DEPR: use_inf_as_na

[jbrockmendel]

Support for use_inf_as_na isn't great, it performs poorly, and we don't have any way to make EAs respect it. Is this worth keeping?

[mahmoudmarayef]

We can convert infinity values to NaN before processing: Instead of relying on the use_inf_as_na parameter, you can manually convert infinity values to NaN before processing the data. This can be done using the replace method in pandas.

[jbrockmendel]

Discussed on last week's dev call, consensus was +1 to deprecate.