[python] add more type hints on LGBMModel methods

[jameslamb]

Contributes to #3756.

Adds type hints on some methods in lightgbm.sklearn.LGBMModel that don't currently have them.

[jmoralez]

Most of these changes raise errors on mypy (logs):

python-package/lightgbm/sklearn.py:830: error: Incompatible return value type (got "None", expected "int")
python-package/lightgbm/sklearn.py:837: error: Incompatible return value type (got "None", expected "int")
python-package/lightgbm/sklearn.py:851: error: Incompatible return value type (got "None", expected "int")
python-package/lightgbm/sklearn.py:858: error: Incompatible return value type (got "Union[str, Callable[[Any, Any], Tuple[Any, Any]], Callable[[Any, Any, Any], Tuple[Any, Any]], None]", expected "Union[str, Union[Callable[[Any, Any], Tuple[Any, Any]], Callable[[Any, Any, Any], Tuple[Any, Any]]]]")
python-package/lightgbm/sklearn.py:1095: error: Incompatible return value type (got "None", expected "int")

Although I believe the if statement is enough to check that most of them aren't None, I'm not sure if mypy isn't able to see it.

[StrikerRUS]

I remember the following discussion about these mypy false alarms: #4837 (review), #4837 (comment).

[jameslamb]

Thanks for finding those links @StrikerRUS ! I remembered those conversations before, and am ok with the conclusion we reached there, that # type: ignore comments could be used to handle these cases where we "just know" that an attribute like n_classes_ will always either be an integer (if model has been fitted) or raise an exception (if not fitted).

I didn't add such comments in this PR because I just wanted to focus on getting the type hints correct, and wanted to leave mypy warnings for separate PRs targeting #3867.

[StrikerRUS]

I didn't add such comments in this PR because I just wanted to focus on getting the type hints correct, and wanted to leave mypy warnings for separate PRs targeting #3867.

Great, totally agree!

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.