Selection inside of label_binarize not supported

[sscdotopen]

This does not work:

train_labels = label_binarize(train_data.helpful_votes > 0, classes=[True, False])

But this works:

train_data['is_helpful'] = train_data['helpful_votes'] > 0
train_labels = label_binarize(train_data['is_helpful'], classes=[True, False])

[stefan-grafberger]

To support this, we need to add support for the pd.Series binops and include them in the DAG. This is something that I already thought about but haven't gotten around to do. If the workaround works for you for now, I'd like to postpone work on this for a bit more. Basically, we need to patch __add__, __mul__, etc., and patch all of the comparison operations. This isn't difficult, but all of this together might take a few hours.

[sscdotopen]

The workaround is fine, no need to address this urgently