Tutorial: Self-installing R packages

In this tutorial, we will demonstrate how to install R packages in your Run:AI workload. This is particularly useful when you need specific packages that are not included in the pre-installed image.

Create a Run:AI Jupyter Lab workload

Follow the instructions in the Jupyter Lab tutorial to create a new workload in Run:AI and connect to the Jupyter Lab interface. You will see the following landing page when the workload is created successfully:

There are three ways of accessing R in this default Jupyter Lab environment:

R notebook
R Console
Terminal

In this tutorial, we will demonstrate how to use the Terminal application to install R packages.

Install R packages via Terminal

Open a new Terminal window by clicking on the Terminal icon in the Jupyter Lab interface.
Create a directory in your PVC to store any newly installed R packages. This ensures that the packages persist across sessions and are not removed after the workload is stopped running.

You can do this by creating a subdirectory in /scratch/${RUNAI_PROJECT}. For example:
```
mkdir -p /scratch/${RUNAI_PROJECT}/my_username/R_libs
```
which will create a directory named R_libs in the project PVC. You may want to change my_username to your own username (e.g., your unikey) or preferred directory name.
Set the R_LIBS_USER environment variable to point to this new directory.
```
export R_LIBS_USER=/scratch/${RUNAI_PROJECT}/my_username/R_libs
```
Start R by executing the command R in the terminal.
Check if the libary has already been installed in the image. You can either run
```
installed.packages()
```
to print out the full list of installed packages, or try loading the package using
```
library(package)
```
You need to replace package with the name of the package you want to check. If the package is not installed, you will receive an error message:

check if the xgboost library is installed
Install the desired R package using the install.packages() function. For example:
```
install.packages("xgboost", repos="http://cran.ms.unimelb.edu.au/")
```
This command will download and install the xgboost package into the directory specified by R_LIBS_USER.

Verify the installation

You can verify the installation by loading the package in R:

library("xgboost")

You can also use the libary in an R notebook or R console by setting the lib variable (e.g., lib="/scratch/sihnextgen/jfan0290/R_libs") when loading the library: