daarem.lasso: NIDAAREM acceleration of penalized regression

In nchenderson/nidaarem: Nesterov Initialized Damped Anderson Acceleration with Epsilon Monotonicity

Description

Uses the NIDAAREM procedure to accelerate proximal gradient algorithms for fitting l1-penalized regression.

Usage

daarem.lasso(par, X, y, lambda, stplngth = NULL, nesterov.init = FALSE, family = c("gaussian", "binomial"), control = list())

Arguments

par	initial value of the parameters
X	the n x p design matrix
y	vector of responses (should have length n)
lambda	the penalty term
stplngth	The step length used in the proximal gradient scheme.
nesterov.init	A logical indicator. If true, the NIDAAREM algorithm is used. If false, the DAAREM algorithm is used.
family	The link function used in the regression. Can be either gaussian or binomial.
control	A list of additional control parameters.

Examples

a = 2

nchenderson/nidaarem documentation built on Feb. 19, 2020, 12:45 p.m.