rlearn.reporting.apply_holms_test

rlearn.reporting.apply_holms_test(imbalanced_experiment, control_oversampler=None)

Use the Holm’s method to adjust the p-values of a paired difference t-test for every combination of classifiers and metrics using a control oversampler.

Parameters:

imbalanced_experiment : object of rlearn.experiment.ImbalancedClassificationExperiment class

The fitted imbalanced experiment object.

control_oversampler : str or None, default=None

The name of the control oversampler. The default oversampler is the last one.

Returns:

holms_test : object of pd.DataFrame class

The results of the test as a pandas DataFrame.