combinationsTest: Significant combinations of variables

In ddgraph: Distinguish direct and indirect interactions with Graphical Modelling

Description

Calculate which combinations of values of variables are significantly different in the two classes (only for binary data). This function takes an DDDataSet and a number of variables and finds those combinations of values of those variables that have significantly different frequencies in the two class labels.

Usage

combinationsTest(obj, selected.vars, cutoff=0.05, p.adjust.method="none", verbose=TRUE)

Arguments

obj	DDDataSet object
selected.vars	indexes or names of variables selected for the test
cutoff	the p value cutoff for reporting (default: 0.05)
p.adjust.method	the multiple adjustment method (default: none)
verbose	if to print progress output and additional information

Value

data.frame with ordered combinatorial patterns of selected variables

Examples

data(mesoBin)
# find significant differences at 0.2 FDR
combinationsTest(mesoBin$Meso, c("Twi 2-4h", "Tin 6-8h", "Mef2 6-8h"), 0.2, "fdr")

ddgraph documentation built on Nov. 17, 2017, 10:50 a.m.