pathCalc: Select Path of Regularization Parameters

In kouroshz/creNet: Fit a GLM with a combination of lasso and overlapping group lasso regularization

Usage

pathCalc(data, index, weights, alpha = 0.95, min.frac = 0.05, nlam = 100, type)

Arguments

data	list with components $x$ and $y$
index	group membership indicator
weights	group weights for regularization penalty
alpha	tradeoff between lasso penalty and group lasso penalty. alpha=1 corresponds to pure lasso, alpha=0 to pure group lasso.
min.frac	minimum regularization parameter lambda as a fraction of the largest
nlam	number of lambda values along the path
type	type of regression model: linear or logit

Value

A vector of nlam lambda values

Author(s)

David Degras

kouroshz/creNet documentation built on May 20, 2019, 1:11 p.m.