Number of classes, 42, does not match size of target_names, 43. Try specifying the labels parameter. about productner HOT 4 CLOSED

This might help:
scikit-learn/scikit-learn#7751

from productner.

In case the unique classes in train and validation are a different number the classification_report fails by default
Line https://github.com/etano/productner/blob/master/classifier.py#L205 need to be fixed by adding labels
as in
print((classification_report(y_val, predicted_classes, target_names=target_names, digits = 6,labels=range(len(self.category_map)))))

from productner.

@ikvision can you make a PR?

from productner.

from productner.