Rob

A note or question for reviewers: With this extra code, is pre-computing uniqueness and monotonicity for indexes still a performance boost? I'm still seeing gains for calling `is_monotonic_increasing` on a...

The pre-computing code should run anytime `Index` has generated its `_engine`, which can come from a couple other methods in addition to `is_unique` and `is_monotonic_increasing`. I think `DataFrame.loc` indirectly calls...

I think `DataFrame.join` uses `is_monotonic_increasing` and `is_unique`, although I'm not sure if they're in a place to benefit from the pre-computation. On net, the code's impact on efficiency looks pretty...

take

The problem appears in several other `from_formula`s, but gets more complicated because multiple formulas are used. For example, `BetaModel` takes both the main formula and an optional `exog_precision_formula`. That second...

Hi, sorry for the delay! The negative offset error should be fixed now. Turns out `Duration`'s `duration_ns` doesn't include its sign and just returns the absolute value of the duration...

Thanks for all the help, @MarcoGorelli! Does this need anything else to merge?

Hi @MarcoGorelli, I'm interested but I think this is pushing the limits of my polars/rust knowledge. For example, I'm a bit confused by `broadcast_try_binary_elementwise` and how to handle series of...

To get a named index, you can use `check_name=True` ([docs](https://pandera.readthedocs.io/en/stable/dataframe_models.html?highlight=check_name#index-name)): ```python class MySchema(SchemaModel): unique_id: Index[int] = Field(check_name=True) value: Series[int] = Field() ``` The default, `check_name=None`, is interpreted as `check_names=False` for...

> This pull request is stale because it has been open for thirty days with no activity. Please [update](https://pandas.pydata.org/pandas-docs/stable/development/contributing.html#updating-your-pull-request) and respond to this comment if you're still interested in working...