Jupyter on Pelle¶

There are multiple IDEs on the UPPMAX clusters, among other Jupyter. Here we describe how to run Jupyter on Pelle.
Jupyter is an IDE specialised for the Python programming language.
Procedure¶
Prefer a video?
This procedure is also demonstrated for Rackham (FIX in this YouTube video
1. Start a Pelle remote desktop environment¶
This can be either:
- Login to the Pelle remote desktop environment using the website
- Login to the Pelle remote desktop environment using a local ThinLinc client
2. Start an interactive session¶
Within the Pelle remote desktop environment, start a terminal. Within that terminal, start an interactive session:
Where [project_number] is your
UPPMAX project, for example:
What is my UPPMAX project number?
See the UPPMAX documentation on how to see your UPPMAX projects
3a. Use software modules to provide Python, Jupyter and python packages needed¶
-
Step 1. Decide what is needed in terms of
- Python version
- Python packages (and versions)
-
Step 2: Check our Python Bundles page and choose compatible modules.
- Check if your needed python packages are compatible with a JupyterLab module.
- If not you may need to go to 3b instead to create a isolated environment in Conda.
Example with latest available versions on Pelle¶
- Within the terminal of the interactive session,
load the
JupyterLabmodule compatible with thefoss2025btoolchain (GCCcore-14.3.0). - You get
Python/3.13.5on the fly. - Also load
SciPy-bundleto getnumpyandpandas - and
matplotlibthat is its own module. - (You can add any other compatible python package module as well)
Forgot what the module system is?
3b Create a conda environment¶
Coming soon!
4. Start the Jupyter notebook¶
Still within the terminal of the interactive session, start a Jupyter Lab like this:
This will start a jupyter server session so leave this terminal open. The terminal will also display multiple URLs.
Copy the URL containing pXXX.
5. Browser to the Jupyter notebook¶
In the remote desktop environment on Pelle, start Firefox.
Set Firefox to the URL addresses from the Jupyter output, which will be similar to http://p115.uppmax.uu.se:8888/lab?token=73178b5ec897ae9bed6ae4b1815137d83dff671562574989