Installing R packages

This exercise aims to give you hands-on experience with installing R packages and contrasting some different approaches. It should take approximately 30 minutes to an hour to complete (potentially more when also conducting the benchmark exercises). To get started, choose an R package that has at least one non-trivial dependency and ideally includes some compiled code (written in C, C++ or Fortran). It is recommended to choose an R package related to your area of research, but if you are really lacking inspiration, you can complete this exercise using the dplyr package.

You can find the relevant documentation in the “R and Julia” section of the GSSI training material and in the R package management section in the VSC Docs.

Here are the different methods you can use to install R packages on a VSC cluster:

• using centrally installed R packages from software modules (e.g., R-bundle-CRAN, R-bundle-Bioconductor, or dedicated modules for specific packages; use module spider to find a package)
• using install.packages() in a personal R library, e.g., under $VSC_DATA with R_LIBS_USER set in ~/.Renviron
• installing specific package versions or packages from Git repositories, e.g., with install_version() from the remotes package, devtools::install_github(), or R CMD INSTALL from a source tarball
• using vsc-rproject, the tool to create and manage RStudio Projects from the command line (see vsc-Rproject in the VSC Docs)
• using a Conda-based environment with R (see conda-based environment managers in the VSC Docs)
• using an Apptainer container (see the building containers exercise)

NOTE: Make sure to keep the different installations separated, mixing approaches will likely cause problems.

After trying the different installation approaches, compare them with respect to the following characteristics:

• ease of use
• time it takes to install the package
• disk space and number of files consumed by the installation
• reproducibility
• portability (e.g., does the same installation work on different clusters, partitions or CPU microarchitectures? R packages with compiled code are sensitive to the -march compiler flag that was used at installation time)
• performance: ideally, use a benchmark (check if the source of the chosen package includes one) to compare runtimes for the different installations. Try to explain possible differences from a theoretical point of view and connect to what you observe in practice.