I’ve heard about Jupytrerhub, lab, etc and thought I’d give it a try. The webinars and some of the tutorials (YouTube) show how to use Notebooks and it seems neat.
I’ve set up a server on AWS and it works as expected
My largest problem is not being able to manage it well. I could list examples however the most pressing is installing system wide python packages.
The read-the-docs, example here make no sense. I understand how to install packages within a. python venv however, I’m not sure to use the jupyter environment to install a package for all users to use.
If possible, please provide a link or some short explanation.
Thanks.
I recently started using JupyterHub and created an instance of The Littlest JupyterHub in a linux virtual machine. What I found helpful to understand installing python packages is the following from this page:
Note that you may have two different Python environments: the ‘hub’ environment and the ‘user’ environment. The hub environment is where JupyterHub itself is running, serving the main web framework for login and managing Jupyter servers on your JupyterHub website. The user environment is where individual’s Jupyter notebooks are running.
For example, in The Littlest JupyterHub, the two environments are usually located in /opt/tljh/hub (for the hub environment) and /opt/tljh/user (for the user environment).
If you open a terminal from within JupyterHub and execute which python3, you should get /opt/tljh/user. Any packages you pip install or conda install from the terminal will be installed in this environment and available to all users, since this is the environment that they automatically have access to.
2 Likes
Thanks. That helps a lot.