02-cluster-software.qmd

# Cluster Software {#sec-cluster-software}

## Software Modules

Unlike a traditional desktop, you must load the different software that you wish to use into the environment via `modulefiles`. The list of supported software can be found on [Software List](https://campuscluster.illinois.edu/resources/software/) or by typing:

``` bash
module avail
```

## Viewing, Retrieving, and Disabling Module Software

The most frequently used module commands are:

``` bash
module list              # See active software modules
module load <software>   # Enable software
module unload <software> # Disable software
module purge             # Removes all active modules
```

Replace `<software>` with the name of the desired software module from `module avail`.

## Latest Version of *R*

As of **September 2021**, the latest version of *R* on ICC is *R* **4.1.1**. We recommend using the latest version of R with the `_sandybridge` suffix. The reason for using `_sandybridge` is to ensure compatibility on older nodes inside of the `stat` partition. For an example of a compatibility error, please see @sec-debugging-errors.

Moreover, with this version, the default library does not contain any non-standard packages.

*R* can be accessed by using:

``` bash
# Load software
module load R/4.1.1_sandybridge
```

**Note:** If the version is not specified during the load, e.g. `module load R`, then a default version of *R* will be used. This default may change without warning.

Once *R* is loaded, the Terminal/non-GUI version of *R* can be started by typing:

``` bash
R
```

To exit an *R* session on the cluster, type inside *R*:

```r
q(save = "no")
```

This will terminate the *R* session without saving any environment values.

## Ask for Help

ICC's help desk (via [help\@campuscluster.illinois.edu](mailto:help@campuscluster.illinois.edu){.email}) can help install software on ICC. Please send them an e-mail and *CC* your advisor.

### Writing a Custom Module

It is possible to compile and create your own modules. For details, see the tutorial [A Modulefile Approach to Compiling *R* on a Cluster](http://thecoatlessprofessor.com/programming/a-modulefile-approach-to-compiling-r-on-a-cluster/).