"A column-vector y was passed when a 1d array was expected" error message

Question

from sklearn.discriminant_analysis import LinearDiscriminantAnalysis

clf = LinearDiscriminantAnalysis()
clf.fit(np.matrix(X_train), np.matrix(y_train))

but I get the error message. Specified above.

I checked the shape of y_train but it's (294,1). tried the ravel() thing but it then it's (1,294) and if I transpose it then it looks back how it did before the ravel().

X_train.shape is (294,8).

Mulham Jarjanazi · Accepted Answer · 2019-10-10T00:51:14.587

3

first, don't use np.matrix, use np.array instead, its no longer recommended to use this class.

Try this:

clf.fit(X_train, np.ravel(y_train))

edited Oct 10 '19 at 00:51

answered Oct 10 '19 at 00:40

Mulham Jarjanazi

386
1
9

Glad is was able to help! – Mulham Jarjanazi Oct 10 '19 at 00:51
One more follow up question. Should also be casting my X_train data to np.array? or is a matrix okay for that one? – financial_physician Oct 10 '19 at 01:04
1

If your data was not already of type array, you have to convert it to np.array before feeding it to the classifier. np.matrix returns an array-like object, thats why its not recommended to use np.matrix in this context. – Mulham Jarjanazi Oct 10 '19 at 01:20

"A column-vector y was passed when a 1d array was expected" error message

1 Answers1