feat: add `Series|Expr` `cum_min` and `cum_max` methods

[FBruzzesi]

What type of PR is this? (check all applicable)

• [ ] 💾 Refactor
• [x] ✨ Feature
• [ ] 🐛 Bug Fix
• [ ] 🔧 Optimization
• [ ] 📝 Documentation
• [ ] ✅ Test
• [ ] 🐳 Other

Related issues

• Related issue #1371

Checklist

• [x] Code follows style guide (ruff)
• [x] Tests added
• [x] Documented the changes

If you have comments or can explain your changes, please do so below

[FBruzzesi]

Pyarrow introduced cumulative_min, cumulative_max and cumulative_prod in 13.0

[AlessandroMiola]

Pyarrow introduced cumulative_min, cumulative_max and cumulative_prod in 13.0

Not sure if relevant, should we maybe also re-map these into their polars native exprs via POLARS_TO_ARROW_AGGREGATIONS for use in group_by like exprs? Found this might be relevant on #1177.

[FBruzzesi]

Pyarrow introduced cumulative_min, cumulative_max and cumulative_prod in 13.0

Not sure if relevant, should we maybe also re-map these into their polars native exprs via POLARS_TO_ARROW_AGGREGATIONS for use in group_by like exprs? Found this might be relevant on #1177.

That's a very good point. However for arrow this is not possible, unless we move operations unsupported natively to use __iter__, but that would most likely be slow (List of pyarrow grouped aggregations).

Regarding pandas, I will check that's possible to achieve and add tests. However I am a bit hesitant due to the impossibility of passing arguments along (reverse=True)

[AlessandroMiola]

That's a very good point. However for arrow this is not possible, unless we move operations unsupported natively to use __iter__, but that would most likely be slow (List of pyarrow grouped aggregations).

ahhh, right, I see ✌️