def test_RobustLinearRegression_settings(self):
"""...Test RobustLinearRegression basic settings
"""
# solver
solver_class_map = RobustLinearRegression._solvers
for solver in RobustLinearRegression._solvers.keys():
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
**Test.specific_solver_kwargs(solver))
solver_class = solver_class_map[solver]
self.assertTrue(isinstance(learner._solver_obj, solver_class))
msg = '^``solver`` must be one of agd, gd, got wrong_name$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1., solver='wrong_name')
# prox
prox_class_map = RobustLinearRegression._penalties
for penalty in RobustLinearRegression._penalties.keys():
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty)
prox_class = prox_class_map[penalty]
self.assertTrue(isinstance(learner._prox_obj, prox_class))
msg = '^``penalty`` must be one of elasticnet, l1, l2, ' \
'none, slope, got wrong_name$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty='wrong_name')
def test_RobustLinearRegression_penalty_C(self):
"""...Test RobustLinearRegression setting of parameter of C
"""
for penalty in RobustLinearRegression._penalties.keys():
if penalty != 'none':
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
C=self.float_1)
self.assertEqual(learner.C, self.float_1)
self.assertEqual(learner._prox_obj.strength, 1. / self.float_1)
learner.C = self.float_2
self.assertEqual(learner.C, self.float_2)
self.assertEqual(learner._prox_obj.strength, 1. / self.float_2)
msg = '^``C`` must be positive, got -1$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
C=-1)
else:
pass
msg = '^You cannot set C for penalty "%s"$' % penalty
with self.assertWarnsRegex(RuntimeWarning, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
C=self.float_1)
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty)
with self.assertWarnsRegex(RuntimeWarning, msg):
learner.C = self.float_1
msg = '^``C`` must be positive, got -2$'
with self.assertRaisesRegex(ValueError, msg):
learner.C = -2
def test_RobustLinearRegression_solver_step(self):
"""...Test RobustLinearRegression setting of step parameter of solver
"""
for solver in RobustLinearRegression._solvers.keys():
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
step=self.float_1,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.step, self.float_1)
self.assertEqual(learner._solver_obj.step, self.float_1)
learner.step = self.float_2
self.assertEqual(learner.step, self.float_2)
self.assertEqual(learner._solver_obj.step, self.float_2)
def test_safe_array_cast(self):
"""...Test error and warnings raised by LogLearner constructor
"""
msg = '^Copying array of size \(5, 5\) to convert it in the ' \
'right format$'
with self.assertWarnsRegex(RuntimeWarning, msg):
RobustLinearRegression._safe_array(self.X.astype(int))
msg = '^Copying array of size \(3, 5\) to create a ' \
'C-contiguous version of it$'
with self.assertWarnsRegex(RuntimeWarning, msg):
RobustLinearRegression._safe_array(self.X[::2])
np.testing.assert_array_equal(
self.X, RobustLinearRegression._safe_array(self.X))
def test_RobustLinearRegression_penalty_elastic_net_ratio(self):
"""...Test RobustLinearRegression setting of parameter of elastic_net_ratio
"""
ratio_1 = 0.6
ratio_2 = 0.3
for penalty in RobustLinearRegression._penalties.keys():
if penalty == 'elasticnet':
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
C=self.float_1,
elastic_net_ratio=ratio_1)
self.assertEqual(learner.C, self.float_1)
self.assertEqual(learner.elastic_net_ratio, ratio_1)
self.assertEqual(learner._prox_obj.strength, 1. / self.float_1)
self.assertEqual(learner._prox_obj.ratio, ratio_1)
learner.elastic_net_ratio = ratio_2
self.assertEqual(learner.C, self.float_1)
self.assertEqual(learner.elastic_net_ratio, ratio_2)
self.assertEqual(learner._prox_obj.ratio, ratio_2)
else:
msg = '^Penalty "%s" has no elastic_net_ratio attribute$$' % \
penalty
with self.assertWarnsRegex(RuntimeWarning, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
elastic_net_ratio=0.8)
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty)
with self.assertWarnsRegex(RuntimeWarning, msg):
learner.elastic_net_ratio = ratio_1
def test_RobustLinearRegression_fdr(self):
"""...Test RobustLinearRegression setting of parameter of C
"""
fdr1 = 0.33
fdr2 = 0.78
learner = RobustLinearRegression(C_sample_intercepts=self.float_1,
fdr=fdr1)
self.assertEqual(learner.fdr, fdr1)
self.assertEqual(learner._prox_intercepts_obj.fdr, fdr1)
learner.fdr = fdr2
self.assertEqual(learner.fdr, fdr2)
self.assertEqual(learner._prox_intercepts_obj.fdr, fdr2)
msg = '^``fdr`` must be in \(0, 1\), got -1.0'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1., fdr=-1.0)
msg = '^``fdr`` must be in \(0, 1\), got 1.5'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1., fdr=1.5)
msg = '^``fdr`` cannot be `None`$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1., fdr=None)
msg = '^``fdr`` must be a finite number, got inf$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1., fdr=np.inf)
def test_RobustLinearRegression_penalty_C_sample_intercepts(self):
"""...Test RobustLinearRegression setting of parameter of C
"""
learner = RobustLinearRegression(C_sample_intercepts=self.float_1)
self.assertEqual(learner.C_sample_intercepts, self.float_1)
self.assertEqual(learner._prox_intercepts_obj.strength,
1. / self.float_1)
learner.C_sample_intercepts = self.float_2
self.assertEqual(learner.C_sample_intercepts, self.float_2)
self.assertEqual(learner._prox_intercepts_obj.strength,
1. / self.float_2)
msg = '^``C_sample_intercepts`` must be positive, got -1.0$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=-1.)
msg = '^``C_sample_intercepts`` cannot be `None`$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=None)
msg = '^``C_sample_intercepts`` cannot be 0.$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=0.)
msg = '^``C_sample_intercepts`` must be a finite number, got inf$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=np.inf)
def test_RobustLinearRegression_warm_start(self):
"""...Test RobustLinearRegression warm start
"""
X, y, weights0, interc0, sample_intercepts0 = self.get_train_data(True)
fit_intercepts = [True, False]
cases = itertools.product(RobustLinearRegression._solvers.keys(),
fit_intercepts)
for solver, fit_intercept in cases:
solver_kwargs = {
'C_sample_intercepts': Test.n_samples / Test.noise_level,
'solver': solver,
'max_iter': 2,
'verbose': False,
'fit_intercept': fit_intercept,
'warm_start': True,
'tol': 0
}
learner = RobustLinearRegression(**solver_kwargs)
learner.fit(X, y)
coeffs_1 = learner.coeffs
learner.fit(X, y)
coeffs_2 = learner.coeffs
# Thanks to warm start objective should have decreased
self.assertLess(learner._solver_obj.objective(coeffs_2),
learner._solver_obj.objective(coeffs_1))
def test_RobustLinearRegression_model_settings(self):
"""...Test RobustLinearRegression setting of parameters of model
"""
for solver in RobustLinearRegression._solvers.keys():
learner = RobustLinearRegression(C_sample_intercepts=1.,
fit_intercept=True,
solver=solver)
self.assertEqual(learner.fit_intercept, True)
self.assertEqual(learner._model_obj.fit_intercept, True)
learner.fit_intercept = False
self.assertEqual(learner.fit_intercept, False)
self.assertEqual(learner._model_obj.fit_intercept, False)
learner = RobustLinearRegression(C_sample_intercepts=1.,
fit_intercept=False,
solver=solver)
self.assertEqual(learner.fit_intercept, False)
self.assertEqual(learner._model_obj.fit_intercept, False)
learner.fit_intercept = True
self.assertEqual(learner.fit_intercept, True)
self.assertEqual(learner._model_obj.fit_intercept, True)
def test_RobustLinearRegression_fit(self):
"""...Test RobustLinearRegression fit with different solvers and penalties
"""
X, y, weights0, interc0, sample_intercepts0 = self.get_train_data(True)
solvers = RobustLinearRegression._solvers.keys()
for i, solver in enumerate(solvers):
learner_keywords = {
'C_sample_intercepts': Test.n_samples / Test.noise_level,
'fit_intercept': True,
'fdr': Test.target_fdr,
'max_iter': 3000,
'tol': 1e-7,
'solver': solver,
'penalty': 'none',
'verbose': False
}
learner = RobustLinearRegression(**learner_keywords)
learner.fit(X, y)
weights = [
1.82145051, 2.32011366, 2.6886905, 2.53289584, 2.86991904
]
interc = -2.9877245464563931
fdp_ = 0.23076923076923078
power = 1.0
np.testing.assert_array_almost_equal(weights, learner.weights, 2)
self.assertAlmostEqual(interc, learner.intercept, 2)
self.assertAlmostEqual(
fdp_, support_fdp(sample_intercepts0,
learner.sample_intercepts), 4)
self.assertAlmostEqual(
power,
support_recall(sample_intercepts0, learner.sample_intercepts),
4)
X, y, weights0, interc0, sample_intercepts0 = self.get_train_data(
False)
for i, solver in enumerate(solvers):
learner_keywords = {
'C_sample_intercepts': Test.n_samples / Test.noise_level,
'fit_intercept': False,
'fdr': Test.target_fdr,
'max_iter': 3000,
'tol': 1e-7,
'solver': solver,
'verbose': False
}
learner = RobustLinearRegression(**learner_keywords)
learner.fit(X, y)
weights = [
1.82341444, 2.3226882, 2.68081823, 2.53942366, 2.86439685
]
interc = None
fdp_ = 0.23076923076923078
power = 1.0
np.testing.assert_array_almost_equal(weights, learner.weights, 2)
self.assertEqual(interc, learner.intercept)
self.assertAlmostEqual(
fdp_, support_fdp(sample_intercepts0,
learner.sample_intercepts), 4)
self.assertAlmostEqual(
power,
support_recall(sample_intercepts0, learner.sample_intercepts),
4)
def test_RobustLinearRegression_solver_basic_settings(self):
"""...Test RobustLinearRegression setting of basic parameters of solver
"""
for solver in RobustLinearRegression._solvers.keys():
# tol
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
tol=self.float_1,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.tol, self.float_1)
self.assertEqual(learner._solver_obj.tol, self.float_1)
learner.tol = self.float_2
self.assertEqual(learner.tol, self.float_2)
self.assertEqual(learner._solver_obj.tol, self.float_2)
# max_iter
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
max_iter=self.int_1,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.max_iter, self.int_1)
self.assertEqual(learner._solver_obj.max_iter, self.int_1)
learner.max_iter = self.int_2
self.assertEqual(learner.max_iter, self.int_2)
self.assertEqual(learner._solver_obj.max_iter, self.int_2)
# verbose
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
verbose=True,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.verbose, True)
self.assertEqual(learner._solver_obj.verbose, True)
learner.verbose = False
self.assertEqual(learner.verbose, False)
self.assertEqual(learner._solver_obj.verbose, False)
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
verbose=False,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.verbose, False)
self.assertEqual(learner._solver_obj.verbose, False)
learner.verbose = True
self.assertEqual(learner.verbose, True)
self.assertEqual(learner._solver_obj.verbose, True)
# print_every
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
print_every=self.int_1,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.print_every, self.int_1)
self.assertEqual(learner._solver_obj.print_every, self.int_1)
learner.print_every = self.int_2
self.assertEqual(learner.print_every, self.int_2)
self.assertEqual(learner._solver_obj.print_every, self.int_2)
# record_every
learner = RobustLinearRegression(
C_sample_intercepts=1.,
solver=solver,
record_every=self.int_1,
**Test.specific_solver_kwargs(solver))
self.assertEqual(learner.record_every, self.int_1)
self.assertEqual(learner._solver_obj.record_every, self.int_1)
learner.record_every = self.int_2
self.assertEqual(learner.record_every, self.int_2)
self.assertEqual(learner._solver_obj.record_every, self.int_2)
def test_RobustLinearRegression_penalty_slope_fdr(self):
"""...Test RobustLinearRegression setting of parameter of slope_fdr
"""
slope_fdr1 = 0.6
slope_fdr2 = 0.3
for penalty in RobustLinearRegression._penalties.keys():
if penalty == 'slope':
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
C=self.float_1,
slope_fdr=slope_fdr1)
self.assertEqual(learner.C, self.float_1)
self.assertEqual(learner.slope_fdr, slope_fdr1)
self.assertEqual(learner._prox_obj.strength, 1. / self.float_1)
self.assertEqual(learner._prox_obj.fdr, slope_fdr1)
learner.slope_fdr = slope_fdr2
self.assertEqual(learner.C, self.float_1)
self.assertEqual(learner.slope_fdr, slope_fdr2)
self.assertEqual(learner._prox_obj.fdr, slope_fdr2)
else:
msg = '^Penalty "%s" has no ``slope_fdr`` attribute$$' % \
penalty
with self.assertWarnsRegex(RuntimeWarning, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty,
slope_fdr=0.8)
learner = RobustLinearRegression(C_sample_intercepts=1.,
penalty=penalty)
msg = '^Penalty "%s" has no ``slope_fdr`` attribute$$' % \
penalty
with self.assertWarnsRegex(RuntimeWarning, msg):
learner.slope_fdr = slope_fdr1
slope_fdr1 = 0.33
slope_fdr2 = 0.78
learner = RobustLinearRegression(C_sample_intercepts=self.float_1,
slope_fdr=slope_fdr1,
penalty='slope')
self.assertEqual(learner.slope_fdr, slope_fdr1)
self.assertEqual(learner._prox_obj.fdr, slope_fdr1)
learner.slope_fdr = slope_fdr2
self.assertEqual(learner.slope_fdr, slope_fdr2)
self.assertEqual(learner._prox_obj.fdr, slope_fdr2)
msg = '^``slope_fdr`` must be in \(0, 1\), got -1.0'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty='slope',
slope_fdr=-1.0)
msg = '^``slope_fdr`` must be in \(0, 1\), got 1.5'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty='slope',
slope_fdr=1.5)
msg = '^``slope_fdr`` cannot be `None`$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty='slope',
slope_fdr=None)
msg = '^``slope_fdr`` must be a finite number, got inf$'
with self.assertRaisesRegex(ValueError, msg):
RobustLinearRegression(C_sample_intercepts=1.,
penalty='slope',
slope_fdr=np.inf)
sample_intercepts0 = weights_sparse_gauss(n_weights=n_samples, nnz=n_outliers)
idx_nnz = sample_intercepts0 != 0
log_linspace = np.log(n_samples * np.linspace(1, 10, n_outliers))
sample_intercepts0[idx_nnz] = outliers_intensity * np.sqrt(2 * log_linspace) \
* np.sign(sample_intercepts0[idx_nnz])
X = features_normal_cov_toeplitz(n_samples, n_features, 0.5)
y = X.dot(weights0) + noise_level * np.random.randn(n_samples) \
+ intercept0 + sample_intercepts0
target_fdr = 0.1
noise_level = std_iqr(y)
learner = RobustLinearRegression(C_sample_intercepts=2 * n_samples /
noise_level,
penalty='none',
fdr=target_fdr,
verbose=False)
learner.fit(X, y)
fdp_ = support_fdp(sample_intercepts0, learner.sample_intercepts)
power_ = support_recall(sample_intercepts0, learner.sample_intercepts)
fig = plt.figure(figsize=(7, 6))
titles = [
'Model weights', 'Learned weights', 'Sample intercepts',
'Learned intercepts'
]
vectors = [
weights0, learner.weights, sample_intercepts0, learner.sample_intercepts
]
