runPCPR2: The Principal Component R-squared method (PC-PR2)

In JoeRothwell/pcpr2: The Principal Component Partial R-squared method (PC-PR2)

Description

Runs the PC-PR2 method on X, an omics matrix of intensities, and Z, the subject metadata to be assessed in the model.

Usage

runPCPR2(X, Z, pct.threshold = 0.8)

Arguments

X	A matrix of omics data. It is recommended that the data be appropriately scaled and log transformed if necessary.
Z	A data frame of subject metadata, with the same number of rows as X, whose influence on the omics data is to be assessed. Categorical variables should be coded as factors.
pct.threshold	The proportion of variability desired to be explained. Defaults to 0.8.

Value

Returns a list with 6 elements:

dimensions	dimensions of the -omics matrix
types	data types for each Z-variable
threshold	the chosen variability to be explained
pcn	the number of principal components used for the chosen threshold
anovaFit	ANOVA output for each PC
pR2	Rpartial2 values for each Z-variable and the overall R2

Examples

output <- runPCPR2(transcripts, Z_metadata)
output$pR2

JoeRothwell/pcpr2 documentation built on June 17, 2021, 8:58 a.m.