math-devcontainer

A devcontainer to build a grad school math course workspace. Opioninated for Python, Latex, and Markdown.

How to use the devcontainer

I generally use this workspace in one of two ways.

• Local development on VS Code using either of my machines (Macbook or Linux box)
• Codespaces on Github (which allows me to work on any machine I can login to Github, including my iPad)

How I structure coursework

I assume that the course is structured in weekly modules and that course work involves some problem set from a textbook and some software assignment.

First, I create a directory called module-{x} where x is that week's module number.

Then, I create a markdown like assignment-example.md. I generally use inline latex in markdown for all the problem sets, see the example assignment for more details.

The devcontainer uses the Run on Save extension to run a command that creates a pdf of any markdown saved using pandoc and pdflatex.

Sometimes pandoc will fail to create a pdf on save because of a formatting error. To see the Run on Save logging and debug the error, I got to the Output in the bottom panel of VS Code to investigate the error.

For any part of the assignment that uses Python, I change directory to the module and then use I use uv which is already installed on the devcontainer to set up Python for each module.

1. cd module-example
2. uv init
3. uv add numpy matplotlib or whatever dependencies you need.
4. uv run software_assignment.py
5. [Optional for linting] uv add ruff
6. [Optional for linting] uv run -m ruff format

I then copy the Python code into the assignment and link any images created from a library like matplotlib. See the example assignment for more details.