perm4PLSC: Permutation for PLSC (as implemented in...

In HerveAbdi/data4PCCAR: Some data sets and R-functions for PCA and CA to accompany Abdi & Beaton (to appear, 2023) Principal Component and Correspondence Analysis with R

Permutation for PLSC (as implemented in TExPosition::tepPLS)

Description

perm4PLSC: Permutation for PLSC (as implemented in TExPosition::tepPLS). Computes an omnibus permutation test and specific tests for the eigenvalues when performing a PLSC from 2 matrices X and Y. Several possible combinations of centering and normalizing are possible (see paramater scale1, scale2, center2, scale2). Used for functions related to PLSC / inter-battery analysis / co-inertia... The different types of normalization are based on the ExPosition::expo.scale function. Two different permutation schemes are currently available (see parameter permType).

Usage

perm4PLSC(
  DATA1,
  DATA2,
  center1 = TRUE,
  center2 = TRUE,
  scale1 = "ss1",
  scale2 = "ss1",
  nIter = 1000,
  permType = "byMat",
  compact = FALSE
)

Arguments

DATA1	an N*I matrix of quantitative data
DATA2	an N*J matrix of quantitative data
center1	when TRUE (default) DATA1 will be centered
center2	when TRUE (default) DATA2 will be centered
scale1	when TRUE (default) DATA1 will be normalized. Depends upon ExPosition function expo.scale whose description is: boolean, text, or (numeric) vector. If boolean or vector, it works just as scale. The following text options are available: 'z': z-score normalization, 'sd': standard deviation normalization, 'rms': root mean square normalization, 'ss1': sum of squares (of columns) equals 1 (i.e., column vector of length of 1).
scale2	when TRUE (default) DATA2 will be normalized (same options as for scale1).
nIter	(Default = 1000). Number of Iterations (i.e., number of permuted samples computed).
permType	what type of permutation is used if 'byMat' (default) only the labels of the observations are permuted, other option is 'byColumns' then all columns of each matrix are independently permuted.
compact	if TRUE (Default) return only p-values for omnibus test.

Value

a list with fixedInertia: the inertia of X'Y data matrix (i.e., sums of squares) fixedEigenvalues: the eigenvalues of the X'Y matrix; pOmnibus: the probability associated to the inertia. If compact is FALSE, return also permInertia: an nIter * 1 vector containing the permuted inertia; pEigenvalues: The probabilites associated to each eigenvalue; If compact is FALSE, returns also permEigenvalues: an nIter * L matrix giving the permuted eigenvalues.

Author(s)

Hervé Abdi

See Also

compS

HerveAbdi/data4PCCAR documentation built on Sept. 11, 2022, 4:19 p.m.