signCorrectionPCA: Sign correction for principal component analysis

In oliviermfmartin/HelpingHand: A helping hand

Description

Corrects the sign ambiguity that exists in PCA by correlating scores with the data.

Usage

signCorrectionPCA(data, loadings, scores)

Arguments

data	matrix; The data matrix with samples in the rows and variables in the columns.
loadings	matrix; The loading matrix with variables in the rows and components in the columns.
scores	matrix; The score matrix with samples in the rows and components in the columns.

Details

The sign of the jth component is defined as: sum(cor(scores[, j], data) * loadings[, j]**2))

Value

A list containing the sign corrected loadings and scores.

See Also

prcomp

oliviermfmartin/HelpingHand documentation built on Oct. 10, 2020, 5:59 a.m.