DriftOutput

class dataeval.detectors.drift.DriftOutput(is_drift: bool, threshold: float, p_val: float, distance: float, feature_drift: ndarray[Any, dtype[bool]], feature_threshold: float, p_vals: ndarray[Any, dtype[float32]], distances: ndarray[Any, dtype[float32]])

Output class for DriftCVM, DriftKS, and DriftUncertainty drift detectors

is_drift

Drift prediction for the images

Type:

bool

threshold

Threshold after multivariate correction if needed

Type:

float

feature_drift

Feature-level array of images detected to have drifted

Type:

NDArray

feature_threshold

Feature-level threshold to determine drift

Type:

float

p_vals

Feature-level p-values

Type:

NDArray

distances

Feature-level distances

Type:

NDArray