Changelog
Source:NEWS.md
bayestestR (development version)
Changes
Support for
posterior::rvar
-type column in data frames. For example, a data framedf
with anrvar
column".pred"
can now be called directly viap_direction(df, rvar_col = ".pred")
.Added support for marginaleffects
The ROPE or threshold ranges in
rope()
,describe_posterior()
,p_significance()
andequivalence_test()
can now be specified as a list. This allows for different ranges for different parameters.Results from objects generated by emmeans (
emmGrid
/emm_list
) now return results with appended grid-data.-
Usability improvements for
p_direction()
:Results from
p_direction()
can directly be used inpd_to_p()
.p_direction()
gets anas_p
argument, to directly convert pd-values into frequentist p-values.p_direction()
gets aremove_na
argument, which defaults toTRUE
, to removeNA
values from the input before calculating the pd-values.Besides the existing
as.numeric()
method,p_direction()
now also has anas.vector()
method.
p_significance()
now accepts non-symmetric ranges for thethreshold
argument.p_to_pd()
now also works with data frames returned byp_direction()
. If a data frame contains apd
,p_direction
orPD
column name, this is assumed to be the pd-values, which are then converted to p-values.p_to_pd()
for data frame inputs gets aas.numeric()
andas.vector()
method.
bayestestR 0.14.0
CRAN release: 2024-07-24
Breaking Changes
- Arguments named
group
,at
,group_by
andsplit_by
will be deprecated in future releases of easystats packages. Please useby
instead. This affects following functions in bayestestR:estimate_density()
.
Changes
bayesian_as_frequentist()
now supports more model families from Bayesian models that can be successfully converted to their frequentists counterparts.bayesfactor_models()
now throws an informative error when Bayes factors for comparisons could not be calculated.
Bug fixes
- Fixed issue in
bayesian_as_frequentist()
for brms models with0 + Intercept
specification in the model formula.
bayestestR 0.13.2
CRAN release: 2024-02-12
Breaking Changes
pd_to_p()
now returns 1 and a warning for values smaller than 0.5.map_estimate()
,p_direction()
,p_map()
, andp_significance()
now return a data-frame when the input is a numeric vector. (making the output consistently a data frame for all inputs.)Argument
posteriors
was renamed intoposterior
. Before, there were a mix of both spellings, now it is consistentlyposterior
.
Bug fixes
Fixed issues in various
format()
methods, which did not work properly for some few functions (likep_direction()
).Fixed issue in
estimate_density()
for double vectors that also had other class attributes.Fixed several minor issues and tests.
bayestestR 0.13.1
CRAN release: 2023-04-07
Changes
Improved speed performance when functions are called using
do.call()
.Improved speed performance to
bayesfactor_models()
forbrmsfit
objects that already included amarglik
element in the model object.
New functionality
-
as.logical()
forbayesfactor_restricted()
results, extracts the boolean vector(s) the mark which draws are part of the order restriction.
Bug fixes
p_map()
gains a newnull
argument to specify any non-0 nulls.Fixed non-working examples for
ci(method = "SI")
.Fixed wrong calculation of rope range for model objects in
describe_posterior()
.Some smaller bug fixes.
bayestestR 0.13.0
CRAN release: 2022-09-18
Breaking
The minimum needed R version has been bumped to
3.6
.contr.equalprior(contrasts = FALSE)
(previouslycontr.orthonorm
) no longer returns an identity matrix, but a shifteddiag(n) - 1/n
, for consistency.
New functionality
-
p_to_bf()
, to convert p-values into Bayes factors. For more accurate approximate Bayes factors, usebic_to_bf()
. -
bayestestR now supports objects of class
rvar
from package posterior. -
contr.equalprior
(previouslycontr.orthonorm
) gains two new functions:contr.equalprior_pairs
andcontr.equalprior_deviations
to aide in setting more intuitive priors.
Changes
- has been renamed
contr.equalprior
to be more explicit about its function. -
p_direction()
now accepts objects of classparameters_model()
(fromparameters::model_parameters()
), to compute probability of direction for parameters of frequentist models.
bayestestR 0.12.1
CRAN release: 2022-05-02
Breaking
Bayesfactor_models()
for frequentist models now relies on the updatedinsight::get_loglikelihood()
. This might change some results for REML based models. See documentation.estimate_density()
argumentgroup_by
is renamedat
.All
distribution_*(random = FALSE)
functions now rely onppoints()
, which will result in slightly different results, especially with smalln
s.Uncertainty estimation now defaults to
"eti"
(formerly was"hdi"
).
Changes
bayestestR functions now support
draws
objects from package posterior.rope_range()
now handles log(normal)-families and models with log-transformed outcomes.New function
spi()
, to compute shortest probability intervals. Furthermore, the"spi"
option was added as new method to compute uncertainty intervals.
Bug fixes
-
bci()
for some objects incorrectly returned the equal-tailed intervals.
bayestestR 0.11.1
New functions
-
describe_posterior()
gains aplot()
method, which is a short cut forplot(estimate_density(describe_posterior()))
.
bayestestR 0.11
Bug fixes
Fixed issues related to last brms update.
Fixed bug in
describe_posterior.BFBayesFactor()
where Bayes factors were missing from out put ( #442 ).
bayestestR 0.10.0
CRAN release: 2021-05-31
Breaking
- All Bayes factors are now returned as
log(BF)
(column namelog_BF
). Printing is unaffected. To retrieve the raw BFs, you can runexp(result$log_BF)
.
New functions
-
bci()
(and its aliasbcai()
) to compute bias-corrected and accelerated bootstrap intervals. Along with this new function,ci()
anddescribe_posterior()
gain a newci_method
type,"bci"
.
bayestestR 0.9.0
CRAN release: 2021-04-08
Breaking
The default
ci
width has been changed to 0.95 instead of 0.89 (see here). This should not come as a surprise to the long-time users ofbayestestR
as we have been warning about this impending change for a while now :)Column names for
bayesfactor_restricted()
are nowp_prior
andp_posterior
(wasPrior_prob
andPosterior_prob
), to be consistent withbayesfactor_inclusion()
output.Removed the experimental function
mhdior
.
General
Support for
blavaan
models.Support for
blrm
models (rmsb).Support for
BGGM
models (BGGM).check_prior()
anddescribe_prior()
should now also work for more ways of prior definition in models from rstanarm or brms.
Bug fixes
Fixed bug in
print()
method for themediation()
function.Fixed remaining inconsistencies with CI values, which were not reported as fraction for
rope()
.Fixed issues with special prior definitions in
check_prior()
,describe_prior()
andsimulate_prior()
.
bayestestR 0.8.2
CRAN release: 2021-01-26
Changes to functions
- All
.stanreg
methods gain acomponent
argument, to also include auxiliary parameters.
Bug fixes
bayesfactor_parameters()
no longer errors for no reason when computing extremely un/likely direction hypotheses.bayesfactor_pointull()
/bf_pointull()
are nowbayesfactor_pointnull()
/bf_pointnull()
(can you spot the difference? #363 ).
bayestestR 0.8.0
CRAN release: 2020-12-05
New functions
-
sexit()
, a function for sequential effect existence and significance testing (SEXIT).
General
Added startup-message to warn users that default ci-width might change in a future update.
Added support for mcmc.list objects.
Bug fixes
unupdate()
gains anewdata
argument to work withbrmsfit_multiple
models.Fixed issue in Bayes factor vignette (don’t evaluate code chunks if packages not available).
bayestestR 0.7.5
CRAN release: 2020-10-22
New functions
Added
as.matrix()
function forbayesfactor_model
arrays.unupdate()
, a utility function to get Bayesian models un-fitted from the data, representing the priors only.
Bug fixes
Fixed issue with default rope range for
BayesFactor
models.Fixed issue in collinearity-check for
rope()
for models with less than two parameters.Fixed issue in print-method for
mediation()
withstanmvreg
-models, which displays the wrong name for the response-value.Fixed issue in
effective_sample()
for models with only one parameter.rope_range()
forBayesFactor
models returns non-NA
values ( #343 )
bayestestR 0.7.2
CRAN release: 2020-07-20
New functions
-
mediation()
, to compute average direct and average causal mediation effects of multivariate response models (brmsfit
,stanmvreg
).
Bug fixes
-
bayesfactor_parameters()
works withR<3.6.0
.
bayestestR 0.7.0
CRAN release: 2020-06-19
Changes to functions
weighted_posteriors()
can now be used with data frames.Revised
print()
fordescribe_posterior()
.Improved value formatting for Bayesfactor functions.
Bug fixes
Link transformation are now taken into account for
emmeans
objets. E.g., indescribe_posterior()
.Fix
diagnostic_posterior()
when algorithm is not “sampling”.Minor revisions to some documentations.
Fix CRAN check issues for win-old-release.
bayestestR 0.6.0
CRAN release: 2020-04-20
Changes to functions
describe_posterior()
now also works oneffectsize::standardize_posteriors()
.p_significance()
now also works onparameters::simulate_model()
.rope_range()
supports more (frequentis) models.
Bug fixes
Fixed issue with
plot()
data.frame
-methods ofp_direction()
andequivalence_test()
.Fix check issues for forthcoming insight-update.
bayestestR 0.5.3
CRAN release: 2020-03-26
Changes to functions
-
estimate_density()
now also works on grouped data frames.
Bug fixes
Fixed bug in
weighted_posteriors()
to properly weight Intercept-onlyBFBayesFactor
models.Fixed bug in
weighted_posteriors()
when models have very low posterior probability ( #286 ).Fixed bug in
describe_posterior()
,rope()
andequivalence_test()
for brmsfit models with monotonic effect.Fixed issues related to latest changes in
as.data.frame.brmsfit()
from the brms package.
bayestestR 0.5.0
CRAN release: 2020-01-18
General
Added
p_pointnull()
as an alias top_MAP()
.Added
si()
function to compute support intervals.Added
weighted_posteriors()
for generating posterior samples averaged across models.Added
plot()
-method forp_significance()
.p_significance()
now also works for brmsfit-objects.estimate_density()
now also works for MCMCglmm-objects.equivalence_test()
getseffects
andcomponent
arguments for stanreg and brmsfit models, to print specific model components.Support for mcmc objects (package coda)
Provide more distributions via
distribution()
.Added
distribution_tweedie()
.Better handling of
stanmvreg
models fordescribe_posterior()
,diagnostic_posterior()
anddescribe_prior()
.
Breaking changes
point_estimate()
: argumentcentrality
default value changed from ‘median’ to ‘all’.p_rope()
, previously as exploratory index, was renamed asmhdior()
(for Max HDI inside/outside ROPE), asp_rope()
will refer torope(..., ci = 1)
( #258 )
Bug fixes
Fixed mistake in description of
p_significance()
.Fixed error when computing BFs with
emmGrid
based on some non-linear models ( #260 ).Fixed wrong output for percentage-values in
print.equivalence_test()
.Fixed issue in
describe_posterior()
forBFBayesFactor
-objects with more than one model.
bayestestR 0.4.0
CRAN release: 2019-10-20
New functions / features
convert_bayesian_to_frequentist()
Convert (refit) Bayesian model as frequentistdistribution_binomial()
for perfect binomial distributionssimulate_ttest()
Simulate data with a mean differencesimulate_correlation()
Simulate correlated datasetsp_significance()
Compute the probability of Practical Significance (ps)overlap()
Compute overlap between two empirical distributionsestimate_density()
:method = "mixture"
argument added for mixture density estimation
Bug fixes
- Fixed bug in
simulate_prior()
for stanreg-models whenautoscale
was set toFALSE
bayestestR 0.3.0
CRAN release: 2019-09-22
General
- revised
print()
-methods for functions likerope()
,p_direction()
,describe_posterior()
etc., in particular for model objects with random effects and/or zero-inflation component
New functions / features
check_prior()
to check if prior is informativesimulate_prior()
to simulate model’s priors as distributionsdistribution_gamma()
to generate a (near-perfect or random) Gamma distributioncontr.bayes
function for orthogonal factor coding (implementation from Singmann & Gronau’sbfrms
, used for proper prior estimation when factor have 3 levels or more. See Bayes factor vignette ## Changes to functionsAdded support for
sim
,sim.merMod
(fromarm::sim()
) andMCMCglmm
-objects to many functions (likehdi()
,ci()
,eti()
,rope()
,p_direction()
,point_estimate()
, …)describe_posterior()
gets aneffects
andcomponent
argument, to include the description of posterior samples from random effects and/or zero-inflation component.More user-friendly warning for non-supported models in
bayesfactor()
-methods
Bug fixes
Fixed bug in
bayesfactor_inclusion()
where the same interaction sometimes appeared more than once (#223)Fixed bug in
describe_posterior()
for stanreg models fitted with fullrank-algorithm
bayestestR 0.2.5
CRAN release: 2019-08-06
Breaking changes
rope_range()
for binomial model has now a different default (-.18; .18 ; instead of -.055; .055)rope()
: returns a proportion (between 0 and 1) instead of a value between 0 and 100p_direction()
: returns a proportion (between 0.5 and 1) instead of a value between 50 and 100 (#168)bayesfactor_savagedickey()
:hypothesis
argument replaced bynull
as part of the newbayesfactor_parameters()
function
New functions / features
density_at()
,p_map()
andmap_estimate()
:method
argument addedrope()
:ci_method
argument addedeti()
: Computes equal-tailed intervalsreshape_ci()
: Reshape CIs between wide/longbayesfactor_parameters()
: New function, replacingbayesfactor_savagedickey()
, allows for computing Bayes factors against a point-null or an interval-nullbayesfactor_restricted()
: Function for computing Bayes factors for order restricted models
Bug fixes
-
bayesfactor_inclusion()
now works withR < 3.6
.
bayestestR 0.2.2
CRAN release: 2019-06-20
Breaking changes
equivalence_test()
: returns capitalized output (e.g.,Rejected
instead ofrejected
)describe_posterior.numeric()
:dispersion
defaults toFALSE
for consistency with the other methods
New functions / features
pd_to_p()
andp_to_pd()
: Functions to convert between probability of direction (pd) and p-valueSupport of
emmGrid
objects:ci()
,rope()
,bayesfactor_savagedickey()
,describe_posterior()
, …
Bug fixes
describe_posterior()
: Fixed column order restorationbayesfactor_inclusion()
: Inclusion BFs for matched models are more inline with JASP results.
bayestestR 0.2.0
CRAN release: 2019-05-29
Breaking changes
plotting functions now require the installation of the
see
packageestimate
argument name indescribe_posterior()
andpoint_estimate()
changed tocentrality
hdi()
,ci()
,rope()
andequivalence_test()
defaultci
to0.89
rnorm_perfect()
deprecated in favour ofdistribution_normal()
map_estimate()
now returns a single value instead of a dataframe and thedensity
parameter has been removed. The MAP density value is now accessible viaattributes(map_output)$MAP_density
New functions / features
describe_posterior()
,describe_prior()
,diagnostic_posterior()
: added wrapper functionpoint_estimate()
added function to compute point estimatesp_direction()
: new argumentmethod
to compute pd based on AUCarea_under_curve()
: compute AUCdistribution()
functions have been addedbayesfactor_savagedickey()
,bayesfactor_models()
andbayesfactor_inclusion()
functions has been addedStarted adding plotting methods (currently in the
see
package) forp_direction()
andhdi()
probability_at()
as alias fordensity_at()
effective_sample()
to return the effective sample size of Stan-modelsmcse()
to return the Monte Carlo standard error of Stan-models
Minor changes
Improved documentation
Improved testing
p_direction()
: improved printingrope()
for model-objects now returns the HDI values for all parameters as attribute in a consistent wayChanges legend-labels in
plot.equivalence_test()
to align plots with the output of theprint()
-method (#78)
bayestestR 0.1.0
CRAN release: 2019-04-08
CRAN initial publication and 0.1.0 release
Added a
NEWS.md
file to track changes to the package