Perform model fit checks for brms models
model_fit(brmsfit, Group = NULL, Prior = FALSE)
The output of a brms model. You can use VarDecomp::brms_model() to produce a brmsfit.
A string containing the name of a grouping variable for the visualization of a posterior predictive check plot (e.g. "sex"). To add multiple grouping variables, use c() (e.g. c("sex", "species")).
A logical argument defining whether the brmsfit
contains prior samples. If set to TRUE
it will produce plots comparing the log distributions of priors and posterior samples for each covariate.
Returns a list containing (a) the maximum R-hat value, (b) the minimum effective sample size, (c) traceplots, (d) posterior predictive check plots, and (e) prior and posterior sample plots (if priors are available).
if (FALSE) { # \dontrun{
md = dplyr::starwars
# Centering variables
md = md %>%
dplyr::select(mass, sex, height, species) %>%
dplyr::mutate(mass = log(mass),
sex = dplyr::recode(sex, "male" = 1,
"female" = -1,
"hermaphroditic" = 0,
"none" = as.numeric(NA)))
# Without random effects
mod = brms_model(Chainset = 2,
Response = "mass",
FixedEffect = c("sex","height"),
Family = "gaussian",
Data = md,
PriorSamples = TRUE)
model_fit(mod, Group = "sex", Prior = TRUE)
} # }