[FIXED] Returning support vectors of OneVsRestClassifier - sklearn

Issue

I'm not able to get the support vectors out of a OneVsRest or OneVsOne Classifier in Python.

The code looks like this but the classifiers don't seem to have a support vector attribute. The normal SVC model has them, so my question is: is there any way to get them out of the classifier?

model_linear = svm.SVC(kernel="linear")
or_linear = OneVsRestClassifier(model_linear)
clf_or_linear = or_linear.fit(X_train, y_train)

print(or_linear.support_vectors_)

oo_linear = OneVsOneClassifier(model_linear)
clf_oo_linear = oo_linear.fit(X_train, y_train)

print(oo_linear.support_vectors_)

Thanks for any help:-)

Solution

The or_linear is an array of SVC estimators.

if you print

print(or_linear.estimators_)

you will get:

[SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear'), SVC(kernel='linear')]

You can loop over every SVC to get every support vector like this

for svc in or_linear.estimators_:
    print(svc.support_vectors_)

Answered By - dotheM