def test_invalid_sequence_ten_matrices_multiplication(self):
# generating random matrix pair
matrix_container = MatrixContainer()
matrix_container.matrix_factory.override(
providers.Factory(ZeroMatrixFactory))
matrix_container.matrix_sequence_factory.override(providers.Factory(
MatrixSequenceFactory, matrix_factory=matrix_container.matrix_factory))
matrix_container.shape_sequence_factory.override(
providers.Factory(InvalidShapeSequenceFactory, error_count_range=(1, 1)))
generate_matrix_sequence = matrix_container.generate_matrix_sequence()
matrices = generate_matrix_sequence(length=2)
# executing command under test
with self.assertRaises(ValueError):
with multiprocessing.Pool() as pool:
result_matrix = multiprocess_matrices_multiplication(
pool=pool, matrices=matrices)
def test_random_matrices_multiplication(self):
# generating random matrix pair
matrix_container = MatrixContainer()
matrix_container.matrix_factory.override(
providers.Factory(RandomMatrixFactory))
matrix_container.matrix_sequence_factory.override(providers.Factory(
MatrixSequenceFactory, matrix_factory=matrix_container.matrix_factory))
matrix_container.shape_sequence_factory.override(
providers.Factory(ValidShapeSequenceFactory))
generate_matrix_sequence = matrix_container.generate_matrix_sequence()
matrices = generate_matrix_sequence(length=2)
# calculating expected result
expected_result_matrix = reduce(lambda m1, m2: numpy.dot(
m1, to_ndarray(m2)), matrices[1:], to_ndarray(matrices[0]))
# executing command under test
with multiprocessing.Pool() as pool:
result_matrix = multiprocess_matrices_multiplication(
pool=pool, matrices=matrices)
numpy.testing.assert_array_almost_equal(
to_ndarray(matrix=result_matrix), expected_result_matrix)
def test_zero_matrices_multiplication(self):
# generating random matrix pair
matrix_container = MatrixContainer()
matrix_container.matrix_factory.override(
providers.Factory(ZeroMatrixFactory))
matrix_container.matrix_sequence_factory.override(providers.Factory(
MatrixSequenceFactory, matrix_factory=matrix_container.matrix_factory))
matrix_container.shape_sequence_factory.override(
providers.Factory(ValidShapeSequenceFactory))
generate_matrix_sequence = matrix_container.generate_matrix_sequence()
matrices = generate_matrix_sequence(length=2)
# calculating expected result
expected_result_matrix = numpy.zeros(
shape=(matrices[0].column_len(), matrices[-1].row_len()))
# executing command under test
with multiprocessing.Pool() as pool:
result_matrix = multiprocess_matrices_multiplication(
pool=pool, matrices=matrices)
self.assertTrue(numpy.array_equal(
to_ndarray(matrix=result_matrix), expected_result_matrix))