R
R can be used through modules, Conda or Miniforge environments, JupyterHub kernels, RStudio when supported, or Apptainer containers.
Important
Cluster availability: This software is available as a module on Pegasus. It is not currently provided as a system module on Triton.
Find and Load R
Search for R:
module avail R
Load R:
module load R/<version>
Check R:
which R
R --version
Run an R script:
Rscript script.R
User R Libraries
Create a project-specific R library:
mkdir -p /scratch/<project>/$USER/Rlibs
export R_LIBS_USER=/scratch/<project>/$USER/Rlibs
Inside R:
.libPaths()
install.packages("ggplot2")
Run R in LSF
Example job script:
#!/bin/bash
#BSUB -P <project>
#BSUB -J r_job
#BSUB -q general
#BSUB -n 1
#BSUB -R "rusage[mem=4000]"
#BSUB -W 01:00
#BSUB -o r_job.%J.out
#BSUB -e r_job.%J.err
module load R/<version>
export R_LIBS_USER=/scratch/<project>/$USER/Rlibs
mkdir -p ${R_LIBS_USER}
Rscript script.R
Recommended Practice
Use R modules when available.
Use project-specific R libraries for reproducible package sets.
Use LSF batch jobs for long-running scripts.
Use Conda/Miniforge or Apptainer when module-based R is not sufficient.