In hsong1/GTV: Graph-based regularization for regression problems with alignment and highly-correlated designs

knitr::opts_chunk$set(echo = TRUE)

Regression with Graph-based Total-Variation Regularization

Description

Implementation of regression with graph-based regularization. See Li et al. (2019)

Installation

Install using devtools package:

devtools::install_github("hsong1/GTV")

Example

Linear regression with simulated data at particular lambdas, e.g. $\lambda_{TV} = 0.1, \lambda_S = 0.1, \lambda_1 = 0.1$.

library(GTV)
data(exampleGTV)
fit0 = with(exampleGTV, gtv(X = X,y = y,Sigma = Sigma,lam_TV = 0.1,lam_S = 0.1,lam_1 = 0.1))
head(coef(fit0))

Linear regression with simulated data where we choose lambdas via 5-fold cross-validation.

set.seed(1234)
fit1 = with(exampleGTV, cv.gtv(X = X,y = y,Sigma = Sigma,parallel = T))
head(coef(fit1))

hsong1/GTV documentation built on Nov. 9, 2019, 2:51 a.m.