feat: add Table.delete_row method

[tlambert03]

cc @jni

would appreciate your input on the API. I considered using Table.drop like the pandas API, but I thought it might be confusing that this one is in-place and that one is not. I'm also wary of the confusion between row indices and row headers (both of which could conceivably be integers), so I forced the usage of a kwarg.

[codecov]

Codecov Report

Attention: 3 lines in your changes are missing coverage. Please review.

Comparison is base (aa38630) 87.75% compared to head (6f41ae5) 87.73%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #610      +/-   ##
==========================================
- Coverage   87.75%   87.73%   -0.02%     
==========================================
  Files          40       40              
  Lines        4705     4722      +17     
==========================================
+ Hits         4129     4143      +14     
- Misses        576      579       +3     

Files	Coverage Δ
src/magicgui/widgets/_table.py	95.56% <82.35%> (-0.58%)	⬇️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jni]

Oh man, I needed a double-take to understand why this wasn't just not isinstance(header, Sequence). That's annoying 😂

[tlambert03]

Yep, classic :) string is a sequence but not "that" kind :)

[jni]

Missing a couple of test coverage lines, but looks good to me. I basically agree with your assessment. I wouldn't look to the pandas API but the draft DataFrame API, which bizarrely has drop_columns but not drop_rows — perhaps they are leaking implementation details into their API choices? 😂 I also couldn't find an issue in their repo. But anyway, drop_columns appears to be in-place (returns Self), so maybe drop_rows would be a good analogous method name? And maybe we should make an issue in their repo, though I don't feel super strongly about it.

[tlambert03]

Thanks!

Btw: I would not assume that it's in place because of -> Self. That means it returns "type of self" (not actual self)

See this type of usage for example: https://peps.python.org/pep-0673/#use-in-classmethod-signatures

[jni]

Fascinating! Thanks for the info!