[python-package] [dask] fix mypy errors about Dask fit() return types

[jameslamb]

Contributes to #3867.

Resolves the following mypy errors.

python-package/lightgbm/dask.py:1178: error: Incompatible return value type (got "_DaskLGBMModel", expected "DaskLGBMClassifier")  [return-value]
python-package/lightgbm/dask.py:1381: error: Incompatible return value type (got "_DaskLGBMModel", expected "DaskLGBMRegressor")  [return-value]
python-package/lightgbm/dask.py:1553: error: Incompatible return value type (got "_DaskLGBMModel", expected "DaskLGBMRanker")  [return-value]

Switches the Dask estimators to returning self, so that e.g. DaskLGBMRegressor.fit() returns a DaskLGBMRegressor. This is useful for other user code and downstream libraries relying on LightGBM, to be able to do isinstance() checks and not need to rely on lightgbm-internal details like _DaskLGBMModel. It could also be relevant in the future is we add attributes to DaskLGBMRegressor, DaskLGBMClassifier, or DaskLGBMRanker that the other Dask estimators don't have.

It is, by the way, exactly the same approach the sklearn estimators use:

LightGBM/python-package/lightgbm/sklearn.py

Lines 989 to 1024 in e666a4b

	class LGBMRegressor(_LGBMRegressorBase, LGBMModel):
	"""LightGBM regressor."""
	def fit( # type: ignore[override]
	self,
	X,
	y,
	sample_weight=None,
	init_score=None,
	eval_set=None,
	eval_names: Optional[List[str]] = None,
	eval_sample_weight=None,
	eval_init_score=None,
	eval_metric: Optional[_LGBM_ScikitEvalMetricType] = None,
	feature_name: _LGBM_FeatureNameConfiguration = 'auto',
	categorical_feature: _LGBM_CategoricalFeatureConfiguration = 'auto',
	callbacks: Optional[List[Callable]] = None,
	init_model: Optional[Union[str, Path, Booster, LGBMModel]] = None
	) -> "LGBMRegressor":
	"""Docstring is inherited from the LGBMModel."""
	super().fit(
	X,
	y,
	sample_weight=sample_weight,
	init_score=init_score,
	eval_set=eval_set,
	eval_names=eval_names,
	eval_sample_weight=eval_sample_weight,
	eval_init_score=eval_init_score,
	eval_metric=eval_metric,
	feature_name=feature_name,
	categorical_feature=categorical_feature,
	callbacks=callbacks,
	init_model=init_model
	)
	return self

Related conversation: #5672 (comment)

[jameslamb]

@jmoralez what do you think about this one?

(no rush)

[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.