8000 Missing multi-output checks in common tests · Issue #13187 · scikit-learn/scikit-learn · GitHub
[go: up one dir, main page]

Skip to content
Missing multi-output checks in common tests #13187
Closed
@rok

Description

@rok

Description

Some classifiers and regressors support multi-output, however we do not have a common test for that. We should add it. See discussion in #11458.

We should also remember to remove redundant individual tests introduced by 95993a4.

Example of code for individual test

from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
from sklearn import datasets

X, y = datasets.make_multilabel_classification(n_classes=3)

# Test multi-output classifier
clf = RandomForestClassifier()
clf.fit(X, y.astype(str)).predict(X)

# Test multi-output regressor
rfr = RandomForestRegressor()
rfr.fit(X, y).predict(X)[:3]

Expected Results

No error is thrown for these checks. Some regressors and classifiers are omitted from this check.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      0