What would you like to say?

When creating an EstimatorReport or a CrossValidationReport where the estimator is itself a GridSearchCV or a RandomizedSearchCV, we should store the results of the cross-validated search in the form of a ComparaisonReport of CrossValidationReport.

We should avoid any recomputation by leveraging the storage of the SearchCV from scikit-learn.