dragonn.metrics module¶

class dragonn.metrics.ClassificationResult(labels, predictions, task_names=None)[source]¶: Bases: object

class dragonn.metrics.IgnoreNumpyErrors[source]¶: Bases: object

dragonn.metrics.auPRC(labels, predictions)[source]¶

dragonn.metrics.auPRG(labels, predictions)[source]¶

dragonn.metrics.auROC(labels, predictions)[source]¶

dragonn.metrics.balanced_accuracy(labels, predictions, threshold=0.5)[source]¶

dragonn.metrics.negative_accuracy(labels, predictions, threshold=0.5)[source]¶

dragonn.metrics.positive_accuracy(labels, predictions, threshold=0.5)[source]¶

dragonn.metrics.recall_at_precision_threshold(labels, predictions, precision_threshold)[source]¶