Check of number_of_features_or_feed of ltn.variable

[ivanDonadello]

Hello,

when we define a variable with ltnw.variable('name', num_feats) if num_feats is a numpy.int_ ltn instantiates a variable with a wrong shape. I guess this is due to the control at line 185 here https://github.com/logictensornetworks/logictensornetworks/blob/master/logictensornetworks.py, as a numpy.int_ is not an int. A check like this if type(number_of_features_or_feed) is int or type(number_of_features_or_feed) is numpy.int_ should work.

[mspranger]

Thanks Ivan! Let me check and update this within the next few days

[sbadredd]

Closing because we are not maintaining the older version of LTN anymore
The solution pointed out by Ivan should work though 👍