snorkelflow.client.metrics

Metric related functions.

Functions

add_metric_to_node(node, metric_name)	Add a metric from the node's settings.
delete_metric_from_node(node, metric_name)	Remove a metric from the node's settings.
get_candidate_extractor_metrics(node[, ...])	Returns a metrics DataFrame for given prediction from candidate extractor against ground truth.
get_df_metrics(node, df, ground_truth_col, ...)	Returns a metrics DataFrame after scoring two columns from a DataFrame.
get_model_metrics(node[, model_uid, split, ...])	Return specified metrics for a model.
list_available_metrics(node)	List available metrics which can be passed to get_model_metrics.
register_custom_metric(node, metric_func, ...)	Register a user-defined metric to available metrics.
tune_threshold_on_valid(node, x_uids, ...[, ...])	Tune decision threshold on ground truth in valid split given predicted_labels and predicted_probs.