glmlasso: Lasso regression for logistics model

In lcw68/G3proj: BIOS 735 Project

Description

Lasso regression for logistics model

Usage

glmlasso(X, y, lambda = 0.1, tol = 1e-06, iter = 100)

Arguments

X	Design Matrix
y	binary outcome
lambda	Lasso penalty parameter
tol	convergence threshold
iter	iteration times

Value

Lasso estimate of coefficients

Examples

set.seed(1232)
Nz = 500
pz = 10
Xz = scale(matrix(rnorm(Nz*pz), ncol=pz))
bz = c(.5, -.5, .25, -.25, .125, -.125, rep(0, pz-6))
yz = rbinom(Nz,1,exp(Xz %*% bz)/(1+exp(Xz %*% bz)))
lambda = .1
require(glmnet)
fit <- glmnet(Xz,yz,family="binomial",lambda = 0.1,intercept = FALSE)
coef(fit,s = 0.1)
fit1 <- glmlasso(Xz,yz,lambda,tol=1e-12)

lcw68/G3proj documentation built on Dec. 21, 2021, 9:46 a.m.