Glmnet

From bwHPC Wiki
Jump to: navigation, search

1 General information

glmnet is a R library for lasso and elastic-net regularized generalized linear models

2 Installation instructions

Consider starting an interactive job for compiling. Copy and paste the following to your shell.

# Load the R software module, e.g.
module load math/R/4.1.2

# Prepare .R directory (if it does not already exists)
mkdir -p ~/.R

# Write the following environment variables to Makevars
# Skip the 2nd and 3rd commands below if Makevars already consist these variables (1st command shows content of Makevars)
cat ~/.R/Makevars
echo "CXX14=icpc" >> ~/.R/Makevars
echo "CXX14FLAGS=-O3 -fPIC -std=c++14 -axCORE-AVX512,CORE-AVX2,AVX -xSSE4.2 -fp-model strict -qopenmp" >> ~/.R/Makevars

# Install the glmnet package from within R session
R -q
> install.packages("glmnet", dependencies=TRUE)

# Run a quick test
> library(glmnet)
> data(QuickStartExample)
> x <- QuickStartExample$x
> y <- QuickStartExample$y
> fit <- glmnet(x, y)
> print(fit)