NAS currently supports three versions of the R machine learning platform:

• R Versions 3.6 and 4.2 are provided in the r3_6 and r4_2 NAS conda environments, respectively.
• R Version 4.1 and 4.2 are provided as a NAS software package.

The versions in conda come preinstalled with many of the popular R packages used in data science and machine learning, including JupyterLab. Although the version provided as a NAS software package does not include JupyterLab, you can configure JupyterLab to work with this version of R by following the instructions below.

Before You Begin: Complete all the steps in the following articles:

• Secure Setup for Using Jupyter Notebook for Machine Learning Development on NAS Systems
• Using Jupyter Notebook for Machine Learning Development on NAS Systems

Complete the following steps to install R and set up JupyterLab to use it:

1. Load the following NetBSD Packages Collection (pkgsrc) module and start R (version 4.1):

   pfe25% module load pkgsrc/2022Q1-rome
   pfe25% R

   Note: A newer version of R (4.2) is available in pkgsrc/2023Q3.

2. Install IRkernel in R:

   > install.packages("IRkernel")

3. Run IRkernel in R:

   > IRkernel::installspec()

4. Copy the IRkernel directory to your Jupyter data directory. For example, if your IRkernel directory is in $HOME/R/x86_64-redhat-linux-gnu-library/4.1, do:

   > cp -R $HOME/R/x86_64-redhat-linux-gnu-library/4.1/IRkernel $HOME/.local/share/jupyter

5. Change the R path (the first "R") in the $HOME/.local/share/jupyter/IRkernel/kernelspec /kernel.json file to:
   /nasa/pkgsrc/toss4/2022Q1-rome/bin/R (or other R)
6. Load Jupyter and run R.