Pandas DEPR: Series.combine with non-Series

The code for this path is

elif [other_is_not_a_series]:
    new_index = self.index
    new_values = np.empty(len(new_index), dtype=object)
    with np.errstate(all="ignore"):
        new_values[:] = [func(lv, other) for lv in self._values]
    new_name = self.name

res_values = self.array._cast_pointwise_result(new_values)
return self._constructor(...)

AFAICT this is equivalent to series.map(lambda x: func(x, other)). (Unless maybe map hasn't been updated to use _cast_pointwise_result?). If so, we should deprecate allowing it and tell users to use map.

Comment From: Lakshya-Upadhyaya

Hi! I'd like to work on this. I tested the behavior on pandas 3.0.0.dev0 and confirmed that series.combine(scalar, func) produces the same result as series.map(lambda x: func(x, scalar)).

I'll add a Pandas4Warning when other is not a Series in Series.combine, directing users to use map instead. Following the same pattern as #62948.

Does this approach sound correct?

Comment From: jbrockmendel

Wait until there is consensus about doing this