Open samwisehawkins opened 5 years ago
OK, I realise this is because I did not specify n_features
,. But perhaps this should be raised earlier and less cryptically?
I had a similar error. It is because when I do not specify the number of features I want to keep n_features
is set to np.inf
at line 233 in mifs.py, and n_features
is used to stop the main loop. Therefore, it crashes after selecting the latest possible feature.
This could be corrected by changing line 233?
if self.n_features == 'auto': n_features = p
Any attempt to call fit() raises an error
ValueError: numpy.nanargmax raises on a.size==0 and axis=None; So Bottleneck too.