prox_nuclear_func: proximal function with respect to low-rank component

In kevinbai92/LSvarEstimate: Low Rank Plus Sparse VAR Model Estimation

Description

proximal function with respect to low-rank component

Usage

prox_nuclear_func(w1, y, A, b, L, lambda, AtA, Atb)

Arguments

w1	the previous proximal vector
y	alternative coefficient vector
A	design matrix in linear regression
b	target vector in linear regression
L	learning rate
lambda	tuning parameter for low-rank penalty, i.e. nuclear penalty
AtA	crossproduct of A
Atb	crossproduct of A and b

Details

This function is to find the proximal function for low rank component, and update the corresponding coefficient.

Value

value of proximal function

Examples

set.seed(100)
w1 <- matrix(stats::rnorm(100), 10, 10)
y <- matrix(stats::rnorm(100), 10, 10)
A <- matrix(stats::rnorm(200), 20, 10)
b <- matrix(stats::rnorm(200), 20, 10)
L <- 0.1
lambda <- 0.5
AtA <- t(A) %*% A
Atb <- t(A) %*% b
prox_nuclear_func(w1, y, A, b, L, lambda, AtA, Atb)

kevinbai92/LSvarEstimate documentation built on May 8, 2020, 1:04 a.m.