cluster_nhds: Generate null-hypothesis distributions for cluster mass...

In dalejbarr/clusterperm: Cluster-Permutation Tests for Time-Series Data from Factorial Experiments

Generate null-hypothesis distributions for cluster mass statistics

Description

Generate null-hypothesis distributions for cluster mass statistics

Usage

cluster_nhds(n, .data, bin, formula, fn, ...)

Arguments

n	Number of Monte Carlo runs
.data	Data frame.
bin	Unquoted name of variable identifying bins.
formula	Model formula, passed to aov_by_bin.
fn	Unquoted name of function to use for exchanging labels.
...	Arguments to be passed to fn.

Details

Generates null-hypothesis distributions by the following algorithm: For each of the n Monte Carlo runs, (1) randomly permute labels in the dataset according to the relabeling function fn; (2) run aov_by_bin on the resulting data, the result of which is passed to (3) detect_clusters_by_effect, and (4) store the maximum cluster mass statistic for each effect on each run.

Value

A data frame with null-hypothesis distributions for each effect in the column nhd.

dalejbarr/clusterperm documentation built on Oct. 29, 2022, 10:01 p.m.