brms R package for Bayesian generalized multivariate non-linear multilevel models using Stan
pathfinder
and laplace
algorithms
in the cmdstanr
backend. (#1591)constant
priors.kfold
via argument joint
.make_stancode
and make_standata
to be aliases of stancode
and
standata
, respectively. Change get_prior
to be an alias of a new generic
method default_prior
. This enable other packages to define new stancode
,
standata
and default_prior
methods to generate Stan code and data, and extract
the default priors, for their own objects building on brms. Thanks to Ven Popov
for helping with this. (#1604)shape
parameter of negbinomial
models
to inv_gamma(0.4, 0.3)
thanks to Aki Vehtari. (#1614)read_csv_as_stanfit
thanks to Ven Popov. (#1619)shinystan
optional. This means that the package has to
be loaded, via library(shinystan)
, before launch_shinystan
can be used. (#1595)summary
output.plot
method by default.N
in the plot
method in favor of argument nvariables
.exact_loo
in method kfold
.combine_models
with moment matching. (#1603)splines2
package version. (#1580)rmulti_normal
thanks to Ven Popov. (#1588)kfold
or reloo
in parallel.horseshoe
and R2D2
priors globally, that is, for all additive predictor terms specified in the same formula. (#1492)as.brmsprior
to transform objects into a brmsprior
. (#1491)lasso
prior as it is not a good shrinkage prior and incompatible with the newly implemented global shrinkage prior framework.newdata
of get_refmodel.brmsfit()
. (#1502)trials
argument after several years of deprecation. (#1501)restructure
. Special thanks to Simon Wood, Ruben Arslan, Marta Kołczyńska, Patrick Hogan, and Urs Kalbitzer. (#1465)unstr
term
thanks to the help of Sebastian Weber. (#1435)hurdle_cumulative
family thanks to Stephen Wild. (#1448)recompile
in post-processing methods that require a compiled Stan model.point_estimate
feature in prepare_predictions
via the new argument ndraws_point_estimate
.lasso
priors. (#1427)cauchit
or softplus
.newdata
than necessary. (#1457, #1459, #1460)s(..., fx = TRUE)
.update
and related methods. (#1373, #1378)drop_unused_levels = FALSE
in brm
and related functions. (#1346)update.brmsfit
. (#1380)dirichlet
priors for more parameter types. (#1165)backend = "cmdstanr"
to stanfit
objects thanks to Simon Mills and Jacob Socolar. (#1331)O1
optimization of brms-generated Stan models thanks to Aki Vehtari. (#1382)sdme
parameters in models with known response standard errors thanks to Solomon Kurz. (#1348)gamma
models with softplus
link.brm_multiple
. (#1383)control_params
returns the right values for models fitted with the cmdstanr
backend. (#1390)subset
addition term. (#1385)lb
and ub
arguments of set_prior
and related functions. (#878, #1094)logistic_normal
for simplex responses. (#1274)future_args
to kfold
and reloo
for additional control over parallel execution via futures.beta_binomial
& zero_inflated_beta_binomial
for potentially over-dispersed and zero-inflated binomial response models thanks to Hayden Rabel. (#1319 & #1311)ppd_*
plots in pp_check
via argument prefix
. (#1313)log
link in binomial and beta type families. (#1316)brms_seed
has been added to get_refmodel.brmsfit()
. (#1287)inits
in favor of init
for consistency with the Stan backends.summary
method for high-dimensional models. (#1330)int_conditions
in conditional_smooths
thanks to Urs Kalbitzer. (#1280)projpred
's K-fold CV. (#1286)make_standata
for bernoulli
families when only 1s are present thanks to Facundo Munoz. (#1298)pp_check
for censored responses to work for all plot types thanks to Hayden Rabel. (#1327)overwrite
in add_criterion
works as expected for all criteria thanks to Andrew Milne. (#1323)launch_shinystan
occurring when warmup draws were saved thanks to Frank Weber. (#1257, #1329)log_lik
for ordinal models. (#1192)posterior
package. (#1204)brms
models
with emmeans
thanks to Mattan S. Ben-Shachar. (#907, #1134)mi
) terms with subset
addition terms. (#1063)get_dpar
for use in the post-processing
of custom families thank to Martin Modrak. (#1131)squareplus
link function in all families and
distributional parameters that also allow for the log
link function.incl_thres
to posterior_linpred.brmsfit()
allowing to
subtract the threshold-excluding linear predictor from the thresholds in case
of an ordinal family. (#1137)"mock"
backend option to facilitate testing
thanks to Martin Modrak. (#1116)file_refit = "always"
to always overwrite models
stored via the file
argument. (#1151)robust
in method hypothesis
. (#1170)loop
of custom_family
. (#1084)cumulative
models. (#1060)regenerate
of method stancode
.expose_functions
for models fitted with the
cmdstanr
backend thanks to Sebastian Weber. (#1176)log_prob
and related functionality in models fitted
with the cmdstanr
backend via function add_rstan_model
. (#1184)cbind
to express multivariate models after
over two years of deprecation (please use mvbind
instead).posterior_linpred(transform = TRUE)
is now equal
to posterior_epred(dpar = "mu")
and no longer deprecated.NA
values in interval censored boundaries as long as
they are unused. (#1070)me
) terms in favor
of the more general and consistent missing value (mi
) terms. (#698)cox
models
thanks to Malcolm Gillies. (#1143)file_refit = "on_change"
if factor level
names have changed thanks to Martin Modrak. (#1128)validate_newdata
even when they are simultaneously
used as predictors and grouping variables thanks to Martin Modrak. (#1141)horseshoe
prior
thanks to Max Joseph. (#1167)normalize
to increase sampling efficiency thanks to Andrew Johnson. (#1017, #1053)posterior_predict
for truncated continuous models even if the required CDF or quantile functions are unavailable.validate_prior
to validate priors supplied by the user.rstan (Stan >= 2.25)
backend.R2D2
to be used in set_prior
.arma
correlation structures in non-normal families.data2
for use in the evaluation of most model terms.file_refit
. (#1058)brm
via the silent
argument. (#1076)stanvars
to alter distributional parameters. (#1061)stanvars
to be used inside threaded likelihoods. (#1111)sratio
and cratio
) thanks to Andrew Johnson. (#1087)multinomial
models with the cmdstanr
backend thanks to Andrew Johnson. (#1033):
operator in autocorrelation terms.wiener
drift diffusion models thanks to the GitHub user yanivabir. (#1085)by
variables thanks to Reece Willoughby. (#1081)emmeans
related methods thanks to Russell V. Lenth. (#1096)normalize
.
to increase sampling efficiency thanks to Andrew Johnson. (#1017, #1053)posterior_predict
for truncated continuous models
even if the required CDF or quantile functions are unavailable.validate_prior
to validate priors supplied by the user.rstan (Stan >= 2.25)
backend.R2D2
to be used in set_prior
.arma
correlation structures in non-normal families.data2
for use in the
evaluation of most model terms.sratio
and cratio
) thanks to Andrew Johnson. (#1087)multinomial
models with the
cmdstanr
backend thanks to Andrew Johnson. (#1033):
operator in autocorrelation terms.wiener
drift diffusion
models thanks to the GitHub user yanivabir. (#1085)by
variables thanks to Reece Willoughby. (#1081)reduce_sum
using argument threads
in brm
thanks to Sebastian Weber. (#892)fixed_param
to sample from fixed parameter values. (#973)NA
values in data
if there are unused because of
the subset
addition argument. (#895)by
variables and within-group correlation matrices
in group-level terms. (#674)robust
to the summary
method. (#976)posterior_predict
and log_lik
methods via argument cores
. (#819)kfold
.print
output
of brmsprior
objects. (#761)unused
of function brmsformula
.emmeans
via
dpar = "mean"
thanks to Russell V. Lenth. (#993)save_pars
and corresponding argument in brm
. (#746)posterior_smooths
to computing predictions
of individual smooth terms. (#738)conditional_effects
using the effects
argument. (#1012)probs
in the conditional_effects
method
in favor of argument prob
.pp_check
inducing wronger observation
orders in time series models thanks to Fiona Seaton. (#1007)loo_moment_match
that prevented
it from working for some more complex models.