant dyadic covariates
(sienaDataCreate.r).
* When a statistic is not plotted in plot.sienaGOF() because
its variance is 0, a note about this is printed to the screen.
* Minimum and maximum of plotted region in plot.sienaGOF()
is calculated without taking into account non-plotted statistics.
* Bug corrected with includeTimeDummy for timeDummy greater than
or equal to 10 (sienaTimeTest.r).
* In case of collinear parameter estimates, standard errors
are reported as NA (phase3.r, print07Report.r).
* At the end of subphases of phase 2, add condition
((!is.na(z$minacor)) & (!is.na(z$maxacor))) to avoid NA errors.
* Arguments main and ylab dropped from plot.sienaGOF(); they did
not work, and their functionality now is covered by the ...
argument (so using main and ylab as arguments now should work).
(Thanks to David Kavaler.)
* Various changes in DESCRIPTION and NAMESPACE files with respect
to Suggests and Depends; for packages in "Depends", require(packageName)
was dropped in all R functions; for package "network",
suppressPackageStartupMessages() was added to require();
for packages in "Suggests", the package name was everywhere
added as packageName:: before the function name.
* GPL requirement changed to >= 3.
Changes in RSienaTest:
* sienaBayes: correction in initialization of truncation rate parameters
based on prior; correction of sampleConstantParameters() (MCversion=0
used the wrong MH ratio, now was dropped altogether).
Changes in RSiena:
* reduceg parameter added (sienaAlgorithmCreate, siena07)
2014-12-11 R-Forge Revision 282
Changes in RSiena and RSienaTest:
* Effects cl.XWX and cl2.XWX corrected (thanks to Christoph Stadtfeld)
(MixedTwoPathTable.cpp).
* interactionType of gwesp.. effects was made dyadic (allEffects.csv)
* Some layout changes in warning message, phase3.r.
* New effects reciPop, reciAct, in3Plus, maxAlt, minAlt, transTrip1,
transTrip2.
* Effect antiInIsolate2 got alias in2Plus.
* inPop is dyadic effect (except for symmetric networks)
(it is sum_j x_{ij} (sum_{h \neq i} x_{hj} + 1)).
* egoX added as effect for non-directed networks
(can be important for representing effects of group-level covariates
in multi-group analyses).
* Components IActors and expectedI added to sienaRI() and print.sienaRI().
* The check for MaxDegree when running siena07 now works properly also
for sienaGroup objects (initializeFRAN).
* Manual introduces the term 'elementary effects'.
Changes in RSienaTest:
* sienaBayes: the stop caused by singularity of the precision
matrix after the multi-group estimation now is circumvented;
still a warning is printed to the screen. This allows the use of
more than one elementary effects having the same target statistic.
* sienaBayes: option priorRatesFromData changed to values 0-1-2,
with 0 = former FALSE, 1 = former TRUE, 2 = robust estimation of prior
for rate parameters from estimates at the end of initialization phase.
* sienaBayes: correction thetaMati (not important,
only for output to screen).
* sienaprint: correction of print.summary.sienaBayesFit for models
with more than one dependent variable.
2014-11-22 R-Forge Revision 281
Changes in RSiena:
* Translation of R messages was updated: changed file
\pkg\RSiena\po\R-RSiena.pot
2014-11-19 R-Forge Revision 280
Changes in RSienaTest:
* Small changes in help pages for sienaGOF and for sienaCompositionChange.
* New parameters nSampVarying and nSampConst in sienaBayes().
2014-11-13 R-Forge Revision 279
Changes in RSiena and RSienaTest:
* Effect AltsAvAlt renamed to avXAlt (with error message).
* effects object no longer used as argument for print01Report
(with error message).
* New effects sameXInPop, transRecTrip2,
totAlt, avInAlt, totInAlt, totRecAlt,
totXAlt, avXInAlt, totXInAlt,
avAltDist2, totAltDist2, avTAltDist2, totAAltDist2,
avXAltDist, totXAltDist2, avTXAltDist2, totAXAltDist2,
avInAltDist2, totInAltDist2, avTInAltDist2, totAInAltDist2,
avXInAltDist2, totXInAltDist2, avTXInAltDist2, totAXInAltDist2,
XWX1, XWX2, cl.XWX1, cl.XWX2.
* gwesp.. effects get endowment=TRUE in allEffects.csv, and thereby
obtain endowment and creation effects.
* In effect group behaviorBipartiteObjective, the following meaningless
effects were dropped: avAlt, avSim, totSim, avSimPopEgo, avSimPopAlt,
behDenseTriads, simDenseTriads.
* Effect class covarBehaviorBipartiteObjective split into
covarABehaviorBipartiteObjective (for covariates on first mode)
and covarBBehaviorBipartiteObjective (for covariates on second mode)
(allEffects.csv, effects.r).
* Added effect class covarABehaviorBipartiteObjective
and covarBBehaviorBipartiteObjective to effectsDocumentation.R.
* Undid duplication of covarNetNetObjective.
* For non-invertible covariance matrices, give diagnostic for the
linear combination that gives trouble at the end of the
algorithm (phase3.r).
* Correction of igraphNetworkExtraction() in sienaGOF-auxiliary.Rd
(the earlier version dropped isolated nodes from simulated networks).
* In sienaGOF-auxiliary.Rd, the example of constraint is replaced
by the example of eigenvector centrality (because constraint
is undefined for isolated nodes, leading to computational problems).
* sienaRIDynamics restored, after corrections.
* "file" parameter of sienaRI dropped (implied platform dependence).
* Parameter showAll added to descriptives.sienaGOF.
* Small correction of print.siena (reporting uponly/downonly).
* Some changes in print.sienaAlgorithm.
* Check in siena07 for incorrect MaxDegree specification.
* Correction of printing errors arising when result of score-type test is NA.
* maxRatio checked for NA or NaN in phase2.r.
* Section in manual about user-defined interaction effects updated.
* Siena_algorithms4.tex renamed Siena_algorithms.tex.
* Set diagonal of observed networks to 0 in sparseMatrixExtraction
(sienaGOF.r).
* Some improvement of error messages for sienaTimeTest.
* p-value for goodness of fit rounded to 3 decimal places (sienaGOF.r)
* Longer string length for output of z$requestedEffects$effectName
in print to output file (print07Report.r).
* File effects.pdf dropped from \inst\doc (it can be created
by effectsDocumentation() so the separate file is superfluous).
* AverageAlterEffect.cpp simplified.
Changes in RSienaTest:
* sienaBayes(): new parameters nImproveMH and priorRatesFromData;
truncate initial rate parameters depending on prior. Small change
in example in help page.
* glueBayes() corrected so that it can be applied sequentially.
* Small improvement in print.summary.sienaBayesFit().
* multipleBayesTest now allows matrix parameter to test
linear combinations.
* Improved plot.multipleBayesTest (shows truncation at 0).
2014-07-24 R-Forge Revision 278
Changes in RSiena and RSienaTest:
* Added s50s to RSiena.rda.
* Corrected se component of sienaFit objects
(should be standard error, was its square).
* new effects totDist2, altInDist2, totInDist2,
totDist2W, altInDist2W, totInDist2W.
* Warnings turned off for calculation of z$regrCoef
and z$regrCor in phase1.r and phase3.r.
* print01Report() improved for changing dyadic covariates,
and some layout improvement for long variable names,
and bug corrected for reporting upOnly/downOnly cases.
Changes in RSienaTest:
* Adapted sienaBayes and glueBayes and print.sienaBayes to allow
inclusion of interaction effects without the corresponding
main effects (use $requestedEffects instead of $effects).
* glueBayes(): checks of equality of prior parameters restricted
to !ratesInVarying; parameter nwarm2 added (sienaBayes.r).
* sienaBayes: Multiplied priorSigma for rate parameters by priorKappa;
changed z$nwarm to 0 if !is.null(prevBayes); change result of
dmvnorm() to -Inf if cov. matrix singular; dropped plotit functionality.
2014-06-22 R-Forge Revision 277
Changes in RSiena and RSienaTest:
* Some documentation of returnChains in siena07.Rd.
Changes in RSienaTest:
* Added function simpleBayesTest() (bayesTest.r, bayesTest.Rd).
* Added function multipleBayesTest() with print and plot
methods (bayesTest.r, bayesTest.Rd).
* Small changes in various functions in sienaBayes.r.
* Corrected starting printing sienaBayesFit at nfirst (sienaprint.r).
2014-06-02 R-Forge Revision 275, but version predated 1.1-276
Changes in RSiena and RSienaTest:
* Higher writefreq for batch operation (phase2.r, phase3.r)
* Error check for name of dependent variable to condition upon (initializeFRAN.r)
* Some corrections to allow behavior dependent variables that are not
coded as integer (in initializeFRAN and unpackBehavior, initializeFRAN.r)
* Parameters fix and test now used in includeEffects() (sienaeffects.r)
* Small bug fix in CalculateDerivative3 (phase3.r)
* Avoidance of ":::" in siena01.r and document.r by small changes.
Changes in RSienaTest:
* Changes in sienaBayes.r and its print and summary methods;
new function glueBayes.
2014-05-05 R-Forge Revision 275
Changes in RSiena and RSienaTest:
* Error message and stop in case of mismatch variable names
in algorithm object and in data (initializeFRAN.r).
* Force behavior and its attribute "change" to be integer in function
unpackBehavior() (initializeFRAN.r).
* Some tabs in .Rd files change to spaces.
2014-04-26 R-Forge Revision 274
Changes in RSiena and RSienaTest:
* Replaced (!strcmp...) by (!(strcmp...)) one more time
in sienainternals07.cpp.
2014-04-25 R-Forge Revision 273
Changes in RSiena and RSienaTest:
* NAMESPACE: removed entries sienaRIDynamics with the corresponding
print, summary, and plot methods (to be entered later when
problems have been solved).
* Removed sienaRIDynamics.Rd (ditto).
* Some changes in sienaRIDynamics.r but not enough to make it work.
* Duplication of outInv and outSqInv effects for two-mode networks
canceled (allEffects.csv).
* Correction of effect homWXClosure.
* Small change in print01Report.r to improve reporting
two files of composition change.
* sienaRI.r: require that <> argument is not NULL for non-Windows
operating systems.
Changes in RSienaTest:
* Minor changes in sienaBayes.r.
2014-04-17 R-Forge Revision 272
Changes in RSiena and RSienaTest:
* sienaRIDynamics.Rd: added
* sienaRIDynamics.r: small changes
* NAMESPACE: added entries sienaRIDynamics with the corresponding
print, summary, and plot methods.
2014-04-17 R-Forge Revision 271
Changes in RSiena and RSienaTest:
* Added component se (standard errors) to sienaFit objects
* Replaced (!strcmp...) by (!(strcmp...)) several times
in sienainternals07.cpp.
2014-04-13 R-Forge Revision 270
Changes in RSiena and RSienaTest:
* Moved plot(RI) in sienaRI.Rd to dontrun{...}.
* File changeLog renamed ChangeLog.
2014-04-13 R-Forge Revision 269
Changes in RSienaTest:
* In one line in sienaRIDynamics.r, "intervalsPerPeriods"
CHANGED TO "intervalsPerPeriod" (this was a misprint).
* Changes of version 1.1.268 in sienaRIDynamics ported from RSiena to RSienaTest.
* Small changes in sienaRI.r.
* sienaRI.Rd ported from RSiena to RSienaTest, with small changes.
* NAMESPACE: added entry sienaRI with the corresponding
print, summary, and plot methods.
* sienaBayes.r, sienaprint.r: updates to let sienaBayes.r accept
a wider range of data and models (e.g., user-defined interactions).
Changes in RSiena:
* sienaRIDynamics with plot and print methods temporarily left out.
* Small changes in sienaRI.r and sienaRI.Rd.
Changes in RSiena and RSienaTest:
* sienaDataCreate.r, sienautils.r, print01Report.r:
In coDyadCovar() and varDyadCovar(), centering now also is optional by the
new option "centered" (like it was done for coCovar() and varCovar()
in revision 1.1-251).
* sienaprint.r: correct bug when printing siena object with symmetric network.
* varDyadCovar: repaired bug occurring when calling it with a named list.
2014-04-08 R-Forge Revision 268
Changes in RSiena:
* sienaRI.Rd: added
* sienaRIDynamics.r: bugfix
* NAMESPACE: added entries sienaRI and sienaRIDynamics with the corresponding
print, summary, and plot methods.
2014-03-30 R-Forge Revision 267
Changes in RSiena:
Added R-files that contain functionalities to calculate, print and plot
relative importances of effects
* sienaRI.r: new
* sienaRIDynamics.r: new
Changes in RSiena and RSienaTest:
* Model.cpp: bugfix
2014-03-17 R-Forge Revision 266
Changes in RSiena and RSienaTest:
* getTargets.r: Added actorTargets function for use with multiple imputation
of behaviour variable.
* sienaModelCreate.r, initializeFRAN.r, setup07setup.cpp, Model.cpp, Model.h,
MLSimulation.cpp, MLSimulation.h: added localML option to sienaAlgorithm.
* allEffects.csv: added "local" column for use with localML option.
2014-03-14 R-Forge Revisions 261-265
Changes in RSiena: Included exactly the same changes in RSiena
as have been included in RSienaTest in Revision 255
2014-03-14 R-Forge Revisions 255-260
Changes in RSienaTest:
* Modified getTargets(...) such that it is possible to return target
statistics of individual actors
as well as change contributions necessary for calculating relative
importances of effects (siena07setup.cpp).
* Call of getTargets(...) is correspondingly adapted (initializeFRAN.r).
* Modified simstats0c(...) such that it is possible to return
statistics of individual actors in simulated networks
as well as statistic contributions of all potential tie flips or
behavior changes for all simulated micro-steps (simstatsc.r).
* Modified model(...) such that it calculates and returns the statistics
newly needed in simstats0c(...) (siena07model.cpp)
* Added methods, functions, constructors, and fields needed due to changes
in getTargets(...) and model(...):
public void needChangeContributions(bool flag) (Model.h, Model.cpp)
public bool needChangeContributions() (Model.h, Model.cpp)
private bool lneedChangeContributions2 (Model.h)
SEXP getChangeContributionsList(...) (siena07utilities.h, siena07utilities.cpp)
SEXP createRObjectAttributes(...) (siena07utilities.h, siena07utilities.cpp)
void getChangeContributionStatistics(...) (siena07internals.h, siena07internals.cpp)
void getActorStatistics(...) (siena07internals.h, siena07internals.cpp)
public StatisticCalculator(..., bool) (StatisticCalculator.h, StatisticCalculator.cpp)
public StatisticCalculator(..., bool, bool) (StatisticCalculator.h, StatisticCalculator.cpp)
public vector staticChangeContributions(...) (StatisticCalculator.h, StatisticCalculator.cpp)
public double * actorStatistics(...) (StatisticCalculator.h, StatisticCalculator.cpp)
private bool lneedActorStatistics (StatisticCalculator.h, StatisticCalculator.cpp)
private bool lcountStaticChangeContributions (StatisticCalculator.h, StatisticCalculator.cpp)
private map lactorStatistics (StatisticCalculator.h, StatisticCalculator.cpp)
private map > lstaticChangeContributions (StatisticCalculator.h, StatisticCalculator.cpp)
* Modifications of functions
calculateNetworkEvaluationStatistics(...) (StatisticCalculator.cpp)
calculateBehaviorStatistics(...) (StatisticCalculator.cpp)
Specifically, assignments of form "this->lstatistics[pInfo] = pEffect->evaluationStatistic()"
are extended in order to also store actor statistics and change contributions.
* Added a field and corresponding get-and set-methods to class MiniStep
for storing potential statistic contributions of changes in micro-steps
private map > * lpChangeContributions (MiniStep.h, MiniStep.cpp)
public map >* changeContributions() (MiniStep.h, MiniStep.cpp)
public void changeContributions(map > * contributions) (MiniStep.h, MiniStep.cpp)
* Added field to class DependentVariable (DependentVariable.h, DependentVariable.cpp)
protected map > * lpChangeContribution;
* Modifications of functions makeChange() and calculateTieFlipProbabilities()
in class NetworkVariable (NetworkVariable.cpp)
calculateProbabilities(), totalEvaluationContribution(),
totalEndowmentContribution(), and totalCreationContribution()
in class BehaviorVariable (BehaviorVariable.cpp)
in order to store statistic contributions of effects to all possible
choices of an actor in the current micro-step
* Added functions to classes NetworkEffect and BehaviorEffect
(NetworkEffect.h, NetworkEffect.cpp, BehaviorEffect.h, BehaviorEffect.cpp)
such that, in addition to the sum of statistics ovar all actors,
individual actor statistics can be returned
public virtual pair statistic(..., bool needActorStatistics)
public virtual pair evaluationStatistic(...,bool needActorStatistics)
public virtual pair endowmentStatistic(..., bool needActorStatistics)
public virtual pair creationStatistic(..., bool needActorStatistics)
2014-02-17 R-Forge Revision 254
Changes in RSiena and RSienaTest:
* Give the capital P to HomCovariateMixedTwoPathFunction.cpp
and HomCovariateMixedTwopathFunction.h.
2014-02-16 R-Forge Revision 253
Changes in RSiena and RSienaTest:
* Give the capital E to HomCovariateTransitiveTripletsEffect.h
and HomCovariateTransitiveTripletsEffect.cpp.
2014-02-15 R-Forge Revision 252
Changes in RSiena:
* Set eoln property to "native" for HomCovariateTransitiveTripletsEffect.h
and HomCovariateTransitiveTripletsEffect.cpp.
Changes in RSienaTest:
* Small change to sienaBayes (mustUpdate).
2014-02-13 R-Forge Revision 251
Changes in RSiena and RSienaTest:
* Also cater for behavior user-specified interactions in includeInteraction
and setEffect (sienaeffects.r).
* New effects: "cl.XWX", "homXTransTrip", "homWXClosure", and "sharedPop".
* Effect "cycle4" extended to non-directed one-mode networks
* (for directed one-mode networks it is "sharedPop").
* Effects "gwesp.." modified; extended to non-directed networks.
* Correction: Effect "to" is not a dyadic effect (allEffects.csv)
* sienaGOF: added the name of the siena fit object
as attribute "sienaFitName" to each of the sienaGofTest objects.
* Manual: added paragraph about how to import results from
xtable() and siena.table() into MS-Word.
* Correction in sparseNetworkExtraction() to avoid errors occurring
when the extracted network has no edges (sienaGOF.r).
* In the help page for sienaGOF-auxiliary, geodesic distances
changed to non-directed; which avoids a further error
when the extracted network has no edges.
* Correction of an error in print.siena for data sets including
other types than "oneMode".
* Changed bandwidth selector for violin plots in plot.sienaGOF to "nrd",
to avoid long violins in cases where all simulations have the same outcome.
* Added parameter "includeRandoms" to print.sienaEffects.
* sienaGOF checks that the sienaFitObject object was not made
by maximum likelihood estimation.
Changes in RSiena:
* Ported effects outRateLog and outTrunc2 from RSienaTest.
* Ported changes (new folders and files: \layers, \iterators, etc.)#
in \src\network\ from RSienaTest.
Changes in RSienaTest:
* Major changes to sienaBayes and print.sienaBayes,
to permit estimation of multilevel stochastic actor-oriented model with
effects varying between groups for only a subset of the effects.
2013-12-04 R-Forge Revision 250
Changes in RSiena and RSienaTest:
* New option "centered" in coCovar and varCovar (sienautils.r,
sienaDataCreate.r, print01Report.r, sienaprint.r).
* setEffect, updateTheta, and prevAns in siena07() now also cater
for user-specified interactions (sienaeffects.r, initializeFRAN.r).
* Wald.RSiena and Multipar.RSiena added (Sienatest.r).
* Error occurrence with message about cvalue in EvaluateTestStatistic
corrected (Sienatest.r).
* Divergent parameters in siena07() get NA for their rows and columns
in the covariance matrix (function phase3.2 in phase3.r).
The following changes in revision 244 were ported from RSienaTest to RSiena:
* In siena08, also report Bonferroni combination
of the two Fisher combinations.
* In phase2.r, rolled back change in truncation from version 1.1-227
to the earlier procedure.
* descriptives.sienaGOF added.
* Minor changes of output in siena.table, print07Report.r, print.siena,
printInitialDescription.r, and in error message for includeEffects.
* Change artificial results from 999 to NA (robmon.r, phase3.r)
* For ML estimation: added autocorrelations during phase 3
to print.summary.sienaFit (sienaprint.r)
2013-11-29 R-Forge Revision 249
Changes in RSiena and RSienaTest:
* Fix warning in EpochSimulation regarding String-plus-int.
Once RSiena moves to C++11 this can be replaced by std::to_string
2013-11-28 R-Forge Revision 247
Changes in RSiena and RSienaTest:
* Fix problem with std::map::const_iterator constructor
call in IncidentTieIterator.cpp
* Fix ambiguous call of abs in AlterCovariateAv/TotSimEffect by
replacing cmath by cstdlib
* Fix problem with using namespace std in DiffusionEffectValueTable
and EffectValueTable
2013-10-31 R-Forge Revision 246
Changes in RSiena and RSienaTest:
* New behavior objective function effects avSimAltX, totSimAltX and
avAltAltX to differentiate sources of peer influence in directed networks.
* Added effect class covarBehaviorNetObjective to effectsDocumentation.R.
* Fix of a bug that occurred in the case of on average decreasing behavior
variables.
2013-10-16 R-Forge Revision 245
Changes in RSienaTest:
* New structural rate effect outRateLog
(allEffects.csv, StatisticCalculator.cpp, DependentVariable.cpp,
siena07Internals.cpp, StructuralRateEffect.cpp, Utils.h, Utils.cpp,
RSiena_Manual.tex).
* In siena08, also report Bonferroni combination
of the two Fisher combinations.
* In phase2.r, rolled back change in truncation from version 1.1-227
to the earlier procedure.
* descriptives.sienaGOF added.
* outTrunc2 effect added.
* Minor changes of output in siena.table, print07Report.r, print.siena,
printInitialDescription.r, and in error message for includeEffects.
* Change artificial results from 999 to NA (robmon.r, phase3.r)
* For ML estimation: added autocorrelations during phase 3
to print.summary.sienaFit (sienaprint.r)
* Small addition to siena_Algorithms4.tex (statistics for rate effects).
* Bit of cleaning up (phase3.r)
* Start of the manual reorganized and partially rewritten (with help from
Zsofia Boda and Andras Voros); instructions for siena01Gui separated
in siena01gui.pdf.
* Larger example for sienaCompositionChange.Rd.
2013-10-16 R-forge revision 244
Changes in RSienaTest:
* Corrected DistanceTwoLayer.cpp
2013-09-17 R-forge revision 243
Changes in RSiena and RSienaTest:
* Correct bug in EffectFactory for isolatePop effect.
* Improved plotting of sienaGOF objects so that observed values
outside of the range of simulated values don't run off the chart.
* Improve treatment of structural values in sienaGOF by
modifying sparseMatrixExtraction(); and simplify
networkExtraction() by building it directly on sparseMatrixExtraction().
Changes in RSiena:
* Add functions AntiIsolateEffect.h and AntiIsolateEffect.cpp
which were forgotten to include in revision 242.
2013-09-10 R-forge revision 242
Changes in RSiena and RSienaTest:
* New effects: anti isolates, anti in-isolates, anti in-near-isolates.
* Effect inIsolatePop dropped (it was shortlived).
* Improved printing of results of siena07() in the case simOnly.
* Prettier response printed to console for includeEffects() and setEffect().
* z$estMeans added to siena objects z:
vector of estimated expected values of statistics;
this is colMeans(z$sf) + z$targets but if dolby,
the regression on the scores is subtracted.
* Correction to Dolby option for the case of more than 2 waves:
in phase1.r and phase3.r, scores are added (instead of averaged)
over waves. (Averaging was wrong, because in phase 2 they are added.)
* All single quotes in .Rd files replaced by double quotes
(first seemed required by the disappearance of inconsolata.sty,
probably superfluous).
Changes in RSienaTest:
* Small additions to Siena_algorithms4.
2013-08-23 R-forge revision 241
Changes in RSiena and RSienaTest:
* Corrected bug that occurred if there was only one option
for choice of alter.
* Dropped scripts.r from \tests.
* Added "silent=TRUE" to various try statements (mainly associated with
inversion of singular matrices) to avoid execution errors.
* Added various degree-related effects to bipartite networks.
* New effect inIsolatePop.
* Drop multiplication of the target statistic for the inPop effect by n.
Changes in RSienaTest:
* Small unimportant changes in sienaeffects.r, robmon.r, sienaprint.r
2013-08-08 R-forge revision 240
Changes in RSienaTest:
* Added the parameter reduceg to siena07()
(sienaModelCreate.r, phase2.r, robmon.r, printDataReport.r, sienaprint.r)
Changes in RSienaTest and RSiena:
* Added effects crprod and inPopIntn to bipartiteBipartiteObjective group
(allEffects.csv, effects.pdf).
* Added tstat (t-statistic for convergence) to sienaFit objects.
* Clarification in sienaCompositionChange.Rd.
* Small change to siena07.Rd.
2013 July 26 - August 2 R-Forge revisions 232-239
Various changes by Mark Ortmann to RSienaTest.
2013-06-18 CRAN version 1.1-232
Changes in RSiena:
* Drop the possibility to use obsolete package snow and rlecuyer
if Rversion < 2.14.0; functionality replaced by
package parallel (siena07.r).
* Correct DESCRIPTION file to satisfy CRAN requirements.
* Correct lines in various help pages that were too long
in the pdf version.
2013-06-15 R-forge revision 231
Changes in RSiena as well as RSienaTest:
* Make the "cumulative" option operational in BehaviorDistribution()
(sienaGOF.r).
* Correct bug in treatment of missing values in sparseMatrixExtraction()
(sienaGOF.r).
* Allow sparse observed data matrices, and structural zeros and ones, in
sparseMatrixExtraction() and networkExtraction(),
and bipartite networks in networkExtraction() (sienaGOF.r).
* In case prevAns exists, recalculate z$dinvv instead of taking it
from prevAns (initializeFRAN.r) (parameter diag may have changed)
* Report correct centering (by overall means) of individual
covariates for multi-group objects (print01Report.r).
* in sienaTimeTest: string in toTest$effectName set to "chisq="
instead of "p=" (sienaTimeTest.r)
* In print method for sienaAlgorithm, report of conditional corrected
in case !x$cconditional (sienaprint.r)
* If there is a composition change object, MoM estimation is forced
to be non-conditional (initializeFRAN.r). This is reported in the help page
(sienaCompositionChange.Rd).
* Small changes in help pages sienaAlgorithm.Rd, plot.sienaTimeTest.Rd,
2013-05-10 R-forge revision 230
Changes in RSiena as well as RSienaTest:
* Add attribute maxObsOutDegree to