def test_ensemble_heterogeneous_estimators_name_validation(X, y, Ensemble):
# raise an error when the name contains dunder
if issubclass(Ensemble, ClassifierMixin):
estimators = [('lr__', LogisticRegression())]
else:
estimators = [('lr__', LinearRegression())]
ensemble = Ensemble(estimators=estimators)
err_msg = r"Estimator names must not contain __: got \['lr__'\]"
with pytest.raises(ValueError, match=err_msg):
ensemble.fit(X, y)
# raise an error when the name is not unique
if issubclass(Ensemble, ClassifierMixin):
estimators = [('lr', LogisticRegression()),
('lr', LogisticRegression())]
else:
estimators = [('lr', LinearRegression()), ('lr', LinearRegression())]
ensemble = Ensemble(estimators=estimators)
err_msg = r"Names provided are not unique: \['lr', 'lr'\]"
with pytest.raises(ValueError, match=err_msg):
ensemble.fit(X, y)
# raise an error when the name conflicts with the parameters
if issubclass(Ensemble, ClassifierMixin):
estimators = [('estimators', LogisticRegression())]
else:
estimators = [('estimators', LinearRegression())]
ensemble = Ensemble(estimators=estimators)
err_msg = "Estimator names conflict with constructor arguments"
with pytest.raises(ValueError, match=err_msg):
ensemble.fit(X, y)
def test_partial_dependence_easy_target(est, power):
# If the target y only depends on one feature in an obvious way (linear or
# quadratic) then the partial dependence for that feature should reflect
# it.
# We here fit a linear regression_data model (with polynomial features if
# needed) and compute r_squared to check that the partial dependence
# correctly reflects the target.
rng = np.random.RandomState(0)
n_samples = 200
target_variable = 2
X = rng.normal(size=(n_samples, 5))
y = X[:, target_variable]**power
est.fit(X, y)
averaged_predictions, values = partial_dependence(
est, features=[target_variable], X=X, grid_resolution=1000)
new_X = values[0].reshape(-1, 1)
new_y = averaged_predictions[0]
# add polynomial features if needed
new_X = PolynomialFeatures(degree=power).fit_transform(new_X)
lr = LinearRegression().fit(new_X, new_y)
r2 = r2_score(new_y, lr.predict(new_X))
assert r2 > .99
def test_plot_partial_dependence_passing_numpy_axes(pyplot, clf_boston,
boston):
grid_resolution = 25
feature_names = boston.feature_names.tolist()
disp1 = plot_partial_dependence(clf_boston, boston.data,
['CRIM', 'ZN'],
grid_resolution=grid_resolution,
feature_names=feature_names)
assert disp1.axes_.shape == (1, 2)
assert disp1.axes_[0, 0].get_ylabel() == "Partial dependence"
assert disp1.axes_[0, 1].get_ylabel() == ""
assert len(disp1.axes_[0, 0].get_lines()) == 1
assert len(disp1.axes_[0, 1].get_lines()) == 1
lr = LinearRegression()
lr.fit(boston.data, boston.target)
disp2 = plot_partial_dependence(lr, boston.data,
['CRIM', 'ZN'],
grid_resolution=grid_resolution,
feature_names=feature_names,
ax=disp1.axes_)
assert np.all(disp1.axes_ == disp2.axes_)
assert len(disp2.axes_[0, 0].get_lines()) == 2
assert len(disp2.axes_[0, 1].get_lines()) == 2
def test_omp_reaches_least_squares():
# Use small simple data; it's a sanity check but OMP can stop early
rng = check_random_state(0)
n_samples, n_features = (10, 8)
n_targets = 3
X = rng.randn(n_samples, n_features)
Y = rng.randn(n_samples, n_targets)
omp = OrthogonalMatchingPursuit(n_nonzero_coefs=n_features)
lstsq = LinearRegression()
omp.fit(X, Y)
lstsq.fit(X, Y)
assert_array_almost_equal(omp.coef_, lstsq.coef_)
def test_linear_regression_sparse(random_state=0):
# Test that linear regression also works with sparse data
random_state = check_random_state(random_state)
for i in range(10):
n = 100
X = sparse.eye(n, n)
beta = random_state.rand(n)
y = X * beta[:, np.newaxis]
ols = LinearRegression()
ols.fit(X, y.ravel())
assert_array_almost_equal(beta, ols.coef_ + ols.intercept_)
assert_array_almost_equal(ols.predict(X) - y.ravel(), 0)
def test_transform_target_regressor_invertible():
X, y = friedman
regr = TransformedTargetRegressor(regressor=LinearRegression(),
func=np.sqrt,
inverse_func=np.log,
check_inverse=True)
assert_warns_message(
UserWarning, "The provided functions or transformer"
" are not strictly inverse of each other.", regr.fit, X, y)
regr = TransformedTargetRegressor(regressor=LinearRegression(),
func=np.sqrt,
inverse_func=np.log)
regr.set_params(check_inverse=False)
assert_no_warnings(regr.fit, X, y)
def test_transform_target_regressor_2d_transformer(X, y):
# Check consistency with transformer accepting only 2D array and a 1D/2D y
# array.
transformer = StandardScaler()
regr = TransformedTargetRegressor(regressor=LinearRegression(),
transformer=transformer)
y_pred = regr.fit(X, y).predict(X)
assert y.shape == y_pred.shape
# consistency forward transform
if y.ndim == 1: # create a 2D array and squeeze results
y_tran = regr.transformer_.transform(y.reshape(-1, 1)).squeeze()
else:
y_tran = regr.transformer_.transform(y)
_check_standard_scaled(y, y_tran)
assert y.shape == y_pred.shape
# consistency inverse transform
assert_allclose(y, regr.transformer_.inverse_transform(y_tran).squeeze())
# consistency of the regressor
lr = LinearRegression()
transformer2 = clone(transformer)
if y.ndim == 1: # create a 2D array and squeeze results
lr.fit(X, transformer2.fit_transform(y.reshape(-1, 1)).squeeze())
else:
lr.fit(X, transformer2.fit_transform(y))
y_lr_pred = lr.predict(X)
assert_allclose(y_pred, transformer2.inverse_transform(y_lr_pred))
assert_allclose(regr.regressor_.coef_, lr.coef_)
def test_plot_partial_dependence_fig_deprecated(pyplot):
# Make sure fig object is correctly used if not None
X, y = make_regression(n_samples=50, random_state=0)
clf = LinearRegression()
clf.fit(X, y)
fig = pyplot.figure()
grid_resolution = 25
msg = ("The fig parameter is deprecated in version 0.22 and will be "
"removed in version 0.24")
with pytest.warns(FutureWarning, match=msg):
plot_partial_dependence(
clf, X, [0, 1], target=0, grid_resolution=grid_resolution, fig=fig)
assert pyplot.gcf() is fig
def test_stacking_regressor_diabetes(cv, final_estimator, predict_params,
passthrough):
# prescale the data to avoid convergence warning without using a pipeline
# for later assert
X_train, X_test, y_train, _ = train_test_split(scale(X_diabetes),
y_diabetes,
random_state=42)
estimators = [('lr', LinearRegression()), ('svr', LinearSVR())]
reg = StackingRegressor(estimators=estimators,
final_estimator=final_estimator,
cv=cv,
passthrough=passthrough)
reg.fit(X_train, y_train)
result = reg.predict(X_test, **predict_params)
expected_result_length = 2 if predict_params else 1
if predict_params:
assert len(result) == expected_result_length
X_trans = reg.transform(X_test)
expected_column_count = 12 if passthrough else 2
assert X_trans.shape[1] == expected_column_count
if passthrough:
assert_allclose(X_test, X_trans[:, -10:])
reg.set_params(lr='drop')
reg.fit(X_train, y_train)
reg.predict(X_test)
X_trans = reg.transform(X_test)
expected_column_count_drop = 11 if passthrough else 1
assert X_trans.shape[1] == expected_column_count_drop
if passthrough:
assert_allclose(X_test, X_trans[:, -10:])
def test_transform_target_regressor_functions_multioutput():
X = friedman[0]
y = np.vstack((friedman[1], friedman[1]**2 + 1)).T
regr = TransformedTargetRegressor(regressor=LinearRegression(),
func=np.log,
inverse_func=np.exp)
y_pred = regr.fit(X, y).predict(X)
# check the transformer output
y_tran = regr.transformer_.transform(y)
assert_allclose(np.log(y), y_tran)
assert_allclose(y, regr.transformer_.inverse_transform(y_tran))
assert y.shape == y_pred.shape
assert_allclose(y_pred, regr.inverse_func(regr.regressor_.predict(X)))
# check the regressor output
lr = LinearRegression().fit(X, regr.func(y))
assert_allclose(regr.regressor_.coef_.ravel(), lr.coef_.ravel())
def test_transform_target_regressor_error():
X, y = friedman
# provide a transformer and functions at the same time
regr = TransformedTargetRegressor(regressor=LinearRegression(),
transformer=StandardScaler(),
func=np.exp,
inverse_func=np.log)
with pytest.raises(ValueError,
match="'transformer' and functions"
" 'func'/'inverse_func' cannot both be set."):
regr.fit(X, y)
# fit with sample_weight with a regressor which does not support it
sample_weight = np.ones((y.shape[0], ))
regr = TransformedTargetRegressor(regressor=Lasso(),
transformer=StandardScaler())
with pytest.raises(TypeError,
match=r"fit\(\) got an unexpected "
"keyword argument 'sample_weight'"):
regr.fit(X, y, sample_weight=sample_weight)
# func is given but inverse_func is not
regr = TransformedTargetRegressor(func=np.exp)
with pytest.raises(ValueError,
match="When 'func' is provided, "
"'inverse_func' must also be provided"):
regr.fit(X, y)
def test_subsamples():
X, y, w, c = gen_toy_problem_4d()
theil_sen = TheilSenRegressor(n_subsamples=X.shape[0],
random_state=0).fit(X, y)
lstq = LinearRegression().fit(X, y)
# Check for exact the same results as Least Squares
assert_array_almost_equal(theil_sen.coef_, lstq.coef_, 9)
def test_transform_target_regressor_functions():
X, y = friedman
regr = TransformedTargetRegressor(regressor=LinearRegression(),
func=np.log,
inverse_func=np.exp)
y_pred = regr.fit(X, y).predict(X)
# check the transformer output
y_tran = regr.transformer_.transform(y.reshape(-1, 1)).squeeze()
assert_allclose(np.log(y), y_tran)
assert_allclose(
y,
regr.transformer_.inverse_transform(y_tran.reshape(-1, 1)).squeeze())
assert y.shape == y_pred.shape
assert_allclose(y_pred, regr.inverse_func(regr.regressor_.predict(X)))
# check the regressor output
lr = LinearRegression().fit(X, regr.func(y))
assert_allclose(regr.regressor_.coef_.ravel(), lr.coef_.ravel())
def test_theil_sen_1d():
X, y, w, c = gen_toy_problem_1d()
# Check that Least Squares fails
lstq = LinearRegression().fit(X, y)
assert np.abs(lstq.coef_ - w) > 0.9
# Check that Theil-Sen works
theil_sen = TheilSenRegressor(random_state=0).fit(X, y)
assert_array_almost_equal(theil_sen.coef_, w, 1)
assert_array_almost_equal(theil_sen.intercept_, c, 1)
def test_theil_sen_1d_no_intercept():
X, y, w, c = gen_toy_problem_1d(intercept=False)
# Check that Least Squares fails
lstq = LinearRegression(fit_intercept=False).fit(X, y)
assert np.abs(lstq.coef_ - w - c) > 0.5
# Check that Theil-Sen works
theil_sen = TheilSenRegressor(fit_intercept=False,
random_state=0).fit(X, y)
assert_array_almost_equal(theil_sen.coef_, w + c, 1)
assert_almost_equal(theil_sen.intercept_, 0.)
def test_fit_intercept():
# Test assertions on betas shape.
X2 = np.array([[0.38349978, 0.61650022], [0.58853682, 0.41146318]])
X3 = np.array([[0.27677969, 0.70693172, 0.01628859],
[0.08385139, 0.20692515, 0.70922346]])
y = np.array([1, 1])
lr2_without_intercept = LinearRegression(fit_intercept=False).fit(X2, y)
lr2_with_intercept = LinearRegression().fit(X2, y)
lr3_without_intercept = LinearRegression(fit_intercept=False).fit(X3, y)
lr3_with_intercept = LinearRegression().fit(X3, y)
assert (
lr2_with_intercept.coef_.shape == lr2_without_intercept.coef_.shape)
assert (
lr3_with_intercept.coef_.shape == lr3_without_intercept.coef_.shape)
assert (
lr2_without_intercept.coef_.ndim == lr3_without_intercept.coef_.ndim)
def test_ransac_stop_score():
base_estimator = LinearRegression()
ransac_estimator = RANSACRegressor(base_estimator,
min_samples=2,
residual_threshold=5,
stop_score=0,
random_state=0)
ransac_estimator.fit(X, y)
assert ransac_estimator.n_trials_ == 1
def test_theil_sen_2d():
X, y, w, c = gen_toy_problem_2d()
# Check that Least Squares fails
lstq = LinearRegression().fit(X, y)
assert norm(lstq.coef_ - w) > 1.0
# Check that Theil-Sen works
theil_sen = TheilSenRegressor(max_subpopulation=1e3,
random_state=0).fit(X, y)
assert_array_almost_equal(theil_sen.coef_, w, 1)
assert_array_almost_equal(theil_sen.intercept_, c, 1)
def test_regressormixin_score_multioutput():
from sklearn_lib.linear_model import LinearRegression
# no warnings when y_type is continuous
X = [[1], [2], [3]]
y = [1, 2, 3]
reg = LinearRegression().fit(X, y)
assert_no_warnings(reg.score, X, y)
# warn when y_type is continuous-multioutput
y = [[1, 2], [2, 3], [3, 4]]
reg = LinearRegression().fit(X, y)
msg = ("The default value of multioutput (not exposed in "
"score method) will change from 'variance_weighted' "
"to 'uniform_average' in 0.23 to keep consistent "
"with 'metrics.r2_score'. To specify the default "
"value manually and avoid the warning, please "
"either call 'metrics.r2_score' directly or make a "
"custom scorer with 'metrics.make_scorer' (the "
"built-in scorer 'r2' uses "
"multioutput='uniform_average').")
assert_warns_message(FutureWarning, msg, reg.score, X, y)
def test_linear_regression_sample_weights():
# TODO: loop over sparse data as well
rng = np.random.RandomState(0)
# It would not work with under-determined systems
for n_samples, n_features in ((6, 5), ):
y = rng.randn(n_samples)
X = rng.randn(n_samples, n_features)
sample_weight = 1.0 + rng.rand(n_samples)
for intercept in (True, False):
# LinearRegression with explicit sample_weight
reg = LinearRegression(fit_intercept=intercept)
reg.fit(X, y, sample_weight=sample_weight)
coefs1 = reg.coef_
inter1 = reg.intercept_
assert reg.coef_.shape == (X.shape[1], ) # sanity checks
assert reg.score(X, y) > 0.5
# Closed form of the weighted least square
# theta = (X^T W X)^(-1) * X^T W y
W = np.diag(sample_weight)
if intercept is False:
X_aug = X
else:
dummy_column = np.ones(shape=(n_samples, 1))
X_aug = np.concatenate((dummy_column, X), axis=1)
coefs2 = linalg.solve(
X_aug.T.dot(W).dot(X_aug),
X_aug.T.dot(W).dot(y))
if intercept is False:
assert_array_almost_equal(coefs1, coefs2)
else:
assert_array_almost_equal(coefs1, coefs2[1:])
assert_almost_equal(inter1, coefs2[0])
def test_classes_property():
X = iris.data
y = iris.target
reg = make_pipeline(SelectKBest(k=1), LinearRegression())
reg.fit(X, y)
assert_raises(AttributeError, getattr, reg, "classes_")
clf = make_pipeline(SelectKBest(k=1), LogisticRegression(random_state=0))
assert_raises(AttributeError, getattr, clf, "classes_")
clf.fit(X, y)
assert_array_equal(clf.classes_, np.unique(y))
def test_linear_regression_multiple_outcome(random_state=0):
# Test multiple-outcome linear regressions
X, y = make_regression(random_state=random_state)
Y = np.vstack((y, y)).T
n_features = X.shape[1]
reg = LinearRegression()
reg.fit((X), Y)
assert reg.coef_.shape == (2, n_features)
Y_pred = reg.predict(X)
reg.fit(X, y)
y_pred = reg.predict(X)
assert_array_almost_equal(np.vstack((y_pred, y_pred)).T, Y_pred, decimal=3)
def test_linear_regression():
# Test LinearRegression on a simple dataset.
# a simple dataset
X = [[1], [2]]
Y = [1, 2]
reg = LinearRegression()
reg.fit(X, Y)
assert_array_almost_equal(reg.coef_, [1])
assert_array_almost_equal(reg.intercept_, [0])
assert_array_almost_equal(reg.predict(X), [1, 2])
# test it also for degenerate input
X = [[1]]
Y = [0]
reg = LinearRegression()
reg.fit(X, Y)
assert_array_almost_equal(reg.coef_, [0])
assert_array_almost_equal(reg.intercept_, [0])
assert_array_almost_equal(reg.predict(X), [0])
def test_linear_regression_sparse_multiple_outcome(random_state=0):
# Test multiple-outcome linear regressions with sparse data
random_state = check_random_state(random_state)
X, y = make_sparse_uncorrelated(random_state=random_state)
X = sparse.coo_matrix(X)
Y = np.vstack((y, y)).T
n_features = X.shape[1]
ols = LinearRegression()
ols.fit(X, Y)
assert ols.coef_.shape == (2, n_features)
Y_pred = ols.predict(X)
ols.fit(X, y.ravel())
y_pred = ols.predict(X)
assert_array_almost_equal(np.vstack((y_pred, y_pred)).T, Y_pred, decimal=3)
def test_linear_regression_sparse_equal_dense(normalize, fit_intercept):
# Test that linear regression agrees between sparse and dense
rng = check_random_state(0)
n_samples = 200
n_features = 2
X = rng.randn(n_samples, n_features)
X[X < 0.1] = 0.
Xcsr = sparse.csr_matrix(X)
y = rng.rand(n_samples)
params = dict(normalize=normalize, fit_intercept=fit_intercept)
clf_dense = LinearRegression(**params)
clf_sparse = LinearRegression(**params)
clf_dense.fit(X, y)
clf_sparse.fit(Xcsr, y)
assert clf_dense.intercept_ == pytest.approx(clf_sparse.intercept_)
assert_allclose(clf_dense.coef_, clf_sparse.coef_)
def test_ransac_predict():
X = np.arange(100)[:, None]
y = np.zeros((100, ))
y[0] = 1
y[1] = 100
base_estimator = LinearRegression()
ransac_estimator = RANSACRegressor(base_estimator,
min_samples=2,
residual_threshold=0.5,
random_state=0)
ransac_estimator.fit(X, y)
assert_array_equal(ransac_estimator.predict(X), np.zeros(100))
def test_ransac_no_valid_model():
def is_model_valid(estimator, X, y):
return False
base_estimator = LinearRegression()
ransac_estimator = RANSACRegressor(base_estimator,
is_model_valid=is_model_valid,
max_trials=5)
msg = ("RANSAC could not find a valid consensus set")
assert_raises_regexp(ValueError, msg, ransac_estimator.fit, X, y)
assert ransac_estimator.n_skips_no_inliers_ == 0
assert ransac_estimator.n_skips_invalid_data_ == 0
assert ransac_estimator.n_skips_invalid_model_ == 5
def test_ransac_is_model_valid():
def is_model_valid(estimator, X, y):
assert X.shape[0] == 2
assert y.shape[0] == 2
return False
base_estimator = LinearRegression()
ransac_estimator = RANSACRegressor(base_estimator,
min_samples=2,
residual_threshold=5,
is_model_valid=is_model_valid,
random_state=0)
assert_raises(ValueError, ransac_estimator.fit, X, y)
def test_less_samples_than_features():
random_state = np.random.RandomState(0)
n_samples, n_features = 10, 20
X = random_state.normal(size=(n_samples, n_features))
y = random_state.normal(size=n_samples)
# Check that Theil-Sen falls back to Least Squares if fit_intercept=False
theil_sen = TheilSenRegressor(fit_intercept=False,
random_state=0).fit(X, y)
lstq = LinearRegression(fit_intercept=False).fit(X, y)
assert_array_almost_equal(theil_sen.coef_, lstq.coef_, 12)
# Check fit_intercept=True case. This will not be equal to the Least
# Squares solution since the intercept is calculated differently.
theil_sen = TheilSenRegressor(fit_intercept=True, random_state=0).fit(X, y)
y_pred = theil_sen.predict(X)
assert_array_almost_equal(y_pred, y, 12)
def test_ransac_default_residual_threshold():
base_estimator = LinearRegression()
ransac_estimator = RANSACRegressor(base_estimator,
min_samples=2,
random_state=0)
# Estimate parameters of corrupted data
ransac_estimator.fit(X, y)
# Ground truth / reference inlier mask
ref_inlier_mask = np.ones_like(ransac_estimator.inlier_mask_).astype(
np.bool_)
ref_inlier_mask[outliers] = False
assert_array_equal(ransac_estimator.inlier_mask_, ref_inlier_mask)
