New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Initial implementation of cross validated SVC #117

Merged
merged 1 commit into from Apr 1, 2011

Conversation

Projects
None yet
2 participants
@fabianp
Member

fabianp commented Apr 1, 2011

Ok, this still needs some work. This pull request is just to share some ideas and because of the 'lookie what I did' factor.

On the API I'm still not sure that it's a great idea to pass multiple parameters as lists or tuples, which is how it's implemented right now. It's handy, but might give some conflicts in the long run:

>>> clf = SVCCV(C=[1., 2.])
>>> clf.fit(X, y)

@fabianp fabianp merged commit 64961c1 into scikit-learn:master Apr 1, 2011

@fabianp fabianp reopened this Apr 1, 2011

@agramfort

This comment has been minimized.

Show comment
Hide comment
@agramfort

agramfort Apr 1, 2011

Member

can you add a test that compares with a pure scikit CV method?
it would be great to have a transparent interface. fit(X, y, cv)
if isinstance(cv, KFold):
then fall back to fast libsvm version

makes sense?

Member

agramfort commented Apr 1, 2011

can you add a test that compares with a pure scikit CV method?
it would be great to have a transparent interface. fit(X, y, cv)
if isinstance(cv, KFold):
then fall back to fast libsvm version

makes sense?

@fabianp

This comment has been minimized.

Show comment
Hide comment
@fabianp

fabianp Apr 8, 2011

Member

Will come back with something better

Member

fabianp commented Apr 8, 2011

Will come back with something better

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment