TomHodson
commented
3 years ago Summary of what I did today:
1) Made a environment.yaml and requirements.txt file using "conda env export --from-history > environment.yaml" which only gives the explicitely installed packages 2) Tested the environment file, turns out bumping matplotlib by a few minor versions already managed to through some of my figures off so I made a matplotlib stylesheet too.
3) Installed nbstripout https://github.com/kynan/nbstripout but couldn’t get it to work/understand the intended usage. https://github.com/mwouts/jupytext was suggested as a possibly better alternative
I wanted to also do some kind of automatic code formatting but given the above issues, I will leave that out for another day or install a manual extension to do it.
I've been working on a project involving doing large Markov Chain Monte Carlo runs of quantum systems for the two year of my PhD thus far. We're about to publish my first paper on the topic and I would love to be able to make the code available. Predictably the code base does not follow particularly good software engineering practices, there are no tests, it's quite environment specific, there is a lot of dead code, the commit history is a mess, etc etc.
I would like to use the workshop as an oppotunity to think about what would need to be done to make the project reproducable or at least useful to someone except me, and start implementing some of those changes. I imagine the lowest hanging fruit would be writing a good README.