✔️ Linting Passed

All linting checks passed. Your pull request is in excellent shape! ☀️

_{Generated for commit: a439d5f. Link to the linter CI: here}

@lesteve ping

I know you don't like people pushing to your branch but that's an empty commit, so hopefully you don't mind too much 😅.

Basically I don't think your previous commit was following the guideline in #28959 which means that this wasn't testing all the global random seeds in the CI. I gave it a go, we'll see what happens ...

With my latest commit, the build log looks like it is testing all the global random seeds for test_dtype_preprocess_data (e.g. 400 tests ran i.e. 4 by random seeds because of the rescale_with_sw=True/False and fit_intercept=True/False parametrization)

OK the tests pass in the CI with all the global random seeds, and the previous commit was actually a normal CI which was green, so let's merge this one!

I know you don't like people pushing to your branch

This case is fine where you actually finished this trivial piece, so thank you @lesteve !
When I have larger PRs, I'd like to get asked first because such a "foreigner's" push might interfere with what I have in the pipe (and not yet pushed).

I don't think your previous commit was following the guideline

I don't understand. I followed the current description on our website and both you and I pushed an empty commit with "[all random seeds]" in the message.

I don't understand. I followed the current description on our website and both you and I pushed an empty commit with "[all random seeds]" in the message.

Note that my commmit message has a second line with the name of the test I wanted to run. Your commit message had a single line and apparently this runs the test suite normally in this case.

As mentioned in #28959 the format of the commit message is something like this:

some message [all random seeds]
test_something
test_some_other_thing

Thanks for the explanation. That knowledge would ideally live in the developer docs, don't you think? On my side: I search the code base for "[all random seeds]" very quickly and I should have read the docstring a bit more focused.