def commit(self):
self.warning(1)
self.error(1)
data = distances = None
if self.data is not None:
metric = _METRICS[self.metric_idx][1]
if not any(a.is_continuous for a in self.data.domain.attributes):
self.error(1, "No continuous features")
data = None
elif (any(a.is_discrete for a in self.data.domain.attributes) or
numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(data.domain.attributes) > 0:
self.warning(1, "Ignoring discrete features")
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.error(1, "Empty data (shape == {})".format(shape))
else:
distances = metric(data, data, 1 - self.axis)
self.send("Distances", distances)
def checks(metric, data):
if data is None:
return
if issparse(data.X) and not metric.supports_sparse:
self.Error.dense_metric_sparse_data()
return
if not any(a.is_continuous for a in data.domain.attributes):
self.Error.no_continuous_features()
return
needs_preprocessing = False
if any(a.is_discrete for a in self.data.domain.attributes):
self.Warning.ignoring_discrete()
needs_preprocessing = True
if not issparse(data.X) and bn.anynan(data.X):
self.Warning.imputing_data()
needs_preprocessing = True
if needs_preprocessing:
# removes discrete features and imputes data
data = distance._preprocess(data)
if not data.X.size:
self.Error.empty_data()
return
return data
def commit(self):
self.warning(1)
self.error(1)
data = distances = None
if self.data is not None:
metric = self.available_metrics[self.metric_idx]
if isinstance(metric, distance.MahalanobisDistance):
metric.fit(self.data, axis=1 - self.axis)
if not any(a.is_continuous for a in self.data.domain.attributes):
self.error(1, "No continuous features")
data = None
elif any(a.is_discrete for a in self.data.domain.attributes) or \
(not issparse(self.data.X) and numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(
data.domain.attributes) > 0:
self.warning(1, "Ignoring discrete features")
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.error(1, "Empty data (shape == {})".format(shape))
else:
distances = metric(data, data, 1 - self.axis, impute=True)
self.send("Distances", distances)
def commit(self):
metric = METRICS[self.metric_idx]
distances = None
data = self.data
if data is not None and issparse(data.X) and \
not metric.supports_sparse:
data = None
self.clear_messages()
if data is not None:
if isinstance(metric, distance.MahalanobisDistance):
metric.fit(self.data, axis=1 - self.axis)
if not any(a.is_continuous for a in self.data.domain.attributes):
self.Error.no_continuous_features()
data = None
elif any(a.is_discrete for a in self.data.domain.attributes) or \
(not issparse(self.data.X) and numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(
data.domain.attributes) > 0:
self.Warning.ignoring_discrete()
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.Error.empty_data(shape)
else:
distances = metric(data, data, 1 - self.axis, impute=True)
self.send("Distances", distances)
def commit(self):
self.warning(1)
self.error(1)
data = distances = None
if self.data is not None:
metric = _METRICS[self.metric_idx][1]
if not any(a.is_continuous for a in self.data.domain.attributes):
self.error(1, "No continuous features")
data = None
elif (any(a.is_discrete for a in self.data.domain.attributes)
or numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(
data.domain.attributes) > 0:
self.warning(1, "Ignoring discrete features")
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.error(1, "Empty data (shape == {})".format(shape))
else:
distances = metric(data, data, 1 - self.axis, impute=True)
self.send("Distances", distances)
def commit(self):
metric = METRICS[self.metric_idx]
distances = None
data = self.data
if data is not None and issparse(data.X) and \
not metric.supports_sparse:
data = None
self.clear_messages()
if data is not None:
if isinstance(metric, distance.MahalanobisDistance):
metric.fit(self.data, axis=1-self.axis)
if not any(a.is_continuous for a in self.data.domain.attributes):
self.Error.no_continuous_features()
data = None
elif any(a.is_discrete for a in self.data.domain.attributes) or \
(not issparse(self.data.X) and numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(data.domain.attributes) > 0:
self.Warning.ignoring_discrete()
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.Error.empty_data(shape)
else:
distances = metric(data, data, 1 - self.axis, impute=True)
self.send("Distances", distances)
def commit(self):
self.warning(1)
self.error(1)
metric = METRICS[self.metric_idx]
distances = None
data = self.data
if data is not None and issparse(data.X) and \
not metric.supports_sparse:
data = None
if data is not None:
if isinstance(metric, distance.MahalanobisDistance):
metric.fit(self.data, axis=1-self.axis)
if not any(a.is_continuous for a in self.data.domain.attributes):
self.error(1, "No continuous features")
data = None
elif any(a.is_discrete for a in self.data.domain.attributes) or \
(not issparse(self.data.X) and numpy.any(numpy.isnan(self.data.X))):
data = distance._preprocess(self.data)
if len(self.data.domain.attributes) - len(data.domain.attributes) > 0:
self.warning(1, "Ignoring discrete features")
else:
data = self.data
if data is not None:
shape = (len(data), len(data.domain.attributes))
if numpy.product(shape) == 0:
self.error(1, "Empty data (shape == {})".format(shape))
else:
distances = metric(data, data, 1 - self.axis, impute=True)
self.send("Distances", distances)
def test_preprocess_multiclass(self):
table = Table('test5.tab')
new_table = _preprocess(table)
np.testing.assert_equal(new_table.Y, table.Y)
self.assertEqual([a.name for a in new_table.domain.attributes],
[a.name for a in table.domain.attributes
if a.is_continuous])
self.assertEqual(new_table.domain.class_vars, table.domain.class_vars)
def test_preprocess_multiclass(self):
table = self.test5
new_table = _preprocess(table)
np.testing.assert_equal(new_table.Y, table.Y)
self.assertEqual(
[a.name for a in new_table.domain.attributes],
[a.name for a in table.domain.attributes if a.is_continuous])
self.assertEqual(new_table.domain.class_vars, table.domain.class_vars)
def test_preprocess(self):
domain = Domain([
ContinuousVariable("c"),
DiscreteVariable("d", values=['a', 'b'])
], [DiscreteVariable("cls", values=['e', 'f'])], [StringVariable("m")])
table = Table(domain, [[1, 'a', 'e', 'm1'], [2, 'b', 'f', 'm2']])
new_table = _preprocess(table)
np.testing.assert_equal(new_table.X, table.X[:, 0].reshape(2, 1))
np.testing.assert_equal(new_table.Y, table.Y)
np.testing.assert_equal(new_table.metas, table.metas)
self.assertEqual(
[a.name for a in new_table.domain.attributes],
[a.name for a in table.domain.attributes if a.is_continuous])
self.assertEqual(new_table.domain.class_vars, table.domain.class_vars)
self.assertEqual(new_table.domain.metas, table.domain.metas)
def commit(self):
self.warning(1)
self.error(1)
distances = None
if self.data is not None:
metric = _METRICS[self.metric_idx][1]
data = distance._preprocess(self.data)
if len(data.domain.attributes) == 0:
self.error(1, "No continuous features")
elif len(self.data.domain.attributes) - len(data.domain.attributes) > 0:
self.warning(1, "Ignoring categoric features")
if len(data.domain.attributes) != 0:
distances = metric(data, data, 1-self.axis)
self.send("Distances", distances)
def test_preprocess(self):
domain = Domain([ContinuousVariable("c"),
DiscreteVariable("d", values=['a', 'b'])],
[DiscreteVariable("cls", values=['e', 'f'])],
[StringVariable("m")])
table = Table(domain, [[1, 'a', 'e', 'm1'],
[2, 'b', 'f', 'm2']])
new_table = _preprocess(table)
np.testing.assert_equal(new_table.X, table.X[:, 0].reshape(2, 1))
np.testing.assert_equal(new_table.Y, table.Y)
np.testing.assert_equal(new_table.metas, table.metas)
self.assertEqual([a.name for a in new_table.domain.attributes],
[a.name for a in table.domain.attributes
if a.is_continuous])
self.assertEqual(new_table.domain.class_vars, table.domain.class_vars)
self.assertEqual(new_table.domain.metas, table.domain.metas)
def commit(self):
self.warning(1)
self.error(1)
distances = None
if self.data is not None:
metric = _METRICS[self.metric_idx][1]
data = distance._preprocess(self.data)
if len(data.domain.attributes) == 0:
self.error(1, "No continuous features")
elif len(self.data.domain.attributes) - len(
data.domain.attributes) > 0:
self.warning(1, "Ignoring categoric features")
if len(data.domain.attributes) != 0:
distances = metric(data, data, 1 - self.axis)
self.send("Distances", distances)
def compute_distances(self, metric, data):
self.clear_messages()
if data is None:
return
if issparse(data.X) and not metric.supports_sparse:
self.Error.dense_metric_sparse_data()
return
if not any(a.is_continuous for a in data.domain.attributes):
self.Error.no_continuous_features()
return
needs_preprocessing = False
if any(a.is_discrete for a in self.data.domain.attributes):
self.Warning.ignoring_discrete()
needs_preprocessing = True
if not issparse(data.X) and bn.anynan(data.X):
self.Warning.imputing_data()
needs_preprocessing = True
if needs_preprocessing:
# removes discrete features and imputes data
data = distance._preprocess(data)
if not data.X.size:
self.Error.empty_data()
return
if isinstance(metric, distance.MahalanobisDistance):
n, m = data.X.shape
if self.axis == 1:
n, m = m, n
if isinstance(metric, distance.MahalanobisDistance):
# Mahalanobis distance has to be trained before it can be used
# to compute distances
try:
metric.fit(data, axis=1 - self.axis)
except (ValueError, MemoryError) as e:
self.Error.mahalanobis_error(e)
return
return metric(data, data, 1 - self.axis, impute=True)
def compute_distances(self, metric, data):
self.clear_messages()
if data is None:
return
if issparse(data.X) and not metric.supports_sparse:
self.Error.dense_metric_sparse_data()
return
if not any(a.is_continuous for a in data.domain.attributes):
self.Error.no_continuous_features()
return
needs_preprocessing = False
if any(a.is_discrete for a in self.data.domain.attributes):
self.Warning.ignoring_discrete()
needs_preprocessing = True
if not issparse(data.X) and bn.anynan(data.X):
self.Warning.imputing_data()
needs_preprocessing = True
if needs_preprocessing:
# removes discrete features and imputes data
data = distance._preprocess(data)
if not data.X.size:
self.Error.empty_data(data.X.shape)
return
if isinstance(metric, distance.MahalanobisDistance):
n, m = data.X.shape
if self.axis == 1:
n, m = m, n
if n <= m:
self.Error.too_few_observations()
return
if isinstance(metric, distance.MahalanobisDistance):
# Mahalanobis distance has to be trained before it can be used
# to compute distances
metric.fit(data, axis=1 - self.axis)
return metric(data, data, 1 - self.axis, impute=True)
def test_preprocess(self):
domain = Domain(
[
ContinuousVariable("c"),
DiscreteVariable("d", values=["a", "b"])
],
[DiscreteVariable("cls", values=["e", "f"])],
[StringVariable("m")],
)
table = Table(domain, [[1, "a", "e", "m1"], [2, "b", "f", "m2"]])
new_table = _preprocess(table)
np.testing.assert_equal(new_table.X, table.X[:, 0].reshape(2, 1))
np.testing.assert_equal(new_table.Y, table.Y)
np.testing.assert_equal(new_table.metas, table.metas)
self.assertEqual(
[a.name for a in new_table.domain.attributes],
[a.name for a in table.domain.attributes if a.is_continuous],
)
self.assertEqual(new_table.domain.class_vars, table.domain.class_vars)
self.assertEqual(new_table.domain.metas, table.domain.metas)
def test_preprocess_impute(self):
new_table = _preprocess(self.test5)
self.assertFalse(np.isnan(new_table.X).any())
def test_preprocess_impute(self):
table = Table('test5.tab')
new_table = _preprocess(table)
self.assertFalse(np.isnan(new_table.X).any())
def test_preprocess_impute(self):
table = Table('test5.tab')
new_table = _preprocess(table)
self.assertFalse(np.isnan(new_table.X).any())
def test_preprocess_impute(self):
new_table = _preprocess(self.test5)
self.assertFalse(np.isnan(new_table.X).any())
def paint(self, painter, *args):
if self._line is None:
self.boundingRect()
painter.save()
painter.setPen(self.pen())
painter.drawLine(self._line)
painter.restore()
def clusters_at_height(root, height):
"""Return a list of clusters by cutting the clustering at `height`.
"""
lower = set()
cluster_list = []
for cl in preorder(root):
if cl in lower:
continue
if cl.value.height < height:
cluster_list.append(cl)
lower.update(preorder(cl))
return cluster_list
if __name__ == "__main__": # pragma: no cover
from Orange import distance
data = Orange.data.Table("iris")
matrix = distance.Euclidean(distance._preprocess(data))
WidgetPreview(OWHierarchicalClustering).run(matrix)
