Brian Vanderwende (Jan 24 2023 at 21:35): Brian Vanderwende (Jan 24 2023 at 21:35):Hello all,
If you were at the ESDS Forum yesterday, you'll recall that I mentioned a new way of @hpcd supporting the R language: Conda environments. This is part of slate of new software installs and kernels we've added to the login and JupyterHub environments today. Here is a quick rundown:
NPL 2023a - New NCAR Package Library conda environment and Jupyter kernel. We've moved to Python 3.9 with this install.
R 4.2 - New conda-based install of R plus a pre-installed collection of packages. Accessible via conda (not environment modules).
Matlab R2022b - The latest Matlab version now has a JupyterHub language kernel.
As mentioned yesterday, the big change is using Conda to curate our R package library instead of source installs. I've installed packages that I know get used by our R users, but if you use R and would like to see other packages included, please let me know. The big benefit here should be more robust R libraries, and easier methods for users to augment with their own. The "r" channel has been added to the list of default channels below conda-forge, to increase the scope of what is available when running conda search.
Moving forward, the next thing on my agenda is to do some cleanup of old kernels in the JupyterHub. Later this week, I'll share a list of kernels we plan to remove and solicit your feedback.
Let me know if you have any questions or suggestions for improvements to our process that could better help you in doing science. Thanks!
Regards,
Brian
Last updated: May 16 2025 at 17:14 UTC