R¶

R is a free software environment for statistical computing and graphics.

Using R in a Jupyter notebook¶

Through Borah-OnDemand you can access a Jupyter notebook running on Borah. (More info about Borah-OnDemand.) Jupyter notebooks can be used to run R, but first you'll need to create a conda environment and install the R-kernel.

Before we can do any of the next steps, we'll need to install a conda/mamba distribution. We recommend miniforge: Installing miniforge.

Info

You may want to run the next steps in an interactive session, so your process is not killed on the login node. Use dev-session to request an interactive session.
Once mamba/conda is installed, we'll use it to create a new environment:
```
mamba create -n r-env -c conda-forge r-recommended r-irkernel jupyter
```
In this example, we're creating an environment called r-env and pulling the packages r-recommended, r-irkernel, and jupyter from the conda-forge channel.

Info

If you installed a different conda/mamba distribution (e.g., miniconda), you may need to replace mamba with conda.
After the environment is created, we'll activate it and install the interactive R kernel:
```
mamba activate r-env
R -e 'IRkernel::installspec()'
```

And that concludes the installation! To use the newly-installed R-kernel, navigate to Borah-OnDemand and request a Jupyter Notebook session through the Interactive Apps tab. On the notebook screen, open the dropdown for a "New" notebook and select "R".

Installing R packages¶

There are many R packages beyond the standard library. If you find yourself getting an error like there is no package called ‘PACKAGENAME’, you probably need to install a package. Here are a couple ways you can install packages in R:

Using Conda/Mamba¶

Many R packages are hosted on the conda-forge channel using the naming convention r-PACKAGENAME where PACKAGENAME is the name of the R package, e.g., r-tidyverse. If you want to check if an R package is available via conda or check what builds/versions are available, try searching r-PACKAGENAME with the name of the package you're looking for) on anaconda.org.

To install an R package into an existing environment:
```
mamba activate r-env
mamba install -c conda-forge r-PACKAGENAME
```
(See above for how to install R in a conda environment.)

To create a new environment containing your R package:

mamba create -n r-env -c conda-forge r-recommended r-PACKAGENAME

For both methods, you'll need to activate the environment each time before calling R:

mamba activate r-env

Using a module¶

Load an R module:
```
module load borah-misc r/4.2.2
```
Start an R terminal and install your desired package:
```
R
```
```
install.packages("PACKAGENAME")
```
where PACKAGENAME is the name of the package you want to install. It will ask if you want to install into a personal library in your home directory--say yes--and ask you to select a mirror--you can just choose the geographically closest one.

Once the package is finished installing, you're ready to go. If you run into issues installing a package, please contact researchcomputing@boisestate.edu.

Submitting an R job to the scheduler¶

Once you have your R packages installed you're ready to submit a job to the scheduler. (More info about scheduling a job.)

Below is an example submission script if you are using R from a module:

r-slurm.sh
#!/bin/bash
#SBATCH -J R_job            # job name
#SBATCH -o log_slurm.o%j    # output and error file name (%j expands to jobID)
#SBATCH -n 48               # total number of tasks requested
#SBATCH -N 1                # number of nodes you want to run on
#SBATCH -p bsudfq           # queue (partition)
#SBATCH -t 12:00:00         # run time (hh:mm:ss)

# Load the module
module load borah-misc r/4.2.2

# Replace myscript.R with the name of the script
Rscript myscript.R

This script can be submitted using sbatch:

sbatch r-slurm.sh

If you are using R from a conda environment, just replace lines 9 and 10 in the above script (highlighted) with the following:

r-slurm.sh
# Activate your environment
~/.bashrc
conda activate r-env