museotoolbox.ai.SuperLearner.get_stats_from_cv

SuperLearner.get_stats_from_cv(confusion_matrix=True, kappa=False, OA=False, F1=False, nTrain=False)[source]

Extract statistics from the Cross-Validation. If Cross-Validation is 5-fold, getStatsFromCV will return 5 confusion matrix, 5 kappas…

Parameters

  • confusion_matrix (bool, default True.) – If True, will return first the Confusion Matrix.

  • kappa (bool, default False.) – If True, will return in kappa.

  • OA (bool, default False.) – If True, will return Overall Accuracy/

  • F1 (bool, default False.) – If True, will return F1 Score per class.

  • nTrain (bool, default False.) – If True, will return number of train samples ordered asc. per label.

Returns

Accuracies – A dictionary of each statistic asked.

Return type

dict

Examples

After having learned with museotoolbox.ai.SuperLearner :

>>> for stats in SL.get_stats_from_cv(confusion_matrix=False,kappa=True):
>>> stats['kappa']
0.942560083148
0.94227598585
0.942560083148
...