autointent.metrics

All the metrics for regexp, retrieval, scoring and decision nodes.

Submodules

• autointent.metrics.custom_types
• autointent.metrics.decision
• autointent.metrics.regexp
• autointent.metrics.retrieval
• autointent.metrics.scoring

Attributes

METRIC_FN

Classes

DecisionMetricFn	Protocol for decision metrics.
RegexpMetricFn	Protocol for regexp metrics.
RetrievalMetricFn	Protocol for retrieval metrics.
ScoringMetricFn	Protocol for scoring metrics.

Functions

decision_accuracy(y_true, y_pred)	Calculate decision accuracy. Supports both multiclass and multilabel.
decision_f1(y_true, y_pred)	Calculate decision f1 score. Supports both multiclass and multilabel.
decision_precision(y_true, y_pred)	Calculate decision precision. Supports both multiclass and multilabel.
decision_recall(y_true, y_pred)	Calculate decision recall. Supports both multiclass and multilabel.
decision_roc_auc(y_true, y_pred)	Calculate ROC AUC for multiclass and multilabel classification.
regexp_partial_accuracy(y_true, y_pred)	Calculate regexp partial accuracy.
regexp_partial_precision(y_true, y_pred)	Calculate regexp partial precision.
retrieval_hit_rate(query_labels, candidates_labels[, k])	Calculate the hit rate at position k.
retrieval_hit_rate_intersecting(query_labels, ...[, k])	Calculate the hit rate at position k for the intersecting labels.
retrieval_hit_rate_macro(query_labels, candidates_labels)	Calculate the hit rate at position k for the intersecting labels.
retrieval_map(query_labels, candidates_labels[, k])	Calculate the mean average precision at position k.
retrieval_map_intersecting(query_labels, candidates_labels)	Calculate the mean average precision at position k for the intersecting labels.
retrieval_map_macro(query_labels, candidates_labels[, k])	Calculate the mean average precision at position k for the intersecting labels.
retrieval_mrr(query_labels, candidates_labels[, k])	Calculate the Mean Reciprocal Rank (MRR) at position k.
retrieval_mrr_intersecting(query_labels, candidates_labels)	Calculate the Mean Reciprocal Rank (MRR) at position k for the intersecting labels.
retrieval_mrr_macro(query_labels, candidates_labels[, k])	Calculate the Mean Reciprocal Rank (MRR) at position k for the intersecting labels.
retrieval_ndcg(query_labels, candidates_labels[, k])	Calculate the Normalized Discounted Cumulative Gain (NDCG) at position k.
retrieval_ndcg_intersecting(query_labels, ...[, k])	Calculate the Normalized Discounted Cumulative Gain (NDCG) at position k for the intersecting labels.
retrieval_ndcg_macro(query_labels, candidates_labels)	Calculate the Normalized Discounted Cumulative Gain (NDCG) at position k for the intersecting labels.
retrieval_precision(query_labels, candidates_labels[, k])	Calculate the precision at position k.
retrieval_precision_intersecting(query_labels, ...[, k])	Calculate the precision at position k for the intersecting labels.
retrieval_precision_macro(query_labels, candidates_labels)	Calculate the precision at position k for the intersecting labels.
scoring_accuracy(labels, scores)	Calculate accuracy for multiclass and multilabel classification.
scoring_f1(labels, scores)	Calculate the F1 score for multiclass and multilabel classification.
scoring_hit_rate(labels, scores)	Calculate the hit rate for multilabel classification.
scoring_log_likelihood(labels, scores[, eps])	Supports multiclass and multilabel cases.
scoring_map(labels, scores)	Calculate the mean average precision (MAP) score for multilabel classification.
scoring_neg_coverage(labels, scores)	Supports multilabel classification.
scoring_neg_ranking_loss(labels, scores)	Supports multilabel.
scoring_precision(labels, scores)	Calculate precision for multiclass and multilabel classification.
scoring_recall(labels, scores)	Calculate recall for multiclass and multilabel classification.
scoring_roc_auc(labels, scores)	Supports multiclass and multilabel cases.

Package Contents

autointent.metrics.METRIC_FN