Best way to deploy Pluto/Jupyterlab for a small class with persistence

Community Teaching & Outreach
, , ,
1

I am a lecturer in agriculture and have developed a crop model written entirely in Julia. I would like to use this model in a course on agriculture and climate change with no more than 20 students. This is not a programming class, so students are not expected to write code, install Julia, or manage packages.

My plan is to provide an interactive interface through Pluto (preferred) or JupyterLab, where students can supply inputs and view outputs from the model. I am looking for a way to host this online that meets the following requirements:

  1. Each student should have their own session (not shared).

  2. Sessions should be persistent so that students can save their work and return to it later.

  3. The solution should not depend on my university’s servers, as arranging that is PITA.

What would be the best way to deploy Pluto (or JupyterLab) for a small class under these conditions? Would options like Binder, Nextjournal, or a self-hosted JupyterHub/PlutoHub on a cloud VM be suitable? Are there other hosting approaches (commercial or open-source) that would better fit these needs?

I have a small teaching grant to support this activity, but my experience with web programming and server setup is limited.

2

I have been looking into CoCalc recently for similar reasons to yours and have had a favourable experience.

1 Like