boot_grf: Simple clustered bootstrap.

In grf: Generalized Random Forests

Simple clustered bootstrap.

Description

Inspired by the 'boot' function in the bootstrap package with clusters + half-sampling added. A future TODO could be to add parallel (not necessarily worth it) https://stat.ethz.ch/R-manual/R-devel/library/parallel/doc/parallel.pdf

Usage

boot_grf(data, statistic, R, clusters, half.sample = TRUE, ...)

Arguments

data	A data frame with the original data.
statistic	A function computing estimate(s) with signature (data, indices, ...) where data is the original data, and indices a vector which defines the bootstrap sample.
R	The number of bootstrap replications.
clusters	Integer vector of cluster assignment, setting to 1:N corresponds to an ordinary unclustered bootstrap.
half.sample	Whether to do half sample bootstrap (half the clusters are drawn). Default is TRUE.
...	Additional arguments passed on to statistic.

Value

A list with the original estimate t0, and bootstrap estimates t.

References

Angelo Canty and Brian Ripley (2021). boot: Bootstrap R (S-Plus) Functions.

grf documentation built on Oct. 1, 2023, 1:07 a.m.