truncatedGaussianKnockoff: Truncated Gaussian knockoff

In CHuanSite/copulaKnockoff: Implementation of Gaussian Copula for Knockoff Variable Selection

Description

Construct Gaussian knockoff by truncating the diagonal variables

Usage

truncatedGaussianKnockoff(x, Sigma, amp = 1)

Arguments

x	original data
Sigma	covariance matrix of the original X
amp	The amplitude of the gaussian copula, default value is set to 1, upper limit is 2.

Value

knockoff generated

Examples

library(mvtnorm)
p = 20
n = 100
u = eigen(cov(matrix(rnorm(p * n), nrow = n)))$vectors
d = runif(p, 1, 10)
Sigma = u %*% diag(d) %*% t(u)
x = rmvnorm(n, rep(0, p), Sigma)
truncatedGaussianKnockoff(x, Sigma, 1)

CHuanSite/copulaKnockoff documentation built on Aug. 1, 2020, 12:50 p.m.