dSVA {dSVA}

direct surrogate variable analysis

Description

Identify hidden factors in high dimensional biomedical data

Usage

dSVA(Y, X, ncomp=0)


 

Arguments

Y

n x m data matrix of n samples and m features.

X

n x p matrix of covariates without intercept.

ncomp

a number of surrogate variables to be estimated. If ncomp=0 (default), ncomp will be estimated using the be method in the num.sv function of the sva package.

Value

Bhat = Bhat.all[idx.test,], BhatSE= BhatSE[idx.test,], Pvalue=Pvalue

Bhat

n x m matrix of the estimated effect sizes of X

BhatSE

n x m matrix of the estimated standard error of Bhat

Pvalue

n x m matrix of the p-values of Bhat

Z

a matrix of the estimated surrogate variable

ncomp

a number of surrogate variables.

Author(s)

Seunggeun Lee

Examples

data(Example)
attach(Example)
out<-dSVA(Y,X, ncomp=0)