[python-package] fix mypy error about pandas categorical features

[jameslamb]

Contributes to #3756.
Contributes to #3867.

Fixes the following error from mypy:

python-package/lightgbm/basic.py:824: error:
Incompatible return value type (got "tuple[ndarray[Any, Any], list[str], list[str] | list[int], list[list[Any]]]", expected "tuple[ndarray[Any, Any], list[str], list[str], list[list[Any]]]")  [return-value]

[jameslamb]

This hint was wrong.

categorical_feature can be a list of integers (indicating 0-based column positions).

LightGBM/python-package/lightgbm/basic.py

Lines 1724 to 1726 in b74528f

	categorical_feature : list of str or int, or 'auto', optional (default="auto")
	Categorical features.
	If list of int, interpreted as indices.

And it's passed back out of this function unmodified if it's provided like that.

LightGBM/python-package/lightgbm/basic.py

Lines 814 to 817 in b74528f

	if categorical_feature == 'auto': # use cat cols from DataFrame
	categorical_feature = cat_cols_not_ordered
	else: # use cat cols specified by user
	categorical_feature = list(categorical_feature) # type: ignore[assignment]

[jameslamb]

categorical_feature can only be a list of strings, a list of integers, or the literal string 'auto'.

So I don't think this else: clause is helpful. When it's reached, under correctly-provided inputs it's just calling list() on something that's already a list, which has no effect.

I guess this could be providing a small bit of help for cases where some other type like a tuple was passed in, but if we want to support that then:

• it should be done at all public interfaces that support categorical_feature
• it shouldn't be buried inside a pandas-specific function

I think this can and should be removed, in the interest of further simplifying the flow of pandas data through the library.