sklearn_to_flow raises TypeError on simple flow #636
Description
I found that the following simple example does not convert to an OpenMLFlow with sklearn_to_flow . I do not know why this happens (lists of lists?). The definition of a flow in code is also quite unclear and has a lots of gaps left with "TODO: add documentation".
import openml
from sklearn.preprocessing import OrdinalEncoder
openml.flows.sklearn_to_flow(OrdinalEncoder(categories=[[0,1],[0,1]]))
raises
... traceback ...
File "D:\repositories/openml-python\openml\flows\sklearn_converter.py", line 534, in _extract_information_from_model
raise TypeError(msg)
TypeError: Second item of tuple does not match assumptions. Expected OpenMLFlow, got <class 'numpy.float32'>