problem

[judelu2021]

cls = Kfda(kernel='rbf', n_components=3)
clf = cls.fit_transform(X_train, y_train)
TypeError: np.matrix is not supported. Please convert to a numpy array with np.asarray.

Although I have converted the data to np.array, he still reminded me that I do not support np.matrix, Please convert to a numpy array with np.array.
These codes used to work normally, but now they keep reporting this error. I look forward to your help to solve this problem. Thank you.

[ManuelZ]

What needs to be converted to a Numpy array is centroids_ in here:

kfda/kfda/kfda.py

Line 112 in f2634c4

self.clf_ = NearestCentroid().fit(centroids_, self.classes_)

With this:

self.clf_ = NearestCentroid().fit(np.asarray(centroids_), self.classes_)

[webbrain-one]

Thanks for reporting this! I’ve opened PR #4 to fix the np.matrix is not supported error by ensuring the data is converted to a proper NumPy array before processing. Please review it when you get a chance.