lassoFDR: Knockoff Variable Selection through Lasso Regression while...

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

Description

Select variable based on knockoff constructed from lasso Regression

Usage

lassoFDR(x, x.knockoff, y, mask = NULL, family, offset = 0, fdr_rate = 0.1)

Arguments

x	Original data
x.knockoff	knockoff copy of the original data
y	Response variable
mask	The variables that are masked from the purpose
family	what kind of
offset	Offset for the knock.thershold, either 0 or 1, default is 0
fdr_rate	Controlling the False Discovery Rate of the knockoff

Examples

out_copula = gaussianCopula(p = 100, n = 500, type = "diag")
out_data = gaussianDistributionGeneration(copula = out_copula, p = 100, n = 500)
out = copulaKnockoff(out_data$data)
lasso_out = LassoGenerationModel(data_signal = out_data$data, p_freq = 0.3)
lassoFDR(x = out_data$data, x.knockoff = out$x.knockoff.copy, y = lasso_out$y, family = "gaussian")

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