ConfusionMatrix#

class openstef_beam.metrics.ConfusionMatrix(true_positives, true_negatives, false_positives, false_negatives, effective_true_positives, ineffective_true_positives)[source]#

Bases: NamedTuple

Confusion matrix components for peak detection in energy forecasting.

This class represents the results of classifying energy load peaks versus non-peaks, with additional effectiveness metrics to account for the direction and magnitude of prediction errors.

Variables:

  • true_positives – Boolean array indicating correctly predicted peaks.

  • true_negatives – Boolean array indicating correctly predicted non-peaks.

  • false_positives – Boolean array indicating incorrectly predicted peaks.

  • false_negatives – Boolean array indicating missed peaks.

  • effective_true_positives – Boolean array indicating true positives that are effective (peak correctly predicted with appropriate magnitude/direction).

  • ineffective_true_positives – Boolean array indicating true positives that are ineffective (peak correctly predicted but with wrong magnitude/direction).

Note

All arrays have shape (num_samples,) and correspond to the same time points in the original forecast evaluation.

true_positives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 0

true_negatives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 1

false_positives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 2

false_negatives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 3

effective_true_positives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 4

ineffective_true_positives: ndarray[tuple[Any, ...], dtype[bool]]#

Alias for field number 5