extractInteraction: Extract interaction

In tdeenes/eegR: Analyze EEG Signals

Description

extractInteraction computes the highest-order (interaction) effect (i.e. the difference between differences) from means of each cell of the design matrix

Usage

extractInteraction(dat, sep = ".", sep_fixed = TRUE)

Arguments

dat	a named list with array elemens, see details below
sep	the character which separates the name of factor levels in the model terms (default: ".")
sep_fixed	logical indicating as sep should be treated "as is" while splitting the model term names

Details

This function is not intended for direct use, and it only works for specifically formatted data. The input data is supposed to be the return value of marginalMeans; a named list which contains the marginal means for each effect of an ANOVA. The names are the names of the factor(s) - for interaction effects the factor names are separated by ":". The list elements are three-dimensional arrays with the following dimension names: modelterm, chan and time. For pure main effects, this function computes all pairwise differences between the levels of the factor. For interaction effects, the function computes all pairwise differences at the level of the highest order interaction. That means for the A:B interaction where factor A has levels Aa & Ab and factor B has levels Ba, Bb and Bc, the function computes the following differences: (Aa-Ab) - (Ba-Bb), (Aa-Ab) - (Ba-Bc), (Aa-Ab) - (Bb-Bc).

Value

A named list is returned with the same names as the input, but the list elements are ia_level x chan x time arrays.

tdeenes/eegR documentation built on April 19, 2021, 4:17 p.m.