coefboot.pffr: Simple bootstrap CIs for pffr

In refund: Regression with Functional Data

Simple bootstrap CIs for pffr

Description

This function resamples observations in the data set to obtain approximate CIs for different terms and coefficient functions that correct for the effects of dependency and heteroskedasticity of the residuals along the index of the functional response, i.e., it aims for correct inference if the residuals along the index of the functional response are not i.i.d.

Usage

coefboot.pffr(
  object,
  n1 = 100,
  n2 = 40,
  n3 = 20,
  B = 100,
  ncpus = getOption("boot.ncpus", 1),
  parallel = c("no", "multicore", "snow"),
  cl = NULL,
  conf = c(0.9, 0.95),
  type = "percent",
  method = c("resample", "residual", "residual.c"),
  showProgress = TRUE,
  ...
)

Arguments

object	a fitted pffr-model
n1	see coef.pffr
n2	see coef.pffr
n3	see coef.pffr
B	number of bootstrap replicates, defaults to (a measly) 100
ncpus	see boot. Defaults to getOption("boot.ncpus", 1L) (like boot).
parallel	see boot
cl	see boot
conf	desired levels of bootstrap CIs, defaults to 0.90 and 0.95
type	type of bootstrap interval, see boot.ci. Defaults to "percent" for percentile-based CIs.
method	either "resample" (default) to resample response trajectories, or "residual" to resample responses as fitted values plus residual trajectories or "residual.c" to resample responses as fitted values plus residual trajectories that are centered at zero for each gridpoint.
showProgress	TRUE/FALSE
...	not used

Value

a list with similar structure as the return value of coef.pffr, containing the original point estimates of the various terms along with their bootstrap CIs.

Author(s)

Fabian Scheipl

refund documentation built on Nov. 14, 2023, 5:07 p.m.