auc

ROC AUC based ranking recommendations metrics.

Functions

calc_auc_metrics(metrics, reco, interactions)

Calculate any ROC AUC based ranking metric.

Classes

AUCFitted(outer_merged_enriched, n_pos, ...)	Container with meta data got from _AUCMetric.fit method.
InsufficientHandling(value)	Strategy for handling insufficient reommendations cases
PAP(k[, debias_config, insufficient_handling])	Partial AUC + precision@k (pAp@k) joint classification and ranking metric.
PartialAUC(k[, debias_config, ...])	Partial AUC at k (pAUC@k).
_AUCMetric(k[, debias_config, ...])	ROC AUC based metric base class.