def test_small_data_time_series(self):
"""
Tests that the time series method returns valid data.
Valid data is defined as not all values equal to 0.
"""
np.set_printoptions(precision=5)
hcbm_mat = generate_hcmb_mat(t_samples=1,
n_size=6,
rho_low=0.25,
rho_high=0.75,
blocks=2,
depth=2,
permute=False)
time_series = time_series_from_dist(corr=hcbm_mat[0],
t_samples=6,
dist="normal").round(5)
flat_series = [
item for sublist in time_series.values.tolist() for item in sublist
]
time_series_sample = [
[0.98732, -0.22608, 0.67257, 1.01905, 0.49815, 0.27573],
[-0.49786, -0.25026, 0.96379, 0.1083, 0.90386, 2.23842],
[-0.65337, -1.29134, -1.17062, 0.10184, 0.418, -0.07461],
[0.00385, -1.2113, 1.10762, -0.20205, 0.9475, -0.23154],
[0.34948, 1.7678, 0.66294, 0.44735, 1.02399, 0.68903],
[-0.70259, -0.05277, -1.17347, -1.01258, -0.37828, -0.6414]
]
flat_sample = [
item for sublist in time_series_sample for item in sublist
]
self.assertCountEqual(flat_series, flat_sample)
def test_matrix_is_symmetrical_hcbm(self):
"""
Tests that the data generated from the HCBM method are symmetrical.
"""
for i in range(self.n_generated):
hcbm_mat = generate_hcmb_mat(
t_samples=self.t_samples[i],
n_size=self.n_size[i],
rho_low=self.rho_lows[i],
rho_high=self.rho_highs[i],
blocks=self.blocks[i],
depth=self.depths[i],
permute=self.permute[i],
)
self.assertTrue(self._data_is_symmetric(hcbm_mat, self.n_size[i]))
def test_data_returned_hcbm(self):
"""
Tests that data generated from HCBM method exists for multiple dimensions and
parameters.
"""
for i in range(self.n_generated):
hcbm_mat = generate_hcmb_mat(
t_samples=self.t_samples[i],
n_size=self.n_size[i],
rho_low=self.rho_lows[i],
rho_high=self.rho_highs[i],
blocks=self.blocks[i],
depth=self.depths[i],
permute=self.permute[i],
)
self.assertTrue(self._data_exists(hcbm_mat))
def test_correct_shapes_hcbm(self):
"""
Tests that the data generated from the HCBM method exists for multiple dimensions and
parameters.
"""
for i in range(self.n_generated):
hcbm_mat = generate_hcmb_mat(
t_samples=self.t_samples[i],
n_size=self.n_size[i],
rho_low=self.rho_lows[i],
rho_high=self.rho_highs[i],
blocks=self.blocks[i],
depth=self.depths[i],
permute=self.permute[i],
)
self.assertTrue(hcbm_mat.shape == (self.t_samples[i],
self.n_size[i], self.n_size[i]))
def test_small_data_hcbm():
"""
Tests that the HCBM method returns valid data.
Valid data is defined as not all values equal to 0.
"""
hcbm_sample = [[[1, 0.51696, 0.25, 0.25, 0.25, 0.25],
[0.51696, 1, 0.25, 0.25, 0.25, 0.25],
[0.25, 0.25, 1, 0.55194, 0.31992, 0.31992],
[0.25, 0.25, 0.55194, 1, 0.31992, 0.31992],
[0.25, 0.25, 0.31992, 0.31992, 1, 0.43508],
[0.25, 0.25, 0.31992, 0.31992, 0.43508, 1]]]
hcbm_mat = generate_hcmb_mat(t_samples=1,
n_size=6,
rho_low=0.25,
rho_high=0.75,
blocks=2,
depth=2,
permute=False)
np.testing.assert_array_almost_equal(hcbm_mat, hcbm_sample, 5)
def test_data_is_correctly_bounded_hcbm(self):
"""
Tests that the data generated from the HCBM method is bounded to the range
rho_low <= data <= rho_high and that its diagonal is always equal to 1s.
"""
for i in range(self.n_generated):
hcbm_mat = generate_hcmb_mat(
t_samples=self.t_samples[i],
n_size=self.n_size[i],
rho_low=self.rho_lows[i],
rho_high=self.rho_highs[i],
blocks=self.blocks[i],
depth=self.depths[i],
permute=self.permute[i],
)
self.assertTrue(
self._data_correctly_bounded(hcbm_mat, self.n_size[i],
self.rho_lows[i],
self.rho_highs[i]))
def test_data_returned_time_series(self):
"""
Tests that data generated from HCBM method exists for multiple dimensions and
parameters.
"""
for i in range(self.n_generated):
hcbm_mat = generate_hcmb_mat(
t_samples=1,
n_size=self.n_size[i],
rho_low=self.rho_lows[i],
rho_high=self.rho_highs[i],
blocks=self.blocks[i],
depth=self.depths[i],
permute=self.permute[i],
)
time_series = time_series_from_dist(corr=hcbm_mat[0],
t_samples=self.ts_samples[i],
dist=self.dists[i])
self.assertTrue(self._data_exists(time_series))