Thanks to visit codestin.com
Credit goes to github.com

Skip to content

[MRG+2] MultiOutputClassifier classes_ attribute #14629

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
Merged
glemaitre merged 10 commits into from
Aug 14, 2019
Merged

[MRG+2] MultiOutputClassifier classes_ attribute #14629

glemaitre merged 10 commits into from
Aug 14, 2019

Conversation

@agamemnonc
Copy link
Contributor

@agamemnonc agamemnonc commented Aug 12, 2019

Reference Issues/PRs

Fixes #14615.

What does this implement/fix? Explain your changes.

MultiOutputClassifier class now has attribute classes_ (same as ClassifierChain)

Any other comments?

@agamemnonc
Copy link
Contributor Author

agamemnonc commented Aug 12, 2019

Shall we include a non-regression test checking the classes_ of a fitted MultiOutputClassifier?

jnothman
jnothman reviewed Aug 12, 2019
View reviewed changes
Copy link
Member

@jnothman jnothman left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please do extend a test to check this.

@agamemnonc
Copy link
Contributor Author

agamemnonc commented Aug 12, 2019

Please do extend a test to check this.

OK done.

Please let me know if anything else is needed. Otherwise please review and change PR title accordingly.

@agamemnonc agamemnonc changed the title Multioutput classes fix MultiOutputClassifier classes_ attribute fix Aug 12, 2019
@agamemnonc agamemnonc changed the title MultiOutputClassifier classes_ attribute fix MultiOutputClassifier classes_ attribute Aug 12, 2019
jnothman
jnothman reviewed Aug 12, 2019
View reviewed changes
@agamemnonc
Copy link
Contributor Author

agamemnonc commented Aug 13, 2019

On a side note, the dataset used in most of the tests is created globally, would it not be better practice to have that wrapped in a function definition, then call that function from within all tests that use the particular dataset?

scikit-learn/sklearn/tests/test_multioutput.py

Lines 151 to 162 in 5a435f4

# Import the data
iris = datasets.load_iris()
# create a multiple targets by randomized shuffling and concatenating y.
X = iris.data
y1 = iris.target
y2 = shuffle(y1, random_state=1)
y3 = shuffle(y1, random_state=2)
y = np.column_stack((y1, y2, y3))
n_samples, n_features = X.shape
n_outputs = y.shape[1]
n_classes = len(np.unique(y1))
classes = list(map(np.unique, (y1, y2, y3)))

@glemaitre
Copy link
Member

glemaitre commented Aug 13, 2019

On a side note, the dataset used in most of the tests is created globally, would it not be better practice to have that wrapped in a function definition, then call that function from within all tests that use the particular dataset?

We could use the pytest.fixture instead. But I would say that this is out of the scope for this PR. We might need to have something consistent across the source code then such that this is easy to know where the fixtures are defined etc. However, this is a good remark

@agamemnonc
Copy link
Contributor Author

agamemnonc commented Aug 13, 2019

On a side note, the dataset used in most of the tests is created globally, would it not be better practice to have that wrapped in a function definition, then call that function from within all tests that use the particular dataset?

We could use the pytest.fixture instead. But I would say that this is out of the scope for this PR. We might need to have something consistent across the source code then such that this is easy to know where the fixtures are defined etc. However, this is a good remark

Agreed. Feel free to open a separate issue then if you wish.

glemaitre
glemaitre requested changes Aug 13, 2019
View reviewed changes
Copy link
Member

@glemaitre glemaitre left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Couple of comments but it looks good

@agamemnonc agamemnonc mentioned this pull request Aug 13, 2019
Closed
NicolasHug
NicolasHug approved these changes Aug 13, 2019
View reviewed changes
Copy link
Member

@NicolasHug NicolasHug left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit but LGTM, thanks @agamemnonc

glemaitre
glemaitre approved these changes Aug 13, 2019
View reviewed changes
Copy link
Member

@glemaitre glemaitre left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@glemaitre
Copy link
Member

glemaitre commented Aug 13, 2019

Could you also merge master into your branch such that we check that the CI for the doc is passing.
It seems just an issue that we earlier fixed

@agamemnonc
Copy link
Contributor Author

agamemnonc commented Aug 14, 2019

Thanks @glemaitre @NicolasHug and @jnothman for feedback.

I think this is now ready.

@agamemnonc agamemnonc changed the title MultiOutputClassifier classes_ attribute [MRG+2] MultiOutputClassifier classes_ attribute Aug 14, 2019
@glemaitre glemaitre merged commit 06dc44c into scikit-learn:master Aug 14, 2019
@glemaitre
Copy link
Member

glemaitre commented Aug 14, 2019

@agamemnonc Thanks for your contribution

@agamemnonc agamemnonc deleted the multioutput_classes_fix branch August 14, 2019 15:15
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@jnothman jnothman jnothman left review comments

@glemaitre glemaitre glemaitre approved these changes

+1 more reviewer

@NicolasHug NicolasHug NicolasHug approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

AttributeError with cross_val_predict(method='predict_proba') when using MultiOuputClassifier

4 participants

@agamemnonc @glemaitre @jnothman @NicolasHug