def load_test_data(self):
script_dir = os.path.dirname(__file__)
absolute_path = os.path.join(script_dir, '../matfiles/')
self.sample_data = GraphDataSet.read(absolute_path, 'example.mat')
self.sample_data_memsave = GraphDataSet.read(absolute_path, 'example.mat', memsave=True)
self.prophnet = ProphNet(self.sample_data)
self.prophnet_memsave = ProphNet(self.sample_data_memsave)
def load_test_data(self):
script_dir = os.path.dirname(__file__)
absolute_path = os.path.join(script_dir, '../matfiles/')
self.sample_data = GraphDataSet.read(absolute_path, 'example.mat')
self.prioritizer = ProphNet(self.sample_data)
self.prio_test = PrioritizationTest(self.prioritizer)
def test_unknown_method_raises_exception(self):
with self.assertRaises(ValueError):
ProphNet(self.sample_data, method="not_valid")
def test_match_matrix_dimensions_matchable(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 1; 0 1; 1 0')
self.assertEquals(mat_b.shape[0], 3)
result = ProphNet.match_matrix_dimensions(mat_b, mat_a)
self.assertEquals(result.shape[0], 2)
def test_match_matrix_dimensions_unmatchable(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 1; 1 0')
with self.assertRaises(ValueError):
ProphNet.match_matrix_dimensions(mat_a, mat_b)
def test_match_matrix_dimensions_matched(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 0 1; 1 0 1; 1 1 0')
result = ProphNet.check_matrix_dimensions(mat_a, mat_b)
self.assertEquals(result, 0)
def test_find_all_paths_single_connection(self):
network = nx.from_numpy_matrix(np.matrix('0 0 1; 0 0 0; 1 0 0'))
paths = ProphNet.find_all_paths(network, 0, 2)
self.assertTrue([0, 2] in paths)
def test_find_all_paths_no_connection(self):
network = nx.from_numpy_matrix(np.matrix('0 0 1; 0 0 0; 1 0 0'))
paths = ProphNet.find_all_paths(network, 1, 2)
self.assertEquals([], paths)
def test_find_all_paths_normal_case(self):
super_adjacency = np.matrix('0 1 1; 1 0 1; 1 1 0')
paths = ProphNet.find_all_paths(nx.from_numpy_matrix(super_adjacency), 0, 2)
self.assertTrue([0, 2] in paths)
self.assertTrue([0, 1, 2] in paths)
class TestProphNetFunctions(unittest.TestCase):
"""
Test for ProphNet class
"""
def load_test_data(self):
script_dir = os.path.dirname(__file__)
absolute_path = os.path.join(script_dir, '../matfiles/')
self.sample_data = GraphDataSet.read(absolute_path, 'example.mat')
self.sample_data_memsave = GraphDataSet.read(absolute_path, 'example.mat', memsave=True)
self.prophnet = ProphNet(self.sample_data)
self.prophnet_memsave = ProphNet(self.sample_data_memsave)
def setUp(self):
self.load_test_data()
def tearDown(self):
"""Function to do cleaning up after the test."""
self.prophnet_memsave.graphdata.cleanup_resources()
def test_find_all_paths_normal_case(self):
super_adjacency = np.matrix('0 1 1; 1 0 1; 1 1 0')
paths = ProphNet.find_all_paths(nx.from_numpy_matrix(super_adjacency), 0, 2)
self.assertTrue([0, 2] in paths)
self.assertTrue([0, 1, 2] in paths)
def test_find_all_paths_no_connection(self):
network = nx.from_numpy_matrix(np.matrix('0 0 1; 0 0 0; 1 0 0'))
paths = ProphNet.find_all_paths(network, 1, 2)
self.assertEquals([], paths)
def test_find_all_paths_single_connection(self):
network = nx.from_numpy_matrix(np.matrix('0 0 1; 0 0 0; 1 0 0'))
paths = ProphNet.find_all_paths(network, 0, 2)
self.assertTrue([0, 2] in paths)
def test_compute_super_adjacency(self):
adj = self.prophnet.graphdata.super_adjacency
self.assertEquals(adj[0, 1], 1)
self.assertEquals(adj[1, 0], 1)
self.assertEquals(adj[0, 0], 0)
self.assertEquals(adj[1, 1], 0)
def test_match_matrix_dimensions_matched(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 0 1; 1 0 1; 1 1 0')
result = ProphNet.check_matrix_dimensions(mat_a, mat_b)
self.assertEquals(result, 0)
def test_match_matrix_dimensions_unmatchable(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 1; 1 0')
with self.assertRaises(ValueError):
ProphNet.match_matrix_dimensions(mat_a, mat_b)
def test_match_matrix_dimensions_matched_non_squared(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 1 0; 0 1 0')
self.assertEquals(mat_b.shape[0], 2)
result = ProphNet.match_matrix_dimensions(mat_b, mat_a)
self.assertEquals(result.shape[0], 2)
def test_match_matrix_dimensions_matchable(self):
mat_a = np.matrix('0 1 1; 1 0 1; 1 1 0')
mat_b = np.matrix('0 1; 0 1; 1 0')
self.assertEquals(mat_b.shape[0], 3)
result = ProphNet.match_matrix_dimensions(mat_b, mat_a)
self.assertEquals(result.shape[0], 2)
def test_unknown_method_raises_exception(self):
with self.assertRaises(ValueError):
ProphNet(self.sample_data, method="not_valid")
def test_network_out_of_bounds_raises_exception(self):
with self.assertRaises(ValueError):
self.prophnet.propagate([3], 0, 5)
def test_unknown_correlation_function_raises_exception(self):
with self.assertRaises(ValueError):
self.prophnet.propagate([3], 0, 1, corr_function="not_valid")
def test_generate_query_vector_matches_network_length(self):
query = [3]
dst = 2
network_length = self.prophnet.graphdata.networks[dst].matrix.shape[0]
query_vector = self.prophnet.generate_query_vector(query, dst)
self.assertEquals(network_length, len(query_vector))
def _within_network_propagation_best_results_matches_query(self, method, query):
names = method.graphdata.networks[0].node_names
scores = method.propagate(query, 0, 0)
sorted_scores = sorted(scores, key=lambda x: x[0], reverse=True)
result_tag = sorted_scores[0][1]
result_index = np.where(names == result_tag)
return (query[0] == result_index[0])
def test_within_network_propagation_memsavemode_same_results(self):
queries = [1, 8, 0]
for q in queries:
result_save = self._within_network_propagation_best_results_matches_query(self.prophnet_memsave, [q])
self.assertTrue(result_save)
def test_within_network_propagation_best_result_is_query_a(self):
result = self._within_network_propagation_best_results_matches_query(self.prophnet, [1])
self.assertTrue(result)
def test_within_network_propagation_best_result_is_query_b(self):
result = self._within_network_propagation_best_results_matches_query(self.prophnet, [8])
self.assertTrue(result)
def test_within_network_propagation_best_result_is_query_c(self):
result = self._within_network_propagation_best_results_matches_query(self.prophnet, [0])
self.assertTrue(result)
def test_query_out_of_bounds_raises_exception(self):
names = self.prophnet.graphdata.networks[0].node_names
query = [len(names)]
with self.assertRaises(ValueError):
scores = self.prophnet.propagate(query, 0, 0)
def _across_network_propagation_dst_names_test(self, method, query, src, dst):
scores = method.propagate(query, src, dst)
names_dst = method.graphdata.networks[dst].node_names
return (scores[0][1] in names_dst)
def test_across_network_propagation_returns_dst_net_tags_forward(self):
result = self._across_network_propagation_dst_names_test(self.prophnet, [1], 0, 1)
self.assertTrue(result)
def test_across_network_propagation_returns_dst_net_tags_reverse(self):
result = self._across_network_propagation_dst_names_test(self.prophnet, [1], 1, 0)
self.assertTrue(result)
def test_across_network_propagation_returns_dst_net_tags_forward_2(self):
result = self._across_network_propagation_dst_names_test(self.prophnet, [1], 0, 2)
self.assertTrue(result)
def test_across_network_propagation_returns_dst_net_tags_forward_memsave(self):
result = self._across_network_propagation_dst_names_test(self.prophnet_memsave, [1], 0, 1)
self.assertTrue(result)
def test_across_network_propagation_returns_dst_net_tags_reverse_memsave(self):
result = self._across_network_propagation_dst_names_test(self.prophnet_memsave, [1], 1, 0)
self.assertTrue(result)
def test_across_network_propagation_returns_dst_net_tags_forward_2_memsave(self):
result = self._across_network_propagation_dst_names_test(self.prophnet_memsave, [1], 0, 2)
self.assertTrue(result)