Closed
Description
Describe the bug
The requires_y tag is missing from the DummyClassifier when I run DummyClassifier()._get_tags() .
I don't think this is a big deal but it seems weird.
The other tags from the other SuperClasses in the DummyClassifier show up.
- multioutput : True
- poor_score : True
- no_validation : True
Steps/Code to Reproduce
from sklearn.dummy import DummyClassifier
DummyClassifier()._get_tags()Expected Results
{'X_types': ['2darray'],
'_skip_test': False,
'allow_nan': False,
'binary_only': False,
'multilabel': False,
'multioutput': True,
'multioutput_only': False,
'no_validation': True,
'non_deterministic': False,
'poor_score': True,
'requires_fit': True,
'requires_positive_X': False,
'requires_positive_y': False,
'requires_y':True,
'stateless': False}Actual Results
{'X_types': ['2darray'],
'_skip_test': False,
'allow_nan': False,
'binary_only': False,
'multilabel': False,
'multioutput': True,
'multioutput_only': False,
'no_validation': True,
'non_deterministic': False,
'poor_score': True,
'requires_fit': True,
'requires_positive_X': False,
'requires_positive_y': False,
'stateless': False}ClassifierMixin
I did a little sleuthing and think I traced it back to the ClassifierMixin
Steps/Code to Reproduce
from sklearn.base import ClassifierMixin
ClassifierMixin()._more_tags()Expected Results
{'requires_y':True}Actual Results
AttributeError: type object 'ClassifierMixin' has no attribute '_more_tags'Versions
Ran this on Google Colab
scikit-learn in /usr/local/lib/python3.6/dist-packages (0.22.2.post1)
https://colab.research.google.com/drive/1E_U_TsvoufIVv1fFMclxKUnBAgU-wt7K?usp=sharing
Metadata
Metadata
Assignees
Labels
No labels