feat: Added ddof arg to std

[anopsy]

This one is spicy, I'm very curious if I did it right

[anopsy]

Okay, I can see what's the problem, I need to check also why it passed the tests

[MarcoGorelli]

Wow, nice! I'm impressed you figured it all out independently, well done!

I think we need a unit test for this, in which we pass different values of ddof. Fancy trying to add one? Either in test_common.py, or in a new test file

[anopsy]

geeeeez, that was tough, I'm still wondering what that "correction: float =1.0" parameter was.
so happy it works

I think we need a unit test for this, in which we pass different values of ddof. Fancy trying to add one? Either in test_common.py, or in a new test file

sure thing!

[MarcoGorelli]

😆 I'd accidentally copied over correction from the array API, glad you figured out it was wrong

[anopsy]

really curious why the pytest-coverage fails, I've run the pytest --cov==narwhals and it didn't complain:

Also: I kind of feel that writing the test with nw.all would be more elegant, let me know what do you think

[MarcoGorelli]

that was fast, cool - you definitely understated your skills 😄 🙌

for the test, I think it's just that your branch is out-of-sync with main - can you run

git fetch upstream
git merge upstream/main
git push origin HEAD

?

You can also remove if df_any is None: return, it's no longer needed (it was just a "trick" I'd introduced to skip running the pyarrow-backed dataframe for CI runs when the pandas version was too old to support them)

[anopsy]

This is what I get, should I then git pull? (as far as I remember it will ask me about rebasing and I have no idea how to proceed)

thank you, I could definitely improve my git game

[MarcoGorelli]

Also: I kind of feel that writing the test with nw.all would be more elegant, let me know what do you think

maybe, but what you've done is good, I think it's ok to merge as-is

[MarcoGorelli]

nice one @anopsy !