hezar.metrics.recall module

class hezar.metrics.recall.Recall(config: RecallConfig, **kwargs)

Bases: Metric

Recall metric for evaluating classification performance using sklearn’s recall_score.

Parameters:

• config (RecallConfig) – Metric configuration object.

• **kwargs – Extra configuration parameters passed as kwargs to update the config.

compute(predictions=None, targets=None, labels=None, pos_label=None, average=None, sample_weight=None, zero_division=None, n_decimals=None, output_keys=None)

Computes the Recall score for the given predictions against targets.

Parameters:

• predictions – Predicted labels.

• targets – Ground truth labels.

• labels – List of labels to include in the calculation.

• pos_label (int) – Label of the positive class.

• average (str) – Type of averaging for the recall score.

• sample_weight (Iterable[float]) – Sample weights for the recall score.

• zero_division (str | float) – Strategy for zero-division, default is 0.0.

• n_decimals (int) – Number of decimals for the final score.

• output_keys (tuple) – Filter the output keys.

Returns:

A dictionary of the metric results, with keys specified by output_keys.

Return type:

dict

required_backends: List[str | Backends] = [Backends.SCIKIT]

class hezar.metrics.recall.RecallConfig(objective: str = 'maximize', output_keys: tuple = ('recall',), n_decimals: int = 4, pos_label: int = 1, average: str = 'macro', sample_weight: Iterable[float] = None, zero_division: str | float = 0.0)

Bases: MetricConfig

Configuration class for Recall metric.

Parameters:

• name (MetricType) – The type of metric, Recall in this case.

• pos_label (int) – Label of the positive class.

• average (str) – Type of averaging for the recall score.

• sample_weight (Iterable[float]) – Sample weights for the recall score.

• zero_division (str | float) – Strategy for zero-division, default is 0.0.

• output_keys (tuple) – Keys to filter the metric results for output.

average: str = 'macro'

name: str = 'recall'

objective: str = 'maximize'

output_keys: tuple = ('recall',)

pos_label: int = 1

sample_weight: Iterable[float] = None

zero_division: str | float = 0.0