8000 sample_weight not tested for random forests · Issue #2752 · scikit-learn/scikit-learn · GitHub
[go: up one dir, main page]
Skip to content

sample_weight not tested for random forests #2752

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.

Sign up for GitHub

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

Jump to bottom
Closed
mblondel opened this issue Jan 14, 2014 · 6 comments
Closed

sample_weight not tested for random forests #2752

mblondel opened this issue Jan 14, 2014 · 6 comments
Labels
Enhancement

Comments

@mblondel
Copy link
Member

Would be nice to add tests, both for the classifier and regressor variants.
@glouppe
Copy link
Contributor
glouppe commented Jan 14, 2014

What kind of tests do you have in mind? sample_weights is tested for single decision trees.

@mblondel
Copy link
Member Author

sample_weights is tested for single decision trees.

I know but it would still be nice to test this code path.

What kind of tests do you have in mind?

No concrete idea for unit tests. A visual example like
http://scikit-learn.org/stable/auto_examples/svm/plot_weighted_samples.html
would be a nice addition too, to check that the code behaves as expected.

Another possible source of inspiration:
#1489 (comment)

@glouppe
Copy link
Contributor
glouppe commented Jan 14, 2014

Another solution would be to add a suite of common tests for all estimators
supporting sample_weights. This way, all estimators would be equally and
consistently tested.

On 14 January 2014 09:47, Mathieu Blondel notifications@github.com wrote:

sample_weights is tested for single decision trees.

I know but it would still be nice to test this code path.

What kind of tests do you have in mind?

No concrete idea for unit tests. A visual example like
http://scikit-learn.org/stable/auto_examples/svm/plot_weighted_samples.html
would be a nice addition too, to check that the code behaves as expected.

Another possible source of inspiration:
#1489 (comment)#1489 (comment)

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/2752#issuecomment-32247416
.

@arjoly
Copy link
Member
arjoly commented Jan 14, 2014

Another solution would be to add a suite of common tests for all estimators
supporting sample_weights. This way, all estimators would be equally and
consistently tested.

+1

@mblondel
Copy link
Member Author

That would be nice but we really need to break the common test suite into parts ;-)

@arjoly
Copy link
Member
arjoly commented Oct 8, 2015

They are now tested. Closing and opening a new issue for common tests.

@arjoly arjoly closed this as completed Oct 8, 2015
@arjoly arjoly mentioned this issue Oct 8, 2015
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
Enhancement
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@mblondel @glouppe @arjoly
0