scikit-multilearn
scikit-multilearn copied to clipboard
MLARAM Attribute error
Walking through the documentation and the example classifiers and am attempting to fit a MLARAM instance with the included emotions dataset as below:
classifier = MLARAM()
classifier.fit(X_train, y_train)
predictions = classifier.predict(X_test)
AttributeError is being thrown during the fit method call as follows:
Have traced the issue back to _normalize_input_space(X)
, line 48 within site-packages/skmultilearn/adapt/mlaram.py
, where .max()
is being called on (originally at least) on a scipy.sparse.lil.lil_matrix
.
Issue rectified by casting to a numpy array (toarray()
), though it did take some digging and perhaps it would be good to require/implement a consistent input format?
Cheers