admmgrouplasso_log: Logistic Loss with group LASSO Penalty

In stanmcgill/680: Linear and Logistic Regression Estimation Using ADMM Algorithm.

Description

Compute the group-LASSO penalized minimum logistic loss solutions

Usage

admmgrouplasso_log(X, y, group, lam, rho = 0.001)

Arguments

X	An n by p design matrix
y	The response vector with n entries
group	The indicator vetor for grouping with n entries
lam	The non-negative tuning parameter value
rho	The penalty parameter in the augmented Lagrangian

Value

This function returns a list with two elements:

beta	The vector of regression coefficient estimates, with p-1 entries
total.iterations	Number of iterations made

Examples

## Not run: 
admmgrouplasso_log(X=X[,-group], y=y, group=X[,group], lam=lam, rho=1e-3)
## End(Not run)

stanmcgill/680 documentation built on Sept. 16, 2019, 2:49 p.m.