MetadataOutput

class dataeval.metrics.bias.MetadataOutput(discrete_factor_names: list[str], discrete_data: ndarray[Any, dtype[int64]], continuous_factor_names: list[str], continuous_data: ndarray[Any, dtype[int64 | float64]] | None, class_labels: ndarray[Any, dtype[int64]], class_names: ndarray[Any, dtype[Any]], total_num_factors: int)

Output class for metadata_binning() function

discrete_factor_names

List containing factor names for the original data that was discrete and the binned continuous data

Type:

list[str]

discrete_data

Array containing values for the original data that was discrete and the binned continuous data

Type:

NDArray[np.int]

continuous_factor_names

List containing factor names for the original continuous data

Type:

list[str]

continuous_data

Array containing values for the original continuous data or None if there was no continuous data

Type:

NDArray[np.int or np.double] | None

class_labels

Numerical class labels for the images/objects

Type:

NDArray[np.int]

class_names

Array of unique class names (for use with plotting)

Type:

NDArray[Any]

total_num_factors

Sum of discrete_factor_names and continuous_factor_names plus 1 for class

Type:

int