fix: Allow index=None in pivot()

[ohines]

Closes #11592

I ran into this 'backlog' issue and decided to implement a fix.
I don't usually write rust code so apologies if something is a little shaky.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 80.38%. Comparing base (a9c8b59) to head (82fd514).
Report is 1 commits behind head on main.

❗ Current head 82fd514 differs from pull request most recent head d03f2c4. Consider uploading reports for the commit d03f2c4 to get more accurate results

Additional details and impacted files

@@             Coverage Diff             @@
##             main   #15855       +/-   ##
===========================================
+ Coverage   39.14%   80.38%   +41.24%     
===========================================
  Files        1364     1263      -101     
  Lines      167705   165371     -2334     
  Branches     3031        0     -3031     
===========================================
+ Hits        65646   132938    +67292     
+ Misses     101592    32433    -69159     
+ Partials      467        0      -467     

Flag	Coverage Δ
python	?
rust	?

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[ritchie46]

@MarcoGorelli any opinion on this? I always have to think hard on pivots. ^^

[MarcoGorelli]

thanks for working on this!

the expected output looks correct to me - it would be like doing

df.with_columns(_tmp=pl.lit("a")).pivot(
    index="_tmp", columns="foo", values=["N", "M"], aggregate_function=None
).drop("_tmp")

right?

I'll take a look at the implementation next week

[MarcoGorelli]

are you sure this renders properly? i think you need an extra blank line

[ohines]

Ah yes I didn't see that. Fixed now.

[MarcoGorelli]

For now I'm going to block this (sorry Oliver!) as I'm not really sure about it, see #11592 (comment)