def test_hawkes_adm4_set_data(self):
"""...Test set_data method of Hawkes ADM4
"""
events = [[
np.array([1, 1.2, 3.4, 5.8, 10.3, 11, 13.4]),
np.array([2, 5, 8.3, 9.10, 15, 18, 20, 33])
],
[
np.array([2, 3.2, 11.4, 12.8, 45]),
np.array([2, 3, 8.8, 9, 15.3, 19])
]]
learner = HawkesADM4(1)
learner._set_data(events)
self.assertEqual(learner.n_nodes, 2)
events = [
np.array([1, 1.2, 3.4, 5.8, 10.3, 11, 13.4]),
np.array([2, 5, 8.3, 9.10, 15, 18, 20, 33])
]
learner = HawkesADM4(1)
learner._set_data(events)
self.assertEqual(learner.n_nodes, 2)
msg = "All realizations should have 2 nodes, but realization 1 has " \
"1 nodes"
with self.assertRaisesRegex(RuntimeError, msg):
events = [[
np.array([1, 1.2, 3.4, 5.8, 10.3, 11, 13.4]),
np.array([2, 5, 8.3, 9.10, 15, 18, 20, 33])
], [np.array([2, 3.2, 11.4, 12.8, 45])]]
learner._set_data(events)
def test_hawkes_adm4_parameters(self):
"""...Test that hawkes adm4 parameters are correctly linked
"""
learner = HawkesADM4(self.float_1)
self.assertEqual(learner.decay, self.float_1)
self.assertEqual(learner._learner.get_decay(), self.float_1)
learner.decay = self.float_2
self.assertEqual(learner.decay, self.float_2)
self.assertEqual(learner._learner.get_decay(), self.float_2)
learner = HawkesADM4(1, rho=self.float_1)
self.assertEqual(learner.rho, self.float_1)
self.assertEqual(learner._learner.get_rho(), self.float_1)
learner.rho = self.float_2
self.assertEqual(learner.rho, self.float_2)
self.assertEqual(learner._learner.get_rho(), self.float_2)
def learn_adm4(events,
end_time,
return_learner=False,
verbose=False,
**kwargs):
learner_mle = HawkesADM4(**kwargs, verbose=verbose)
learner_mle.fit(events, end_time)
if return_learner:
return learner_mle
return learner_mle.baseline, learner_mle.adjacency
def test_sparse(self):
"""...Test that original coeffs are correctly retrieved when some
realizations are empty
"""
baseline, adjacency, events = self.simulate_sparse_realization()
learner = HawkesADM4(self.decay, verbose=False)
learner.fit(events)
np.testing.assert_array_almost_equal(learner.baseline,
baseline,
decimal=1)
np.testing.assert_array_almost_equal(learner.adjacency,
adjacency,
decimal=1)
def test_hawkes_adm4_solution(self):
"""...Test solution obtained by HawkesADM4 on toy timestamps
"""
events = [[
np.array([1, 1.2, 3.4, 5.8, 10.3, 11, 13.4]),
np.array([2, 5, 8.3, 9.10, 15, 18, 20, 33])
],
[
np.array([2, 3.2, 11.4, 12.8, 45]),
np.array([2, 3, 8.8, 9, 15.3, 19])
]]
n_nodes = len(events[0])
rho = 0.5
C = 10
lasso_nuclear_ratio = 0.7
baseline_start = np.zeros(n_nodes) + .2
adjacency_start = np.zeros((n_nodes, n_nodes)) + .2
learner = HawkesADM4(self.decay,
rho=rho,
C=C,
lasso_nuclear_ratio=lasso_nuclear_ratio,
n_threads=3,
max_iter=11,
verbose=False,
em_max_iter=3,
record_every=1)
learner.fit(events[0],
baseline_start=baseline_start,
adjacency_start=adjacency_start)
baseline = np.array([0.14551, 0.239859])
adjacency = np.array([[2.275416e-01, 8.234672e-02],
[1.195861e-02, 4.070548e-10]])
np.testing.assert_array_almost_equal(learner.baseline,
baseline,
decimal=6)
np.testing.assert_array_almost_equal(learner.adjacency,
adjacency,
decimal=6)
def test_hawkes_adm4_score(self):
"""...Test HawkesADM4 score method
"""
n_nodes = 2
n_realizations = 3
train_events = [[
np.cumsum(np.random.rand(4 + i)) for i in range(n_nodes)
] for _ in range(n_realizations)]
test_events = [[
np.cumsum(np.random.rand(4 + i)) for i in range(n_nodes)
] for _ in range(n_realizations)]
learner = HawkesADM4(self.decay, record_every=1)
msg = '^You must either call `fit` before `score` or provide events$'
with self.assertRaisesRegex(ValueError, msg):
learner.score()
given_baseline = np.random.rand(n_nodes)
given_adjacency = np.random.rand(n_nodes, n_nodes)
learner.fit(train_events)
train_score_current_coeffs = learner.score()
self.assertAlmostEqual(train_score_current_coeffs, 0.12029826)
train_score_given_coeffs = learner.score(baseline=given_baseline,
adjacency=given_adjacency)
self.assertAlmostEqual(train_score_given_coeffs, -0.15247511)
test_score_current_coeffs = learner.score(test_events)
self.assertAlmostEqual(test_score_current_coeffs, 0.17640007)
test_score_given_coeffs = learner.score(test_events,
baseline=given_baseline,
adjacency=given_adjacency)
self.assertAlmostEqual(test_score_given_coeffs, -0.07973875)