Skip to contents

Confidence intervals from a fast AME bootstrap

Source: R/ame_als_bootstrap.R
confint.boot_ame.Rd

Bootstrap confidence intervals for the intercept and regression coefficients of a fast AME fit.

Usage

# S3 method for class 'boot_ame'
confint(
  object,
  parm = NULL,
  level = 0.95,
  ci_type = c("percentile", "basic"),
  which = c("all", "beta", "vc", "a", "b", "U", "V"),
  ...
)

Arguments

object

a boot_ame object.

parm

character vector of parameter names, or integer indices. If NULL (default), all coefficients are returned.

level

confidence level (default 0.95).

ci_type

interval type: "percentile" (default) uses the replicate quantiles directly and matches the intervals reported by summary() and tidy() on the same fit; "basic" reflects the replicate quantiles about the point estimate (\(2\hat\theta - q\)), which corrects first-order bias and can be preferable for the IRLS binary/Poisson point estimators on small samples.

which

which uncertainty channel to return: "all" (default, intercept + regression coefficients + variance components + sender effects

  • receiver effects + multiplicative U/V), or a single channel: "beta", "vc", "a", "b", "U", "V".

...

ignored.

Value

A matrix with one row per parameter and lower/upper bound columns.