calc_classification_metrics

rectools.metrics.classification.calc_classification_metrics(metrics: Dict[str, Union[ClassificationMetric, SimpleClassificationMetric]], merged: DataFrame, catalog: Optional[Collection[Union[str, int]]] = None) → Dict[str, float][source]

Calculate any classification metrics.

Works with prepared data.

Warning: It is not recommended to use this function directly. Use calc_metrics instead.

Parameters

  • metrics (dict(str -> (ClassificationMetric | SimpleClassificationMetric))) – Dict of metric objects to calculate, where key is a metric name and value is a metric object.

  • merged (pd.DataFrame) – Result of merging recommendations and interactions tables. Can be obtained using merge_reco function.

  • catalog (collection, optional) – Collection of unique item ids that could be used for recommendations. Obligatory only if metrics contains ClassificationMetric instances.

Returns

Dictionary where keys are the same as keys in metrics and values are metric calculation results.

Return type

dict(str->float)

Raises

  • ValueError – If n_items is not passed and ClassificationMetric is present in metrics.

  • TypeError – If unexpected metric is present in metrics.