Handling of categorical columns about ppscore HOT 2 CLOSED

The main issue here was that the task was inferred wrongly based on the dtype. In the next release, the dtype will determine the task and the issue won't happen any more

from ppscore.

Thank you for your report.
It seems to me that this warning appears in your example because you only generate 10 rows. Therefore, per default all columns are inferred as categorical because they all only have 10 unique values.

However, the values of the columns are float and sklearn does not like to receive float values when using a classifier:
https://stackoverflow.com/questions/41925157/logisticregression-unknown-label-type-continuous-using-sklearn-in-python

TODO: We might want to handle this case in the code via explicitly applying a LabelEncoder to a series which is inferred as categoric but has the dtype float

Thank you,
Florian

from ppscore.