def test_variance_drop_all(self):
data, label = get_data_label(load_boston())
method = SelectionMethod.Variance(threshold=100000)
selector = Selective(method)
try:
selector.fit(data)
selector.transform(data)
except ValueError:
pass
def test_correlation(self):
data, label = get_data_label(load_boston())
method = SelectionMethod.Correlation(0.60)
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
self.assertListEqual(list(subset.columns), ['CRIM', 'ZN', 'INDUS', 'CHAS', 'RM', 'PTRATIO', 'B'])
def test_correlation_small(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=["CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"])
method = SelectionMethod.Correlation(0.60)
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
self.assertListEqual(list(subset.columns), ['CRIM', 'ZN', 'AGE', 'B'])
def test_tree_invalid_num_features(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.TreeBased(num_features=100)
selector = Selective(method)
with self.assertRaises(ValueError):
selector.fit(data, label)
def test_variance_zero_threshold(self):
data, label = get_data_label(load_boston())
method = SelectionMethod.Variance(threshold=0)
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(data.shape[1], 13)
self.assertEqual(subset.shape[1], 13)
def test_anova_classif_top_percentile(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.Statistical(num_features=0.5)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
self.assertListEqual(list(subset.columns), ['petal length (cm)', 'petal width (cm)'])
def test_chi_regress_top_percentile_invalid(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.Statistical(num_features=0.6,
method="chi_square")
selector = Selective(method)
with self.assertRaises(TypeError):
selector.fit(data, label)
def test_tree_estimator_lightgbm_classif_top_k(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.TreeBased(
num_features=2,
estimator=XGBClassifier(random_state=Constants.default_seed))
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
def test_variance_drop_target(self):
data, label = get_data_label(load_boston())
method = SelectionMethod.Variance(threshold=85)
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(data.shape[1], 13)
self.assertEqual(subset.shape[1], 4)
self.assertListEqual(list(subset.columns), ['ZN', 'AGE', 'TAX', 'B'])
def test_tree_classif_top_k(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.TreeBased(num_features=2)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
self.assertListEqual(list(subset.columns),
['petal length (cm)', 'petal width (cm)'])
def test_anova_regress_top_percentile_all(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=["CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"])
method = SelectionMethod.Statistical(num_features=1.0)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(data.shape[1], subset.shape[1])
self.assertListEqual(list(data.columns), list(subset.columns))
def test_vif_top_k_no_label(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.Statistical(num_features=2,
method="variance_inflation")
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
self.assertListEqual(list(subset.columns),
['sepal width (cm)', 'petal width (cm)'])
def test_linear_regress_top_k_all(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.Linear(num_features=5)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertListEqual(list(data.columns), list(subset.columns))
def test_tree_regress_top_percentile(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.TreeBased(num_features=0.6)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 3)
self.assertListEqual(list(subset.columns), ['CRIM', 'AGE', 'LSTAT'])
def test_tree_estimator_adaboost_classif_top_k(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.TreeBased(
num_features=2,
estimator=AdaBoostClassifier(random_state=Constants.default_seed))
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
self.assertListEqual(list(subset.columns),
['petal length (cm)', 'petal width (cm)'])
def test_linear_classif_top_percentile_all(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.Linear(num_features=1.0)
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 4)
self.assertListEqual(list(subset.columns), [
'sepal length (cm)', 'sepal width (cm)', 'petal length (cm)',
'petal width (cm)'
])
def test_ridge_regress_top_k(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.Linear(num_features=3, regularization="ridge")
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 3)
self.assertListEqual(list(subset.columns), ['CRIM', 'AGE', 'LSTAT'])
def test_tree_estimator_xgboost_regress_top_k(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.TreeBased(
num_features=3,
estimator=XGBRegressor(random_state=Constants.default_seed))
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 3)
def test_mutual_classif_top_k_all(self):
data, label = get_data_label(load_iris())
method = SelectionMethod.Statistical(num_features=4,
method="mutual_info")
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 4)
self.assertListEqual(list(subset.columns), [
'sepal length (cm)', 'sepal width (cm)', 'petal length (cm)',
'petal width (cm)'
])
def test_variance_lt1(self):
data, label = get_data_label(load_boston())
method = SelectionMethod.Variance(threshold=1.0)
selector = Selective(method)
selector.fit(data)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(data.shape[1], 13)
self.assertEqual(subset.shape[1], 10)
self.assertListEqual(list(subset.columns), [
'CRIM', 'ZN', 'INDUS', 'AGE', 'DIS', 'RAD', 'TAX', 'PTRATIO', 'B',
'LSTAT'
])
def test_vif_top_k_regression(self):
data, label = get_data_label(load_boston())
data = data.drop(columns=[
"CHAS", "NOX", "RM", "DIS", "RAD", "TAX", "PTRATIO", "INDUS"
])
method = SelectionMethod.Statistical(num_features=2,
method="variance_inflation")
selector = Selective(method)
selector.fit(data, label)
subset = selector.transform(data)
# Reduced columns
self.assertEqual(subset.shape[1], 2)
self.assertListEqual(list(subset.columns), ['CRIM', 'ZN'])
