confusion_matrix()

audplot.confusion_matrix(truth, prediction, *, labels=None, label_aliases=None, percentage=False, show_both=False, ax=None)

Confusion matrix between ground truth and prediction.

The confusion matrix is calculated by audmetric.confusion_matrix.

Parameters

• truth (Union[Sequence, Series]) – truth values

• prediction (Union[Sequence, Series]) – predicted values

• labels (Optional[Sequence]) – labels to be included in confusion matrix

• label_aliases (Optional[Dict]) – mapping to alias names for labels to be presented in the plot

• percentage (bool) – if True present the confusion matrix with percentage values instead of absolute numbers

• show_both (bool) – if True and percentage is True it shows absolute numbers in brackets below percentage values. If True and percentage is False it shows the percentage in brackets below absolute numbers

• ax (Optional[Axes]) – pre-existing axes for the plot. Otherwise, calls matplotlib.pyplot.gca() internally

Examples

>>> truth = [0, 1, 1, 1, 2, 2, 2] * 1000
>>> prediction = [0, 1, 2, 2, 0, 0, 2] * 1000
>>> confusion_matrix(truth, prediction)

>>> confusion_matrix(truth, prediction, percentage=True)

>>> confusion_matrix(truth, prediction, show_both=True)

>>> confusion_matrix(truth, prediction, percentage=True, show_both=True)

>>> confusion_matrix(truth, prediction, labels=[0, 1, 2, 3])

>>> confusion_matrix(truth, prediction, label_aliases={0: "A", 1: "B", 2: "C"})