Skip to contents

Extract information from a brmsfit object

Source: R/utils.R
fit_info.Rd

Extract information from a brmsfit object

Usage

fit_info(fit, what)

Arguments

fit

A brmsfit object, or a list of brmsfit objects

what

String. What to return:

  • "time" for the sampling time per chain

  • "time_mean" for the mean sampling time

  • "inv_metric" for the adapted inverse metric per chain

  • "step_size" for the adapted step size per chain

Value

Depends on what and the class of fit. For brmsfit objects, information about the single fit is returned. For brmsfit_list objects, a list or data.frame with the information for each fit is returned.

  • "time": A data.frame with the warmup, sampling and total time per chain

  • "time_mean": A named numeric vector with the mean warmup, sampling and total time across chains

  • "inv_metric": A list with one numeric vector per chain (the diagonal), or one matrix per chain for metric = "dense_e"

  • "step_size": A list with one number per chain

Every chain of the fit is reported, also in a fit combined from several fits with brms::combine_models() or extended with add_chains(). NULL is returned for what the sampler did not record: anything for a fit without a stanfit or without chains, "time" and "time_mean" for variational and Laplace runs, "inv_metric" and "step_size" for any algorithm other than NUTS, and "inv_metric" for metric = "unit_e". In a fit combined from several fits, a unit_e chain next to chains with a metric has a NULL element in the "inv_metric" list. Values read from Stan's printed adaptation output (rstan fits, and fits combined from several fits) have six significant digits by default. A cmdstanr unit_e chain carries no printed step size, so in a fit combined from several fits its step size is that of its last draw.

Examples

if (FALSE) { # identical(Sys.getenv("BMM_EXAMPLES"), "true")
fit <- bmm(
  formula = bmmformula(c ~ 1, kappa ~ 1),
  data = data.frame(y = rsdm(1000)),
  model = sdm(resp_error = "y")
)

fit_info(fit, "time")
}