IEPY raises ValueError ColumnFilter eliminates all columns! about iepy HOT 4 CLOSED

This is simply because there's no enough information yet, so no column pass the filter.
We should skip current IEPY-loop-round to ask human for more information, instead of crashing

from iepy.

This filter can be trained using the entired data set because it doesn't need labeled data. I think this could be a good idea.

from iepy.

Yes, I see that as a good opportunity, but I rather prefer to fix this issue with the minimum effort, and consider seriously your recommendation on a separated issue/feature. Not because I don't like your idea, but because it may imply more time to invest, and with the minimum here I want to round the round6 of experiments.

So, I'll try for now to only catch this kind of errors and keep running asking a bit more of info to the Human.

from iepy.

Ack

from iepy.