lale.lib.rasl.metrics module

class lale.lib.rasl.metrics.MetricMonoidFactory(*args, **kwargs)

Bases: MonoidFactory[Tuple[Union[Series, ndarray], Union[Series, ndarray], DataFrame], float, _M], Protocol

Abstract base class for factories that create metrics with an associative monoid interface.

abstract score_data(y_true: Series, y_pred: Series, X: Optional[DataFrame] = None) → float

score_data_batched(batches: Iterable[Tuple[Union[Series, ndarray], Union[Series, ndarray], DataFrame]]) → float

abstract score_estimator(estimator: TrainedOperator, X: DataFrame, y: Series) → float

score_estimator_batched(estimator: TrainedOperator, batches: Iterable[Tuple[DataFrame, Series]]) → float

abstract to_monoid(batch: Tuple[Union[Series, ndarray], Union[Series, ndarray], DataFrame]) → _M

Create a monoid instance representing the input data

lale.lib.rasl.metrics.accuracy_score(y_true: Series, y_pred: Series) → float

Replacement for sklearn’s accuracy_score function.

lale.lib.rasl.metrics.balanced_accuracy_score(y_true: Series, y_pred: Series) → float

Replacement for sklearn’s balanced_accuracy_score function.

lale.lib.rasl.metrics.f1_score(y_true: Series, y_pred: Series, pos_label: Union[int, float, str] = 1) → float

Replacement for sklearn’s f1_score function.

lale.lib.rasl.metrics.get_scorer(scoring: str, **kwargs) → MetricMonoidFactory

Replacement for sklearn’s get_scorer function.

lale.lib.rasl.metrics.r2_score(y_true: Series, y_pred: Series) → float

Replacement for sklearn’s r2_score function.