BoostingOverfit#

class BoostingOverfit[source]#

Check for overfit caused by using too many iterations in a gradient boosted model.

The check runs a pred-defined number of steps, and in each step it limits the boosting model to use up to X estimators (number of estimators is monotonic increasing). It plots the given score calculated for each step for both the train dataset and the test dataset.

Parameters
scorerUnion[Callable, str] , default: None

Scorer used to verify the model, either function or sklearn scorer name.

scorer_namestr , default: None

Name to be displayed in the plot on y-axis. must be used together with ‘scorer’

num_stepsint , default: 20

Number of splits of the model iterations to check.

n_samplesint , default: 1_000_000

number of samples to use for this check.

random_stateint, default: 42

random seed for all check internals.

__init__(alternative_scorer: Optional[Tuple[str, Union[str, Callable]]] = None, num_steps: int = 20, n_samples: int = 1000000, random_state: int = 42, **kwargs)[source]#
__new__(*args, **kwargs)#

Methods

BoostingOverfit.add_condition(name, ...)

Add new condition function to the check.

BoostingOverfit.add_condition_test_score_percent_decline_less_than([...])

Add condition.

BoostingOverfit.clean_conditions()

Remove all conditions from this check instance.

BoostingOverfit.conditions_decision(result)

Run conditions on given result.

BoostingOverfit.config([include_version, ...])

Return check instance config.

BoostingOverfit.from_config(conf[, ...])

Return check object from a CheckConfig object.

BoostingOverfit.from_json(conf[, ...])

Deserialize check instance from JSON string.

BoostingOverfit.metadata([with_doc_link])

Return check metadata.

BoostingOverfit.name()

Name of class in split camel case.

BoostingOverfit.params([show_defaults])

Return parameters to show when printing the check.

BoostingOverfit.remove_condition(index)

Remove given condition by index.

BoostingOverfit.run(train_dataset, test_dataset)

Run check.

BoostingOverfit.run_logic(context)

Run check.

BoostingOverfit.to_json([indent, ...])

Serialize check instance to JSON string.

Examples#