def test_leave_discrete(self):
s = [0] * 50 + [1] * 50
X1 = np.array(s).reshape((100, 1))
X2 = np.arange(100).reshape((100, 1))
X3 = np.ones((100, 1))
X = np.hstack([X1, X2, X3])
domain = data.Domain([
data.DiscreteVariable("a", values="MF"),
data.ContinuousVariable("b"),
data.DiscreteVariable("c", values="AB")
], data.ContinuousVariable("d"))
table = data.Table(domain, X, X1)
dom = discretize.DomainDiscretizer(table)
self.assertIs(dom[0], table.domain[0])
self.assertEqual(dom[1].compute_value.points, [24.5, 49.5, 74.5])
self.assertIs(dom[2], table.domain[2])
self.assertIs(dom.class_var, table.domain.class_var)
domain = data.Domain([
data.DiscreteVariable("a", values="MF"),
data.ContinuousVariable("b"),
data.DiscreteVariable("c", values="AB")
], data.DiscreteVariable("d"))
table = data.Table(domain, X, X1)
dom = discretize.DomainDiscretizer(table)
self.assertIs(dom[0], table.domain[0])
self.assertEqual(dom[1].compute_value.points, [24.5, 49.5, 74.5])
self.assertIs(dom[2], table.domain[2])
self.assertIs(dom.class_var, table.domain.class_var)
def test_method(self):
dom = discretize.DomainDiscretizer(self.table_class)
self.assertEqual(len(dom[1].values), 4)
dom = discretize.DomainDiscretizer(self.table_class,
method=discretize.EqualWidth(n=2))
self.assertEqual(len(dom[1].values), 2)
def test_discretize_class(self):
dom = discretize.DomainDiscretizer(self.table_class)
self.assertIs(dom.class_var, self.table_class.domain.class_var)
dom = discretize.DomainDiscretizer(self.table_class,
discretize_class=True)
self.assertIs(dom.class_var, self.table_class.domain.class_var)
def test_fixed(self):
dom = discretize.DomainDiscretizer(self.table_no_class,
method=discretize.EqualWidth(n=2),
fixed={"Feature 2": [1, 11]})
self.assertEqual(len(dom.attributes), 2)
self.assertEqual(dom[0].compute_value.points, [0.5])
self.assertEqual(dom[1].compute_value.points, [6])
def test_discretize_exclude_constant(self):
dom = discretize.DomainDiscretizer(self.table_no_class)
self.assertEqual(len(dom.attributes), 2)
self.assertEqual(dom[0].compute_value.points, [0.5])
self.assertEqual(dom[1].compute_value.points, [24.5, 49.5, 74.5])
dom = discretize.DomainDiscretizer(self.table_no_class, clean=False)
self.assertEqual(len(dom.attributes), 3)
self.assertEqual(dom[0].compute_value.points, [0.5])
self.assertEqual(dom[1].compute_value.points, [24.5, 49.5, 74.5])
self.assertEqual(dom[2].compute_value.points, [])
dom = discretize.DomainDiscretizer(self.table_class)
self.assertEqual(len(dom.attributes), 2)
self.assertEqual(dom[0].compute_value.points, [0.5])
self.assertEqual(dom[1].compute_value.points, [24.5, 49.5, 74.5])
def test_remove_mapping_after_compute_value(self):
housing = Table("housing")
method = discretize.EqualFreq(n=3)
discretizer = discretize.DomainDiscretizer(discretize_class=True,
method=method)
domain = discretizer(housing)
data = housing.transform(domain)
val12 = np.nonzero(data.Y > 0)[0]
data = data[val12]
remover = Remove(class_flags=Remove.RemoveUnusedValues)
cleaned = remover(data)
np.testing.assert_equal(cleaned.Y, data.Y - 1)
def test_purge_discretized(self):
housing = Table("housing")
method = discretize.EqualFreq(n=3)
discretizer = discretize.DomainDiscretizer(discretize_class=True,
method=method)
domain = discretizer(housing)
data = housing.transform(domain)
widget = self.widget_with_context(domain, [["MEDV", 101, 2, (2, 3)]])
widget.purge_classes = True
self.send_signal(widget.Inputs.data, data)
out = self.get_output(widget.Outputs.matching_data)
expected = data.Y[(data.Y == 1) + (data.Y == 2)]
expected = (expected == 2).astype(float)
np.testing.assert_equal(out.Y, expected)
