proximalOperators: Proximal operators

Description Usage Arguments Value

Description

Proximal operators for the lasso, ridge, group lasso and nuclear norm penalties.

Usage

1
2
3
4
5
6
7
proxL1(x, step, lambda)

proxL2(x, step, lambda)

proxE2(x, step, lambda)

proxNuclear(x, step, lambda, nrow, ncol)

Arguments

x

value of the coefficients

step

value of the step size. Should be between 0 and the inverse of the Lipschitw constant of grad f

lambda

value of the penalization parameter

nrow

the number of rows of the matrix of coefficients

ncol

the number of columns of the matrix of coefficients

Value

the updated vector of coefficients


bozenne/lava.penalty documentation built on May 13, 2019, 1:41 a.m.