def tangent_riemannian_submersion_test_data(self):
random_data = []
for n, n_samples in zip(self.n_list, self.n_samples_list):
bundle = CorrelationMatricesBundle(n)
mat = bundle.random_point()
point = bundle.riemannian_submersion(mat)
vec = bundle.random_point(n_samples)
random_data.append(dict(n=n, vec=vec, point=point))
return self.generate_tests([], random_data)
def vertical_projection_is_vertical_test_data(self):
random_data = []
for n, n_samples in zip(self.n_list, self.n_samples_list):
bundle = CorrelationMatricesBundle(n)
mat = bundle.random_point()
vec = bundle.random_point(n_samples)
tangent_vec = bundle.base.to_tangent(vec, mat)
random_data.append(dict(n=n, tangent_vec=tangent_vec, mat=mat))
return self.generate_tests([], random_data)
def horizontal_projection_test_data(self):
random_data = []
for n in self.n_list:
bundle = CorrelationMatricesBundle(n)
mat = bundle.random_point()
vec = bundle.random_point()
random_data.append(dict(n=n, vec=vec, mat=mat))
return self.generate_tests([], random_data)
def vertical_projection_tangent_submersion_test_data(self):
random_data = []
for n in self.n_list:
bundle = CorrelationMatricesBundle(n)
mat = bundle.random_point(2)
vec = SymmetricMatrices(n).random_point(2)
random_data.append(dict(n=n, vec=vec, mat=mat))
return self.generate_tests([], random_data)
def log_after_align_is_horizontal_test_data(self):
n_list = [2, 3]
random_data = []
for n in n_list:
bundle = CorrelationMatricesBundle(n)
point = bundle.random_point(2)
random_data.append(dict(n=n, point_a=point[0], point_b=point[1]))
return self.generate_tests([], random_data)
def exp_belongs_test_data(self):
bundle = CorrelationMatricesBundle(3)
base_point = bundle.base.random_point()
tangent_vec = bundle.base.to_tangent(bundle.random_point(), base_point)
smoke_data = [
dict(dim=3, tangent_vec=tangent_vec, base_point=base_point)
]
return self.generate_tests(smoke_data)
def exp_log_composition_test_data(self):
bundle = CorrelationMatricesBundle(3)
point = bundle.riemannian_submersion(bundle.random_point(2))
random_data = [dict(dim=3, point=point)]
return self.generate_tests([], random_data)