Version 1.7.5 of the BAS package adds an internal function to count the number of models
that satisfy "hereditary" constraints. This is used in the force.heredity option
in bas.lm to reduce the number of models considered in the sampling process and should
reduce the memory requirements and speed up the sampling process. This currently works only
for factors included in the model formula, but not with factors always included in the model
orwith other hereditaty constraints such as with polynomials. (theforce.heredity option
does work with these other constraints). This is a first step in reducing the number of models
aallocated in the sampling process. Future updates will include other hereditary constraints.
added hexagon sticker
add copyright to source files to satisfy OpenSSF Best Practice
fixed (issue #89) reported as Error on CRAN Check page for compiling BAS under R-devel
with clang19. Removed legacy definitions of MACHEPS and MAXNUM from Cephes
and replaced with DLB_EPSILON and DBL_MAX in R. Files in src/ mconf.h and const.c are no longer
used and will be reomved from in the future.
fixed (issue #87) prior inclusion probabilities using a Bernoulli prior other than 0.5
were incorrect if include.always was used to include some variables always.
Added unit test in test-priorprobs.R
SETLENGTH (issue #82) to address stack imbalance issues seen in interactive checks and not flagged in R CMD check.added method="AMCMC" for bas.lm to use adaptive independent Metropolis Hastings for sampling
models. With option importance.sampling = TRUE the adaptive independent proposal
and be used for importance sampling with improved estimation of model probabilities and inclusion probabilities based on the Horivitz-Thompsom / Hajek estimator.
added unit tests for link functions implemented in family.c
fixed (issue #81) Removed legacy definitions of ‘PI’ and ‘Free’ and replaced with‘M_PI’ and ‘R_Free’ to comply with ‘STRICT_R_HEADERS’ prevent package removal on 9/23/2024
fixed (issue #82) avoid SETLENGTH as non-API function when truncating vectors
Initialized vector se via memset and disp = 1.0 in fit_glm.c (issue #72)
Initialized variables in hyp1f1.c from testthat (issue #75)
Removed models that have zero prior probability in bas.lm and bas.glm (issue #74)
Fixed error in bayesglm.fit to check arguments x or y for correct type before calling C and added unit test (issue #67)
Added support for Gamma regression for bas.glm, with unit tests and
example (Code contributed by @betsyberrson)
added error if supplied initial model for the bas.lm sampling methods "MCMC" and "MCMC+BAS" had prior probability zero.
fixed printing problems as identified via checks
fixed indexing error for bas.lm and method = "MCMC+BAS" as bas.lm using method = "MCMC+BAS" crashed with a segmentation fault if bestmodel is not NULL or the null model. GitHub issue #69
fixed error in predict.bas with se.fit=TRUE if there is only one predictor. GitHub issue #68 reported by @AleCarminati
added unit test to test-predict.R
Fixed error in coef for bas.glm objects when using a betaprior of class
IC, including AIC and BIC Github issue #65
Fixed error when using Jeffreys prior in bas.glm with the
include.always option and added unit test in test-bas-glm.R.
Github issue #61
Fixed error for extracting coefficients from the median probability model
when a formula is passed as an object rather than a literal, and added
a unit test to test-coefficients.R Github issues #39 and #56
pivot=FALSE for bas.lm as default uses pivoting and documentation
indicates that pivot=FALSE should only be used in the full rank case so that
users should not encounter this issue in practice. Users will continue to see
a warning of NA's are returned, but should be aware that not all platforms may
produce a warning (such as M1mac). Github issue #62Added checks and unit-tests to see if modelprior is of class 'prior' resolving Github Issue #57
Removed polevl.c, psi.c and gamma.c from Cephes as no longer used after switching to R's internal functions
DOUBLE_EPS with DBL_EPSILON for R 4.2.0 release (in two places) so restore on CRANreplaced deprecated DOUBLE_EPS with DBL_EPSILON for R 4.2.0 release
fixed warnings from CRAN checks for under R devel (use of | and if with class)
added a function trCCH that uses integration to compute the normalizing constant in the
Truncated Compound Confluent Hypergeometric distribution that provides the
correct normalizing constant for Gordy (1998) and is more stable for large values
compared to the current phi1 function. This is now used in the TCCH prior for bas.glm.
Rewrote phi1 function to use direct numerical integration (phi1_int) when Wald statistic is large so that marginal likelihoods are not NA as suggested by Daniel Heeman and Alexander Ly (see below). This should improve stability of estimates of Bayes Factors and model probabilities from bas.glm that used the HyperTwo function, including coefficient priors for hyper.g.n(), robust(), and intrinsic(). Added additional unit tests.
Added thin as an option for bas.glm
added unit tests and examples to show the connections between the special
functions trCCH, phi1, 1F1 and 2F1
added internal function for phi1_int when the original HyperTwo function returns NA Issue #55 See more details above.
corrected the shrinkage estimate under the CCH prior that did not include
terms involving the beta function.
USE_FC_LEN_T for character stringslog_laplace_F21 which had an uninitialized variable
leading to NaN being returned from R function hypergeometric2F1Fixed WARNING under fedora-clang-devel. Added climate.dat file to package for building vignette so that package does not violate CRAN's policy for accessing internet resources and is more permanent if file location/url changes locally.
Fixed testthat errors under Solaris. Default settings for force.heredity is
set back to FALSE in bas.lm and bas.glm so that methods work on all
platforms. For Solaris, users who wish to impose the force.heredity
constraint may use the post-processing function.
Fixed valgrind error in src/ZS_approx_null_np.c for invalid write noted in CRAN checks
fixed function declaration type-mismatch and argument errors identified by LTO noted in CRAN checks
Added contrast=NULL argument to bas.lm and bas.glm so that non-NULL contrasts do not
trigger warning in model.matrix as of R 3.6.0. Bug #44
Added check for sample size equal to zero due to subsetting or missing data Bug #37
Fixed errors identified on cran checks https://cran.r-project.org/web/checks/check_results_BAS.html
initialize R2_m = 0.0 in lm_mcmcbas.c (lead to NA's with clang on debian and fedora )
switch to default of pivot = TRUE in bas.lm, adding tol as an argument to control tolerance in cholregpovot for improved stability across platforms with singular or nearly singular designs.
valgrind messages: Conditional jump or move depends on uninitialized value(s). Initialize vectors allocated via R_alloc in lm_deterministic.c and glm_deterministic.c.
Included an option pivot=TRUE in bas.lm to fit the models using a pivoted Cholesky decomposition to allow models that are rank-deficient. Enhancement #24 and Bug #21. Currently coefficients that are not-estimable are set to zero so that predict and other methods will work as before. The vector rank is added to the output (see documentation for bas.lm) and the degrees of freedom methods that assume a uniform prior for obtaining estimates (AIC and BIC) are adjusted to use rank rather than size.
Added option force.heredity=TRUEto force lower order terms to be included if higher order terms are present (hierarchical constraint) for method='MCMC' and method='BAS' with bas.lm and bas.glm. Updated Vignette to illustrate. enhancement #19. Checks to see if parents are included using include.always pass issue #26.
Added option drop.always.included to image.bas so that variables that are always included may be excluded from the image. By default all are shown enhancement #23
Added option drop.always.included and subset to plot.bas so that variables that are always included may be excluded from the plot showing the marginal posterior inclusion probabilities (which=4). By default all are shown enhancement #23
update fitted.bas to use predict so that code covers both GLM and LM cases with type='link' or type='response'
Updates to package for CII Best Practices Badge certification
Added Code Coverage support and more extensive tests using test_that.
fixed issue #36 Errors in prior = "ZS-null" when R2 is not finite or out of range due to model being not full rank. Change in gexpectations function in file bayesreg.c
fixed issue #35 for method="MCMC+BAS" in bas.glm in glm_mcmcbas.c when no values are provided for MCMC.iterations or n.models and defaults are used. Added unit test in test-bas-glm.R
fixed issue #34 for bas.glm where variables in include.always had marginal inclusion probabilities that were incorrect. Added unit test in test-bas-glm.R
fixed issue #33 for Jeffreys prior where marginal inclusion probabilities were not renormalized after dropping intercept model
fixed issue #32
to allow vectorization for phi1 function in R/cch.R
and added unit test to "tests/testthat/test-special-functions.R"
fixed issue #31 to coerce g to be a REAL for g.prior prior and IC.prior in bas.glm; added unit-test "tests/testthat/test-bas-glm.R"
fixed issue #30 added n as hyper-parameter if NULL and coerced to be a REAL for intrinsic prior in bas.glm; added unit-test
fixed issue #29 added n as hyper-parameter if NULL and coerced to be a REAL for beta.prime prior in bas.glm; added unit-test
fixed issue #28 fixed length of MCMC estimates of marginal inclusion probabilities; added unit-test
fixed issue #27 where expected shrinkage with the JZS prior was greater than 1. Added unit test.
fixed output include.always to include the intercept issue #26 always so that drop.always.included = TRUE drops the intercept and any other variables that are forced in. include.always and force.heredity=TRUE can now be used together with method="BAS".
added warning if marginal likelihoods/posterior probabilities are NA with default model fitting method with suggestion that models be rerun with pivot = TRUE. This uses a modified Cholesky decomposition with pivoting so that if the model is rank deficient or nearly singular the dimensionality is reduced. Bug #21.
corrected count for first model with method='MCMC' which lead to potential model with 0 probability and errors in image.
coerced predicted values to be a vector under BMA (was a matrix)
fixed size with using method=deterministic in bas.glm (was not updated)
fixed problem in confint with horizontal=TRUE when intervals are point mass at zero.
suppress warning when sampling probabilities are 1 or 0 and the number of models is decremented
Issue #25
changed force.heredity.bas to re-normalize the prior probabilities rather than to use a new prior probability based on heredity constraints. For future, add new priors for models based on heredity. See comment on issue #26.
Changed License to GPL 3.0
variable.names to extract variable names in the highest probability model, median probability
model, and best probability model for objects created by predict.predict.basglm which had that type = "link" was the default for prediction issue #18add na.action for handling NA's for predict methods issue #10
added include.always as new argument to bas.lm. This allows a formula to specify which terms should always be included in all models. By default the intercept is always included.
added a section to the vignette to illustrate weighted regression and the
force.heredity.bas function to group levels of a factor so that they enter
or leave the model together.
fixed problem if there is only one model for image function;
github issue #11
fixed error in bas.lm with non-equal weights where R2 was incorrect.
issue #17
predict argument in predict.bas, predict.basglm and internal functions as it is not utilizedconfint.coef.bas when parm is a character stringBayes.outlier if prior
probability of no outliers is providedfixed issue with scoping in eval of data in predict.bas if dataname is defined in local env.
fixed issue 10 in github (predict for estimator='BPM' failed if there were NA's in the X data. Delete NA's before finding the closest model.
fixed bug in 'JZS' prior - merged pull request #12 from vandenman/master
fixed bug in bas.glm when default betaprior (CCH) is used and inputs were INTEGER instead of REAL
removed warning with use of 'ZS-null' for backwards compatibility
updated print.bas to reflect changes in print.lm
Added Bayes.outlier function to calculate posterior probabilities of outliers using the method from Chaloner & Brant for linear models.
Added new method for bas.lm to obtain marginal likelihoods with the Zellner-Siow Priors for "prior= 'JZS' using QUADPATH routines for numerical integration. The optional hyper parameter alpha may now be used to adjust the scaling of the ZS prior where g ~ G(1/2, alpha*n/2) as in the BayesFactor package of Morey, with a default of alpha=1 corresponding to the ZS prior used in Liang et al (2008). This also uses more stable evaluations of log(1 + x) to prevent underflow/overflow.
Priors ZS-full for bas.lm is planned to be deprecated.
replaced math functions to use portable C code from Rmath and consolidated header files
Fixed unprotected ANS in C code in glm_sampleworep.c and sampleworep.c after call to PutRNGstate and possible stack imbalance in glm_mcmc.
Fixed problem with predict for estimator=BPM when newdata was one row
Fixed non-conformable error with predict when new data was from a dataframe with one row.
Fixed problem with missing weights for prediction using the median probability model with no new data.
Extract coefficient summaries, credible intervals and plots for the HPM and MPM in addition to the default BMA by adding a new estimator argument to the coef function. The new n.models argument to coef provides summaries based on the top n.models highest probability models to reduce computation time. 'n.models = 1' is equivalent to the highest probability model.
use of newdata that is a vector is now deprecated for predict.bas; newdata must be a dataframe or missing, in which case fitted values based on the dataframe used in fitting is used
factor levels are handled as in lm or glm for prediction when there may be only level of a factor in the newdata
fixed issue for prediction when newdata has just one row
fixed missing id in plot.bas for which=3
bas.lm to agree with documentationrenormalize that selects whether the Monte Carlo frequencies are used to estimate posterior model and marginal inclusion probabilities (default renormalize = FALSE) or that marginal likelihoods time prior probabilities that are renormalized to sum to 1 are used. (the latter is the only option for the other methods); new slots for probne0.MCMC, probne0.RN, postprobs.RN and postprobs.MCMC.coefficients function.bas.lm and bas.glmna.action for bas.lm and bas.glm to omit missing data.confint.pred.bas or confint.coef.bas. See the help files for an example or the vignette.se.fit option in predict.basglm.testBF as a betaprior option for bas.glm to implement Bayes Factors based on the likelihood ratio statistic's distribution for GLMs.A vignette has been added at long last! This illustrates several of the new features in BAS such as
confint.pred.bas()confint.coef.bas()predict.bas()type to specify estimator in fitted.bas and replaced with estimator so that predict() and fitted() are compatible with other S3 methods.bas to avoid NAMESPACE conflicts with other librariesfitted.bas or predict.basbas.glmdiagnostic() function for checking convergence of bas objects created with method = "MCMC""plot.bas that appears with Sweavecoef.bma when there is just one predictorbas rather than
bma to avoid name conflicts with other packagesfixed problem with duplicate models if n.models was > 2^(p-1) by restricting n.models
hypergeometric2F1 function that is callable by R
predict.bma to center newdata using the mean(X)-Allowed pruning of model tree to eliminate rejected models
method = "MCMC+BAS")-Cleaned up all .Call routines so that all objects are duplicated or allocated within code