def test_PoissonRegression_settings(self):
"""...Test PoissonRegression basic settings
"""
# solver
solver_class_map = PoissonRegression._solvers
for solver in PoissonRegression._solvers.keys():
learner = PoissonRegression(solver=solver)
solver_class = solver_class_map[solver]
self.assertTrue(isinstance(learner._solver_obj, solver_class))
msg = '^``solver`` must be one of agd, bfgs, gd, sgd, svrg, ' \
'got wrong_name$'
with self.assertRaisesRegex(ValueError, msg):
PoissonRegression(solver='wrong_name')
prox_class_map = PoissonRegression._penalties
for penalty in PoissonRegression._penalties.keys():
if penalty == 'binarsity':
learner = PoissonRegression(penalty=penalty, blocks_start=[0],
blocks_length=[1])
else:
learner = PoissonRegression(penalty=penalty)
prox_class = prox_class_map[penalty]
self.assertTrue(isinstance(learner._prox_obj, prox_class))
msg = '^``penalty`` must be one of binarsity, elasticnet, l1, l2, ' \
'none, tv, got wrong_name$'
with self.assertRaisesRegex(ValueError, msg):
PoissonRegression(penalty='wrong_name')
def test_loglik(self):
"""...Test PoissonRegression loglik function
"""
X_train, y_train, _, _ = self.get_train_data(n_samples=200,
n_features=12)
learner = PoissonRegression(random_state=32789, tol=1e-9)
learner.fit(X_train, y_train)
X_test, y_test, _, _ = self.get_train_data(n_samples=5, n_features=12)
np.testing.assert_array_almost_equal(
learner.loglik(X_test, y_test), 1.8254, decimal=4)
def test_predict(self):
"""...Test PoissonRegression predict
"""
X_train, y_train, _, _ = self.get_train_data(n_samples=200,
n_features=12)
learner = PoissonRegression(random_state=32789, tol=1e-9)
learner.fit(X_train, y_train)
X_test, y_test, _, _ = self.get_train_data(n_samples=5, n_features=12)
y_pred = np.array([1., 5., 0., 5., 6.])
np.testing.assert_array_almost_equal(learner.predict(X_test), y_pred)
def test_decision_function(self):
"""...Test PoissonRegression decision function
"""
X_train, y_train, _, _ = self.get_train_data(n_samples=200,
n_features=12)
learner = PoissonRegression(random_state=32789, tol=1e-9)
learner.fit(X_train, y_train)
X_test, y_test, _, _ = self.get_train_data(n_samples=5, n_features=12)
y_pred = np.array([1.1448, 5.2194, 0.2624, 4.5525, 6.4168])
np.testing.assert_array_almost_equal(
learner.decision_function(X_test), y_pred, decimal=4)
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):
PoissonRegression._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):
PoissonRegression._safe_array(self.X[::2])
np.testing.assert_array_equal(self.X,
PoissonRegression._safe_array(self.X))
def test_PoissonRegression_fit(self):
"""...Test PoissonRegression fit with default parameters
"""
n_samples = 2000
n_features = 20
for fit_intercept in [False, True]:
X, y, weights0, intercept0 = self.get_train_data(
n_samples=n_samples, n_features=n_features,
fit_intercept=fit_intercept)
learner = PoissonRegression(C=1e3, verbose=False,
fit_intercept=fit_intercept,
solver='bfgs')
learner.fit(X, y)
err = np.linalg.norm(learner.weights - weights0) / n_features
self.assertLess(err, 1e-2)
if fit_intercept:
self.assertLess(np.abs(learner.intercept - intercept0), 1e-1)
def test_PoissonRegression_penalty_elastic_net_ratio(self):
"""...Test PoissonRegression setting of parameter of elastic_net_ratio
"""
ratio_1 = 0.6
ratio_2 = 0.3
for penalty in PoissonRegression._penalties.keys():
if penalty == 'elasticnet':
learner = PoissonRegression(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):
if penalty == 'binarsity':
PoissonRegression(penalty=penalty,
elastic_net_ratio=0.8,
blocks_start=[0],
blocks_length=[1])
else:
PoissonRegression(penalty=penalty,
elastic_net_ratio=0.8)
if penalty == 'binarsity':
learner = PoissonRegression(penalty=penalty,
blocks_start=[0],
blocks_length=[1])
else:
learner = PoissonRegression(penalty=penalty)
with self.assertWarnsRegex(RuntimeWarning, msg):
learner.elastic_net_ratio = ratio_1
def test_PoissonRegression_run(self):
"""...Test PoissonRegression runs with different solvers and penalties
"""
n_samples = 200
n_features = 10
for fit_intercept in [False, True]:
X, y, weights0, intercept0 = self.get_train_data(
n_samples=n_samples, n_features=n_features,
fit_intercept=fit_intercept)
for solver, penalty in product(PoissonRegression._solvers,
PoissonRegression._penalties):
if solver == 'bfgs' and (penalty not in ['zero', 'l2']):
continue
if penalty == 'binarsity':
learner = PoissonRegression(
verbose=False, fit_intercept=fit_intercept,
solver=solver, penalty=penalty, max_iter=1, step=1e-5,
blocks_start=[0], blocks_length=[1])
else:
learner = PoissonRegression(
verbose=False, fit_intercept=fit_intercept,
solver=solver, penalty=penalty, max_iter=1, step=1e-5)
learner.fit(X, y)
self.assertTrue(np.isfinite(learner.weights).all())
if fit_intercept:
self.assertTrue(np.isfinite(learner.intercept))
def test_PoissonRegression_solver_step(self):
"""...Test LogisticRegression setting of step parameter of solver
"""
for solver in PoissonRegression._solvers.keys():
if solver == 'bfgs':
learner = PoissonRegression(solver=solver)
self.assertIsNone(learner.step)
learner = PoissonRegression(solver=solver, step=self.float_1)
self.assertIsNone(learner.step)
msg = '^Solver "bfgs" has no settable step$'
with self.assertWarnsRegex(RuntimeWarning, msg):
learner.step = self.float_2
self.assertIsNone(learner.step)
else:
learner = PoissonRegression(solver=solver, step=self.float_1)
self.assertEqual(learner.step, self.float_1)
self.assertEqual(learner._solver_obj.step, self.float_1)
learner.step = self.float_2
self.assertEqual(learner.step, self.float_2)
self.assertEqual(learner._solver_obj.step, self.float_2)
def test_PoissonRegression_solver_random_state(self):
"""...Test PoissonRegression setting of random_state parameter of solver
"""
for solver in PoissonRegression._solvers.keys():
if solver in ['agd', 'gd', 'bfgs']:
msg = '^Solver "%s" has no settable random_state$' % solver
with self.assertWarnsRegex(RuntimeWarning, msg):
learner = PoissonRegression(solver=solver, random_state=1)
self.assertIsNone(learner.random_state)
else:
learner = PoissonRegression(solver=solver,
random_state=self.int_1)
self.assertEqual(learner.random_state, self.int_1)
self.assertEqual(learner._solver_obj.seed, self.int_1)
msg = '^random_state must be positive, got -1$'
with self.assertRaisesRegex(ValueError, msg):
PoissonRegression(solver=solver, random_state=-1)
msg = '^random_state is readonly in PoissonRegression$'
with self.assertRaisesRegex(AttributeError, msg):
learner = PoissonRegression(solver=solver)
learner.random_state = self.int_2
def test_PoissonRegression_model_settings(self):
"""...Test LogisticRegression setting of parameters of model
"""
for solver in PoissonRegression._solvers.keys():
learner = PoissonRegression(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 = PoissonRegression(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_PoissonRegression_solver_basic_settings(self):
"""...Test LogisticRegression setting of basic parameters of solver
"""
for solver in PoissonRegression._solvers.keys():
# tol
learner = PoissonRegression(solver=solver, tol=self.float_1)
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 = PoissonRegression(solver=solver, max_iter=self.int_1)
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 = PoissonRegression(solver=solver, verbose=True)
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 = PoissonRegression(solver=solver, verbose=False)
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 = PoissonRegression(solver=solver, print_every=self.int_1)
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 = PoissonRegression(solver=solver, record_every=self.int_1)
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_PoissonRegression_penalty_C(self):
"""...Test PoissonRegression setting of parameter of C
"""
for penalty in PoissonRegression._penalties.keys():
if penalty != 'none':
if penalty == 'binarsity':
learner = PoissonRegression(
penalty=penalty, C=self.float_1, blocks_start=[0],
blocks_length=[1])
else:
learner = PoissonRegression(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$'
with self.assertRaisesRegex(ValueError, msg):
if penalty == 'binarsity':
PoissonRegression(penalty=penalty, C=-1,
blocks_start=[0], blocks_length=[1])
else:
PoissonRegression(penalty=penalty, C=-1)
else:
msg = '^You cannot set C for penalty "%s"$' % penalty
with self.assertWarnsRegex(RuntimeWarning, msg):
PoissonRegression(penalty=penalty, C=self.float_1)
learner = PoissonRegression(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
import matplotlib.pyplot as plt
from tick.simulation import weights_sparse_gauss
from tick.linear_model import SimuPoisReg, PoissonRegression
from tick.plot import plot_history
n_samples = 50000
n_features = 100
np.random.seed(123)
weight0 = weights_sparse_gauss(n_features, nnz=int(n_features-1)) / 20.
intercept0 = -0.1
X, y = SimuPoisReg(weight0, intercept0, n_samples=n_samples,
verbose=False, seed=123).simulate()
opts = {'verbose': False, 'record_every': 1, 'tol': 1e-8, 'max_iter': 40}
poisson_regressions = [
PoissonRegression(solver='gd', **opts),
PoissonRegression(solver='agd', **opts),
PoissonRegression(solver='svrg', random_state=1234, **opts),
PoissonRegression(solver='bfgs', **opts)
]
for poisson_regression in poisson_regressions:
poisson_regression.fit(X, y)
plot_history(poisson_regressions, log_scale=True, dist_min=True)
plt.title('Solvers comparison for Poisson regression', fontsize=16)
plt.tight_layout()