- Check that running a piece of code gives an expected result
- Important terms: regression tests/unit tests, 'code coverage'
More detail: - Things to cover: both 'good' and 'bad' data
- pytest, unittest { some particular tips w/ pytest? assert }
- Where do tests 'live'? with the package (MDAKit cookie cutter) vs. separate (MDA)
- Running locally, running automatically -> GitHub actions + CI
- codecov (+ automating)
- Check that software still works in general (installation, pass tests) as time progresses
- Related: 'upstream dependencies'+ keeping an eye on upcoming changes
- Both changes you make + that happen upstream can affect code in unexpected ways!
- GitHub actions + scheduling on push, at X time, etc