def test_feature_importances_2d_coef():
X, y = datasets.make_classification(n_samples=1000,
n_features=10,
n_informative=3,
n_redundant=0,
n_repeated=0,
shuffle=False,
random_state=0,
n_classes=4)
est = LogisticRegression()
for threshold, func in zip(["mean", "median"], [np.mean, np.median]):
for order in [1, 2, np.inf]:
# Fit SelectFromModel a multi-class problem
transformer = SelectFromModel(estimator=LogisticRegression(),
threshold=threshold,
norm_order=order)
transformer.fit(X, y)
assert_true(hasattr(transformer.estimator_, 'coef_'))
X_new = transformer.transform(X)
assert_less(X_new.shape[1], X.shape[1])
# Manually check that the norm is correctly performed
est.fit(X, y)
importances = norm(est.coef_, axis=0, ord=order)
feature_mask = importances > func(importances)
assert_array_equal(X_new, X[:, feature_mask])
def test_feature_importances_2d_coef():
X, y = datasets.make_classification(
n_samples=1000,
n_features=10,
n_informative=3,
n_redundant=0,
n_repeated=0,
shuffle=False,
random_state=0,
n_classes=4,
)
est = LogisticRegression()
for threshold, func in zip(["mean", "median"], [np.mean, np.median]):
for order in [1, 2, np.inf]:
# Fit SelectFromModel a multi-class problem
transformer = SelectFromModel(estimator=LogisticRegression(), threshold=threshold, norm_order=order)
transformer.fit(X, y)
assert_true(hasattr(transformer.estimator_, "coef_"))
X_new = transformer.transform(X)
assert_less(X_new.shape[1], X.shape[1])
# Manually check that the norm is correctly performed
est.fit(X, y)
importances = norm(est.coef_, axis=0, ord=order)
feature_mask = importances > func(importances)
assert_array_equal(X_new, X[:, feature_mask])
def test_norm():
X = np.array([[-2, 4, 5], [1, 3, -4], [0, 0, 8], [0, 0, 0]]).astype(float)
# Test various axis and order
assert_equal(math.sqrt(135), norm(X))
assert_array_equal(np.array([math.sqrt(5),
math.sqrt(25),
math.sqrt(105)]), norm(X, axis=0))
assert_array_equal(np.array([3, 7, 17]), norm(X, axis=0, ord=1))
assert_array_equal(np.array([2, 4, 8]), norm(X, axis=0, ord=np.inf))
assert_array_equal(np.array([0, 0, 0]), norm(X, axis=0, ord=-np.inf))
assert_array_equal(np.array([11, 8, 8, 0]), norm(X, axis=1, ord=1))
# Test shapes
assert_equal((), norm(X).shape)
assert_equal((3, ), norm(X, axis=0).shape)
assert_equal((4, ), norm(X, axis=1).shape)
def test_norm():
X = np.array([[-2, 4, 5],
[1, 3, -4],
[0, 0, 8],
[0, 0, 0]]).astype(float)
# Test various axis and order
assert_equal(math.sqrt(135), norm(X))
assert_array_equal(
np.array([math.sqrt(5), math.sqrt(25), math.sqrt(105)]),
norm(X, axis=0)
)
assert_array_equal(np.array([3, 7, 17]), norm(X, axis=0, ord=1))
assert_array_equal(np.array([2, 4, 8]), norm(X, axis=0, ord=np.inf))
assert_array_equal(np.array([0, 0, 0]), norm(X, axis=0, ord=-np.inf))
assert_array_equal(np.array([11, 8, 8, 0]), norm(X, axis=1, ord=1))
# Test shapes
assert_equal((), norm(X).shape)
assert_equal((3,), norm(X, axis=0).shape)
assert_equal((4,), norm(X, axis=1).shape)
