def test_get_best_booster(self) -> None:
unexpected_value = 20 # out of scope.
params = {"verbose": -1, "lambda_l1": unexpected_value} # type: Dict
dataset = lgb.Dataset(np.zeros((10, 10)))
study = optuna.create_study()
tuner = LightGBMTuner(params, dataset, valid_sets=dataset, study=study)
with pytest.raises(ValueError):
tuner.get_best_booster()
with mock.patch.object(BaseTuner, "_get_booster_best_score", return_value=0.0):
tuner.tune_regularization_factors()
best_booster = tuner.get_best_booster()
assert best_booster.params["lambda_l1"] != unexpected_value
# TODO(toshihikoyanase): Remove this check when LightGBMTuner.best_booster is removed.
with pytest.warns(DeprecationWarning):
tuner.best_booster
tuner2 = LightGBMTuner(params, dataset, valid_sets=dataset, study=study)
# Resumed study does not have the best booster.
with pytest.raises(ValueError):
tuner2.get_best_booster()
def test_best_booster_with_model_dir(self) -> None:
params = {"verbose": -1} # type: Dict
dataset = lgb.Dataset(np.zeros((10, 10)))
study = optuna.create_study()
with TemporaryDirectory() as tmpdir:
tuner = LightGBMTuner(params,
dataset,
valid_sets=dataset,
study=study,
model_dir=tmpdir)
with mock.patch.object(BaseTuner,
"_get_booster_best_score",
return_value=0.0):
tuner.tune_regularization_factors()
best_booster = tuner.get_best_booster()
tuner2 = LightGBMTuner(params,
dataset,
valid_sets=dataset,
study=study,
model_dir=tmpdir)
best_booster2 = tuner2.get_best_booster()
assert best_booster.params == best_booster2.params
def train(*args: Any, **kwargs: Any) -> Any:
"""Wrapper of LightGBM Training API to tune hyperparameters.
It tunes important hyperparameters (e.g., `min_child_samples` and `feature_fraction`) in a
stepwise manner. Arguments and keyword arguments for `lightgbm.train()
<https://lightgbm.readthedocs.io/en/latest/pythonapi/lightgbm.train.html>`_ can be passed.
"""
_check_lightgbm_availability()
auto_booster = LightGBMTuner(*args, **kwargs)
auto_booster.run()
return auto_booster.get_best_booster()
def train(*args: Any, **kwargs: Any) -> Any:
"""Wrapper of LightGBM Training API to tune hyperparameters.
It tunes important hyperparameters (e.g., ``min_child_samples`` and ``feature_fraction``) in a
stepwise manner. It is a drop-in replacement for `lightgbm.train()`_. See
`a simple example of LightGBM Tuner <https://github.com/optuna/optuna/blob/master/examples/lig
htgbm_tuner_simple.py>`_ which optimizes the validation log loss of cancer detection.
:func:`~optuna.integration.lightgbm.train` is a wrapper function of
:class:`~optuna.integration.lightgbm_tuner.LightGBMTuner`. To use feature in Optuna such as
suspended/resumed optimization and/or parallelization, refer to
:class:`~optuna.integration.lightgbm_tuner.LightGBMTuner` instead of this function.
Arguments and keyword arguments for `lightgbm.train()`_ can be passed.
.. _lightgbm.train(): https://lightgbm.readthedocs.io/en/latest/pythonapi/lightgbm.train.html
"""
_imports.check()
auto_booster = LightGBMTuner(*args, **kwargs)
auto_booster.run()
return auto_booster.get_best_booster()