Skip to contents

Delete-d (generalised) jackknife

Source: R/bootstrap_methods.R
delete_d_jackknife.Rd

Generalised jackknife removing d observations per replicate; all subsets are enumerated when \(C(n, d) \le\) max_subsets and Monte-Carlo sampled otherwise. resample exposes the equivalent generalised jackknife API via resample::jackknife (cross-referenced).

Usage

delete_d_jackknife(
  data,
  statistic,
  d = 2L,
  ci_level = 0.95,
  max_subsets = 5000L,
  seed = 42L
)

Arguments

data

Numeric vector or matrix.

statistic

Function returning a scalar.

d

Number of observations to delete per replicate.

ci_level

Confidence level.

max_subsets

Maximum subsets to evaluate.

seed

Random seed.

Value

A morie_jackknife_result.

See also

resample::jackknife.

Examples

if (requireNamespace("coin", quietly = TRUE)) {
  set.seed(1)
  x <- rnorm(8)
  r <- delete_d_jackknife(x, mean, d = 2L)
  r$estimate
}
#> [1] 0.1314544