knnClassifier.py 文件源码

def predict_training(self, folds=5):
        """Do cross-validation and return probabilities for each data-point.

        Args:
            folds (int): Number of folds used for prediction on training data.
        """
        prediction = np.zeros((len(self.strings), self.num_classes))
        skf = StratifiedKFold(n_splits=folds)
        for train_index, test_index in skf.split(self.strings, self.labels):
            # prepare the training and test data
            training_strings = self.strings[train_index]
            test_strings = self.strings[test_index]
            training_labels = self.labels[train_index]
            # predicting the results
            part_prediction = self.find_knn(training_strings, training_labels,
                                            test_strings)
            prediction[test_index] = part_prediction
        return prediction