brms R package for Bayesian generalized multivariate non-linear multilevel models using Stan
cox
. (#230, #962)loo_moment_match
, which can be used to
update a loo
object when Pareto k estimates are large.sample_new_levels = "uncertainty"
. (#956)id
in function mo
to ensure conditionally monotonic effects. (#924)rtdists
as additional backend of wiener
distribution functions thanks to the help of Henrik Singmann. (#385)constant
priors on some coefficients thanks to Frank Weber. (#919)conditional_effects
occuring for categorical
models with matrix predictors thanks to Jamie Cranston. (#933)rate
addition term so that it also
affects the shape
parameter in negbinomial
models thanks to
Edward Abraham. (#915)threshold
in ordinal family functions thanks to the help of
Marta Kołczyńska.posterior_linpred
as method in conditional_effects
.std_normal
in the Stan code for improved efficiency.cor
, id
, and cov
to the functions gr
and
mm
for easy specification of group-level correlation structures.int_conditions
in conditional_effects
to
work for all predictors not just interactions.data2
in
brm_multiple
. (#886)emmeans
package thanks to the help
of Russell V. Lenth. (#418)stanvar
using the position
argument.me
terms
thanks to Chris Chatham. (#855, #856)std_normal
in set_prior
thanks to Ben Goodrich. (#867)weibull
, frechet
,
or inverse.gaussian
families thanks to Brian Huey and Jack Caster. (#879)gp
for increased efficiency.parse_bf
to brmsterms
and deprecate the former function.extract_draws
to prepare_predictions
and deprecate
the former function.rescor
default.cov_ranef
in brm
and related functions.prior
argument. (#783)sigma
in combination with fixed correlation matrices
via autocorrelation term fcor
.data2
in brm
and related functions to pass
data objects which cannot be passed via data
. The usage of data2
will be extended in future versions.log_lik
for
non-factorizable Student-t models. (#705)posterior_predict
for multinomial
models
thanks to Ivan Ukhov.re_formula
in
multivariate models thanks to Maxime Dahirel. (#834)re_formula
thanks to @ferberkl. (#844)posterior_predict
again thanks to Mattew Kay. (#838)NA
values more consistently in posterior_table
thanks to Anna Hake. (#845)offset
variables to offsets
in the generated Stan
code as the former will be reserved in the new stanc3 compiler.loo
package.summary
output. (#824)newdata
thanks to Andrew Milne. (#830)resp_thres
. (#675)loo_subsample
for performing approximate
leave-one-out cross-validation for large data.add_criterion
. (#793)sample_new_levels = "uncertainty"
thanks to Dominic Magirr. (#779)pp_check
on
censored models thanks to Andrew Milne. (#744)zero_inflated_binomial
models thanks to Raoul Wolf. (#756)subset
thanks to Ruben Arslan.reloo
or kfold
with CAR models.fitted(..., scale = "linear")
with
multinomial models thanks to Santiago Olivella. (#770)as.mcmc
method for thinned models
thanks to @hoxo-m. (#811)marginal_effects
to conditional_effects
and
marginal_smooths
to conditional_smooths
. (#735)stanplot
to mcmc_plot
.pp_expect
as an alias of fitted
. (#644)add_criterion
are now
stored in the brmsfit$criteria
slot.resp_cat
in favor of resp_thres
.model_weights
.intercept
in favor of Intercept
.exact_match
in favor of fixed
.add_loo
and add_waic
in favor of add_criterion
.summary
output. (#712)vreal
and vint
. (#707)cor_cosy
. (#403)sigma
in combination with several
autocorrelation structures. (#403)rate
to conveniently handle
denominators of rate responses in log-linear models.cor_car
thanks to the case study
and help of Mitzi Morris.marginal_effects
if not specified otherwise. (#718)me
terms with
grouping factors thanks to the GitHub user tatters. (#706)horseshoe
prior in categorical and
related models thanks to the Github user tatters. (#678)prior_samples
thanks to Jonas Kristoffer Lindelov. (#696)marginal_smooths
thanks to Gavin Simpson. (#740)softplus
link function in various families. (#622)decomp
of brmsformula
thanks to the help of Ben Goodrich. (#640)sparse
separately for each model formula.bayes_R2
and loo_R2
with ordinal models. (#639)cor_arma
in non-normal models. (#648)cor_arr
and cor_bsts
correlation structures after a year of deprecation.marginal_effects
to measurement error models thanks to Jonathan A. Nations. (#636)marginal_effects
.brm_multiple
without sampling thanks to Will Petry. (#671)multinomial
. (#463)dirichlet
. (#463)categorical
and
multinomial
families together with non-linear formula syntax. (#560)categorical
and related
families via argument refcat
of the corresponding family functions.subset
. (#360)center
of brmsformula
and related functions.update
method for brmsfit_multiple
objects. (#615)group
in the kfold
method. (#619)compare_ic
and instead recommend loo_compare
for the
comparison of loo
objects to ensure consistency between packages. (#414)mvbind
to eventually replace cbind
in the formula syntax of multivariate models.brm
before compiling the Stan model. (#576)get_y
which is used to extract response
values from brmsfit
objects.re_formula
in bayes_R2
thanks to the GitHub user emieldl. (#592)resp
of marginal_effects
in
univariate models thanks to Vassilis Kehayas. (#589)ndt
in drift diffusion models.kfold
thanks to
the GitHub user gcolitti. (#602)VarCorr
method to
meta-analytic models thanks to Michael Scharkow. (#616)gp
. (#540)brm_multiple
via the future package. (#364)kfold_predict
. (#468)oos
of extract_draws
. (#539)marginal_effects
more robust to the usage of non-standard variable names.fitted(..., scale = "linear")
with ordinal models thanks to Andrew Milne. (#557)marginal_smooths
with ordinal models thanks to Andrew Milne. (#570)me
terms thanks to the GitHub user hlluik. (#571)warmup
samples when using update.brmsfit
.