Parametric bootstrap inference for a VCMoE fit

Parametric bootstrap inference for a VCMoE fit

Usage

vcmoe_bootstrap(
  fit,
  data,
  u = NULL,
  B = 200L,
  coefficient_set = c("expert", "gating"),
  seed = NULL,
  control = list(),
  min_successful = max(20L, ceiling(0.5 * B)),
  keep_fits = FALSE,
  verbose = FALSE
)

Arguments

fit

A vcmoe fit with k = 2:10.

data

Original data frame used to fit fit.

u

Optional original u values or column name.

B

Number of bootstrap replicates.

coefficient_set

Coefficient sets to store.

seed

Optional random seed.

control

Control overrides for bootstrap refits.

min_successful

Minimum successful replicates for reliable inference.

keep_fits

Whether to store successful bootstrap fit objects.

verbose

Whether to message progress.

Value

An object of class vcmoe_bootstrap.

Details

Bootstrap refits preserve the reference fitting engine. A joint-path reference is therefore refitted with vcmoe_fit(..., engine = "joint_path_em") rather than silently falling back to local-grid EM.