ensure compatibility with numpy 2.0
Closed this issue · 4 comments
See the NumPy 2.0 migration guide
It's probably straightforward to run Ruff's rule once on the pingouin code to conform to Numpy 2.0. But @raphaelvallat are you cool with simultanously replacing both Black and Flake8 with Ruff? It should be configurable within pyproject.toml to match the current rules of both. They also offer a drop-in GitHub Action workflow that could replace pingouin's current Black Action. I've implemented this before on a recent project and found it a convenient one-stop solution for these things.
Yeah, I keep hearing good things about ruff so let's change
Adding a checklist for this Issue that combines goals of numpy 2.0 compatibility with ruff replacement.
- Replace black with ruff in
pyproject.toml - Replace flake8 with ruff in
pyproject.toml - Replace black GitHub Action with ruff GitHub action
- Run ruff's formatter (in case of minor deviations from black)
- Run ruff's linter (in case of minor deviations from flake8)
- Run ruff's numpy 2.0 compatibility enforcer
- Replace black and flake8 with ruff in Contributing sections in docs and README.
@raphaelvallat this got addressed when #446 was merged, so you can close it as completed 👍 ✅