dataeval.metrics.stats.labelstats¶

dataeval.metrics.stats.labelstats(dataset)¶

Calculates statistics for data labels.

This function computes counting metrics (e.g., total per class, total per image) on the labels.

Parameters:¶

dataset : Metadata or ImageClassificationDataset or ObjectDetect¶

Returns:¶

A dataclass containing the computed counting metrics for the labels.

Return type:¶

LabelStatsOutput

Examples

Calculate basic statistics on labels for a dataset.

>>> from dataeval.data import Metadata
>>> stats = labelstats(Metadata(dataset))
>>> print(stats.to_table())
Class Count: 5
Label Count: 15
Average # Labels per Image: 1.88
--------------------------------------
  Label: Total Count - Image Count
  horse:      2      -      2
    cow:      4      -      3
  sheep:      2      -      2
    pig:      2      -      2
chicken:      5      -      5