class Chi2(SklScorer):
"""
A wrapper for `${sklname}`. The following is the documentation
from `scikit-learn <http://scikit-learn.org>`_.
${skldoc}
"""
__wraps__ = skl_fss.chi2
feature_type = DiscreteVariable
class_type = DiscreteVariable
preprocessors = [Discretize()]
def score(self, X, y):
f, p = skl_fss.chi2(X, y)
return f
def test_bad_learner(self):
"""
Some learners on input might raise error.
GH-38
"""
w = self.widget
self.assertFalse(w.Error.fitting_failed.is_shown())
learner = LogisticRegressionLearner()
learner.preprocessors = [Discretize()]
self.send_signal(w.Inputs.learner, learner)
self.send_signal(w.Inputs.data, self.iris)
self.assertTrue(w.Error.fitting_failed.is_shown())
learner.preprocessors = []
self.send_signal(w.Inputs.learner, learner)
self.assertFalse(w.Error.fitting_failed.is_shown())
def test_set_preprocessor(self):
"""
Test preprocessor set
"""
w = self.widget
preprocessor = Continuize()
# check if empty
self.assertIn(w.preprocessors, [[], None])
self.send_signal(w.Inputs.preprocessor, preprocessor)
# check preprocessor is set
self.assertEqual(w.preprocessors, [preprocessor])
self.assertIn(preprocessor,
self.get_output(w.Outputs.learner).preprocessors)
# remove preprocessor
self.send_signal(w.Inputs.preprocessor, None)
self.assertIn(w.preprocessors, [[], None])
self.assertNotIn(preprocessor,
self.get_output(w.Outputs.learner).preprocessors)
# set it again
preprocessor = Discretize()
self.send_signal(w.Inputs.preprocessor, preprocessor)
# check preprocessor is set
self.assertEqual(w.preprocessors, [preprocessor])
self.assertIn(preprocessor,
self.get_output(w.Outputs.learner).preprocessors)
# change preprocessor
preprocessor = Continuize()
self.send_signal(w.Inputs.preprocessor, preprocessor)
# check preprocessor is set
self.assertEqual(w.preprocessors, [preprocessor])
self.assertIn(preprocessor,
self.get_output(w.Outputs.learner).preprocessors)
class ClassificationScorer(Scorer):
"""
Base class for feature scores in a class-labeled data set.
Parameters
----------
feature : int, string, Orange.data.Variable
Feature id
data : Orange.data.Table
Data set
Attributes
----------
feature_type : Orange.data.Variable
Required type of features.
class_type : Orange.data.Variable
Required type of class variable.
"""
feature_type = DiscreteVariable
class_type = DiscreteVariable
supports_sparse_data = True
preprocessors = Scorer.preprocessors + [
Discretize(remove_const=False)
]
def score_data(self, data, feature):
instances_with_class = \
np.sum(distribution.Discrete(data, data.domain.class_var))
def score_from_contingency(f):
cont = contingency.Discrete(data, f)
return self.from_contingency(
cont, 1. - np.sum(cont.unknowns)/instances_with_class)
scores = [score_from_contingency(f) for f in data.domain.attributes]
if feature is not None:
return scores[0]
return scores
self.set_output_label_text()
def merge_data(self):
attributes = getattr(self.data.domain, 'attributes')
cls_vars = getattr(self.data.domain, 'class_vars')
metas_v = getattr(self.data.domain, 'metas')\
+ getattr(self.transformed_data.domain, 'attributes')
domain = Domain(attributes, cls_vars, metas_v)
X = self.data.X
Y = self.data.Y
metas = np.hstack((self.data.metas, self.transformed_data.X))
table = Table.from_numpy(domain, X, Y, metas)
table.name = getattr(self.data, 'name', '')
table.attributes = getattr(self.data, 'attributes', {})
table.ids = self.data.ids
return table
def send_report(self):
if self.preprocessor is not None:
self.report_items("Settings",
(("Preprocessor", self.preprocessor), ))
if self.data is not None:
self.report_data("Data", self.data)
if self.transformed_data is not None:
self.report_data("Transformed data", self.transformed_data)
if __name__ == "__main__": # pragma: no cover
WidgetPreview(OWTransform).run(set_data=Table("iris"),
set_preprocessor=Discretize())
except Exception as ex: # pylint: disable=broad-except
self.Error.pp_error(ex)
self.Outputs.transformed_data.send(self.transformed_data)
self.set_preprocessor_label_text()
self.set_output_label_text()
def send_report(self):
if self.preprocessor is not None:
self.report_items("Settings",
(("Preprocessor", self.preprocessor), ))
if self.data is not None:
self.report_data("Data", self.data)
if self.transformed_data is not None:
self.report_data("Transformed data", self.transformed_data)
if __name__ == "__main__":
from AnyQt.QtWidgets import QApplication
app = QApplication([])
ow = OWTransform()
d = Table("iris")
pp = Discretize()
ow.set_data(d)
ow.set_preprocessor(pp)
ow.handleNewSignals()
ow.show()
app.exec_()
ow.saveSettings()