DriftDetectorPipelineElement#

class capymoa.stream.preprocessing.DriftDetectorPipelineElement[source]#

Pipeline element that wraps around a drift detector

__init__( drift_detector: BaseDriftDetector, prepare_drift_detector_input_func: Callable, )[source]#

Initializes the pipeline element with a drift detector.

Parameters#

drift_detector: BaseDriftDetector: The drift detector that associated with the pipeline element
prepare_drift_detector_input_func: Callable: The function that prepares the input of the drift detector. The function signature should start with the instance and the prediction. E.g., prediction_is_correct(instance, pred). The output of that function gets passed to the drift detector

pass_forward( instance: Instance, ) → Instance[source]#

Simply returns the instance. The drift detector gets updated in pass_forward_predict.

pass_forward_predict( instance: Instance, prediction: Any = None, ) → Tuple[Instance, Any][source]#

Updates the drift detector; returns the instance and the prediction that were provided to the function

instance: Instance:: The instance
prediction: Any: The prediction from the previous pipeline steps. This can be None (e.g., when monitoring the the instance), an integer (e.g., when monitoring a classifier), or a float (when monitoring a regressor). It can also be anything else, but it must be compatible with prepare_drift_detector_input_func

Tuple[Instance, Any]: The instance and prediction that were provided as input