logitCdaC: Optimize a logistic regression model by coordinate descent...

In iilasso: Independently Interpretable Lasso

Description

Optimize a logistic regression model by coordinate descent algorithm using a design matrix

Usage

logitCdaC(X_tilde, y, lambda, R, init_beta, delta = 0, maxit = 10000,
  eps = 1e-04, warm = "lambda", strong = TRUE)

Arguments

X_tilde	standardized matrix of explanatory variables
y	vector of objective variable
lambda	lambda sequence
R	matrix using exclusive penalty term
init_beta	initial values of beta
delta	ratio of regularization between l1 and exclusive penalty terms
maxit	max iteration
eps	convergence threshold for optimization
warm	warm start direction: "lambda" (default) or "delta"
strong	whether use strong screening or not

Value

standardized beta

iilasso documentation built on May 2, 2019, 10:14 a.m.