Decision function for the OneVsRestClassifier. Return the distance of each sample from the decision boundary for each class. This can only be used with estimators which implement the `decision_function` method. Parameters ---------- X : array-like of
(self, X)
| 563 | |
| 564 | @available_if(_estimators_has("decision_function")) |
| 565 | def decision_function(self, X): |
| 566 | """Decision function for the OneVsRestClassifier. |
| 567 | |
| 568 | Return the distance of each sample from the decision boundary for each |
| 569 | class. This can only be used with estimators which implement the |
| 570 | `decision_function` method. |
| 571 | |
| 572 | Parameters |
| 573 | ---------- |
| 574 | X : array-like of shape (n_samples, n_features) |
| 575 | Input data. |
| 576 | |
| 577 | Returns |
| 578 | ------- |
| 579 | T : array-like of shape (n_samples, n_classes) or (n_samples,) for \ |
| 580 | binary classification. |
| 581 | Result of calling `decision_function` on the final estimator. |
| 582 | |
| 583 | .. versionchanged:: 0.19 |
| 584 | output shape changed to ``(n_samples,)`` to conform to |
| 585 | scikit-learn conventions for binary classification. |
| 586 | """ |
| 587 | check_is_fitted(self) |
| 588 | if len(self.estimators_) == 1: |
| 589 | return self.estimators_[0].decision_function(X) |
| 590 | return np.array( |
| 591 | [est.decision_function(X).ravel() for est in self.estimators_] |
| 592 | ).T |
| 593 | |
| 594 | @property |
| 595 | def multilabel_(self): |