dataeval.bias.BalanceOutput

class dataeval.bias.BalanceOutput

Output class for the Balance bias evaluator.

Contains three polars DataFrames with normalized mutual information scores and threshold flags.

balance

DataFrame with global class-to-factor normalized mutual information:

  • factor_name: str - Name of the metadata factor. Includes “class_label” which represents the self-information (always 1.0).

  • mi_value: float - Normalized mutual information value between this factor and class labels

Type:

pl.DataFrame

factors

DataFrame with inter-factor normalized mutual information correlations:

  • factor1: str - Name of the first factor

  • factor2: str - Name of the second factor

  • mi_value: float - Normalized mutual information value

  • is_correlated: bool - True if mi_value > factor_correlation_threshold

Type:

pl.DataFrame

classwise

DataFrame with per-class-to-factor normalized mutual information:

  • class_name: str - Name of the class

  • factor_name: str - Name of the metadata factor

  • mi_value: float - Normalized mutual information value

  • is_imbalanced: bool - True if mi_value > class_imbalance_threshold

Type:

pl.DataFrame

data()

Return the output data as a dictionary.

Return type:

dict[str, Any]

meta()

Metadata about the execution of the function or method for the Output class.

Return type:

ExecutionMetadata