Using RStudio via OnDemand

RStudio is a code development environment for R. It provides an intuitive web interface for developing your R workflow interactively. Using RStudio on Apocrita via OnDemand allows you to scale to much larger datasets and workloads without putting a strain on your local machine. Please refer to the overview section for instructions on how to login to OnDemand.

Starting an RStudio session

Select an RStudio Server version from the Servers list of the Interactive Apps drop-down menu, or from the My Interactive Sessions page.

Once you have chosen the resources your job will need, including the version of RStudio Server required, click Launch to queue the job. Once resources are allocated, click the blue Launch RStudio Server button to connect to your session.

Choose your core request wisely

R will only use multiple cores if your code calls an appropriate parallelisation library such as parallel. When selecting a parallelisation library please be aware that detectCores() is not cluster aware and will return all cores on a node even if they are not actually available. Please ensure your code is parallelised in short test sessions before requesting a large number of cores for a long runtime. See the main R application page for further details: R - Example Jobs

Choose your RAM request wisely

Please establish a baseline of expected RAM usage in short test sessions before requesting a large amount of RAM for a long runtime.

Once connected, the familiar RStudio interface is presented, and you will be able to use the allocated resources, and access your research data located on Apocrita.

Installing packages

Don't use renv for versions older than RStudio 2026!

The renv project manager has become popular, but we only support its use when using RStudio 2026 and newer, as we explicitly changed how R runs for those versions (see this blog post for more information).

renv aggressively caches existing binaries. When using RStudio 2025 or older on Apocrita, sometimes the cached binaries for personal installs aren't compatible with all environments and can cause issues. If you want to integrate renv into your work on Apocrita, please use RStudio 2026 or newer.

It's likely your scripts will require additional R libraries; these can be installed using the install.packages() command in the console window of RStudio. If the package cannot be installed due to missing dependencies, please contact us and we will try to add them to the container RStudio runs from (for RStudio 2026 or newer) or the Spack environment (for RStudio 2025 or older) that is activated alongside RStudio.

Further tips and guidance regarding R package installations can be found on our blog: Untangle installations

Known issues

Web browser security

For security reasons, modern browsers will restrict OnDemand from embedding certain content in the RStudio application.

Rprojects and renv

As detailed above, the use of renv is only recommended when using RStudio 2026 or newer.

If you use the Rproject feature of RStudio, the default is to use renv as well. We recommend un-checking this box when using RStudio 2025 or older:

Exiting the session

If a session exceeds the requested running time, it will be killed. You may receive a message "The previous R session was abnormally terminated...". Click OK to acknowledge the message and continue. To avoid this message, it's good practice to exit the session cleanly when you have finished.

To cleanly exit RStudio, click File -> Quit Session... and then release resources back to the cluster queues by clicking the red Delete button for the relevant session on the My Interactive Sessions page.

For further guidance on exiting RStudio sessions please see this section of our blog: Exit intentionally

Useful additional tips

Further tips regarding the usage of RStudio can be found under our blog site: R Top Tips