ScreenData: Fast p-value computation for simple 2-sample or K-sample...

In markvdwiel/ShrinkBayes: Bayesian analysis of high-dimensional omics data, either Gaussian or counts

Fast p-value computation for simple 2-sample or K-sample designs.

Description

This function provides fast p-value combination for very large data sets based on t-tests, F-tests, Wilcoxon two-sample tests or Kruskal-Wallis tests. It may be used as an initial screen and in combination with FastScreenP to filter data and speed up computations.

Usage

ScreenData(dat, variable, np = TRUE, ncpus = 2, updateby = 10000)

Arguments

dat	Matrix or data frame containing the data. Rows are features, columns are samples.
variable	String. Name of the factor variable that contains the group information for each sample
np	Boolean. Should a nonparametric test be used or not?
ncpus	Integer. The number of cpus to use for parallel computations.
updateby	Integer. Show an update for each updateby number of features executed.

Details

If np=TRUE then a Wilcoxon two-sample test is performed when the number of levels of variable equals 2, and a Kruskal-Wallis test is performed when the number of levels exceeds 2. If np=FALSE, a two-sample t-test or F-test is used instead.

Value

Numerical vector containing the p-values.

Author(s)

Mark A. van de Wiel

See Also

wilcox.test, kruskal.test, t.test, anova.

Examples

# See FastScreenP

markvdwiel/ShrinkBayes documentation built on March 27, 2022, 7:47 p.m.