X_decor {armada}R Documentation

Decorrelation of a matrix X, given a response variable Y.

Description

Decorrelation of a matrix X, given a response variable Y.

Usage

X_decor(X, Y, tree = NULL, nclust = 1, maxnbfactors = 10)

Arguments

X

the matrix (or data.frame) of covariates, dimension n*p (n is the sample size, p the number of covariates). X must have colnames and rownames.

Y

the vector of the response, length n.

tree

the dendrogram of the covariates (object obtained before by the function clustering). By default, tree=NULL.

nclust

integer, the number of clusters in the covariates (1 by default).

maxnbfactors

integer, the maximum number of factors in the clusters. By default: maxnbfactors=10.

Details

The function X_decor applies the factor analysis method FAMT in the different clusters of variables. The clusters must have been defined before (with the function "clustering").

Value

a matrix X.decorrele, with the same dimension, same rownames and same colnames than X.

Examples

toys.data
X<-toys.data$x
Y<-toys.data$Y
Tree <- clustering(X,plot=FALSE)
nclust <- Tree[[2]]
tree <- Tree[[1]]
library(ClustOfVar)
library(FAMT)
X.deco<- X_decor(X, Y, tree, nclust, maxnbfactors=10)

[Package armada version 0.1.0 Index]