Once again reminded of @rmcelreath.bsky.social's "Science as Amateur Software Development" (https://www.youtube.com/watch?v=8qzVV7eEiaI&ab_channel=RichardMcElreath). I know that time is a finite resource, but better coding would definitely be worth it -- it's a genuinely useful skill, outside of academia as well.
Comments
circa 1995, Leonard S Lerman, member of the National Academy, etc etc, wrote to Phil Sharp, then chair of Biology/MIT
Leonard said MIT PhD students should take a course in programming a part of the program
Phil wrote back a dismissive answer, saying this was a waste of time
*Which to be clear beats the hell out of any paper not sharing code
- Using different coding languages depending on the problem
- Writing unit and integration tests so we can know my code is doing what we hope it's doing
- Writing self documenting, readable code to the extent possible+ using comments for "why" instead of "what"
I also think there are people in academia who are currently coding rings around a lot of people in industry (myself included) Way less about ability and more about structures
Unit tests: rarely make sense
Docker: I know some folks love for repro, I'm not convinced
And 100% agree there's a lot of low hanging fruit we can go after! I've just run into enough folks who seem to think the software engineering-ish stuff is esoteric or "over the top" vs Docker is tricky but would solve many reproducibility problems??
something like:
It isn't that old scientists learn new things; rather they are replaced by young people who are not taught the errors of the past