README.md

In yang0117/TVsMiss: Variable Selection for Missing Data

TVsMiss

TVsMiss

The goal of TVsMiss is to select variable in both low and high demesion through regularization likelihood method.

Installation

You can install TVsMiss from github with:

# install.packages("devtools")
devtools::install_github("yang0117/TVsMiss")

Example

rm(list = ls())
library(TVsMiss)
n <- 50
p <- 8
beta <- c(3,0,1.5,0,2,rep(0,p-5))
xm <- matrix(rnorm(n*p),ncol = p, nrow = n)
y <- xm %*% beta + rnorm(n)
colnames(xm) <- paste0("Var_",1:p)
fit02 <- tvsmiss(x=xm,y=y,method = "BIC")
fit02$selection_beta
#>     Var_1     Var_2     Var_3     Var_4     Var_5     Var_6     Var_7 
#> 0.5614923 0.0000000 0.1417931 0.0000000 0.2872000 0.0000000 0.0000000 
#>     Var_8 
#> 0.0000000
plot(fit02,x.log=TRUE,label = TRUE)

yang0117/TVsMiss documentation built on July 14, 2020, 2:56 a.m.