def __save_and_load_test (self, name, A ):
save_scipy_matrix_as_mat ( data_path, name, A )
A_load = load_scipy_matrix_from_mat ( data_path, name )
np.testing.assert_equal ( A_load.shape, A.shape )
if type(A) != np.ndarray:
np.testing.assert_equal( A_load.data, A.data )
np.testing.assert_equal( A_load.indices, A.indices )
np.testing.assert_equal( A_load.indptr, A.indptr )
else:
np.testing.assert_equal( A_load, A )
def __save_and_load_test(self, name, A):
save_scipy_matrix_as_mat(data_path, name, A)
A_load = load_scipy_matrix_from_mat(data_path, name)
np.testing.assert_equal(A_load.shape, A.shape)
if type(A) != np.ndarray:
np.testing.assert_equal(A_load.data, A.data)
np.testing.assert_equal(A_load.indices, A.indices)
np.testing.assert_equal(A_load.indptr, A.indptr)
else:
np.testing.assert_equal(A_load, A)
def save_matrices(path, id, A, b):
"""
Save the system matrix and the excitation vector.
The matrices will be saved in <path>/<id>/A.mat and <path>/<id>/b.mat
@param path: the folder where the matrices are to be saved
@param id: a unique problem identifier
@param A: the system matrix
@param b: the excitation vector
"""
full_path = os.path.join(path, id)
save_scipy_matrix_as_mat(full_path, 'A', A)
save_scipy_matrix_as_mat(full_path, 'b', b)
dofs = A.shape[0]
f = file(os.path.join(full_path, '%dDOFs.txt' % dofs), 'w')
f.write('DOFs = %d\n' % dofs)
f.close()
def test_save_scipy_matrix ( self ):
import scipy.sparse
N = 1000;
A = scipy.sparse.rand ( N, N, format='csr' )
save_scipy_matrix_as_mat ( data_path, 'A', A )
def test_save_scipy_matrix(self):
import scipy.sparse
N = 1000
A = scipy.sparse.rand(N, N, format='csr')
save_scipy_matrix_as_mat(data_path, 'A', A)