def setUp(self):
# read in sample data
current_directory = os.path.dirname(__file__)
self.data_dir = os.path.join(current_directory, 'data')
pos_train = os.path.abspath(self.data_dir +
'/ppismall/pos_train_edges')
# read data into graph
training_graph = CSFGraph(pos_train)
worddictionary = training_graph.get_node_to_index_map()
self.reverse_worddictionary = training_graph.get_index_to_node_map()
# generate random walks
n2v_graph = N2vGraph(training_graph, 1, 1)
walks = n2v_graph.simulate_walks(5, 10)
# learn embeddings
self.model = ContinuousBagOfWordsWord2Vec(
walks,
worddictionary=worddictionary,
reverse_worddictionary=self.reverse_worddictionary,
num_epochs=2)
# create temporary directory to write data to
self.temp_dir_loc = os.path.abspath(self.data_dir + '/temp')
os.mkdir(self.temp_dir_loc)
return None
def setUp(self):
curdir = os.path.dirname(__file__)
pos_train = os.path.join(curdir, 'data/ppismall/pos_train_edges')
pos_train = os.path.abspath(pos_train)
training_graph = CSFGraph(pos_train)
# obtain data needed to build model
worddictionary = training_graph.get_node_to_index_map()
reverse_worddictionary = training_graph.get_index_to_node_map()
# initialize n2v object
p, q = 1, 1
self.number_of_nodes_in_training = training_graph.node_count()
self.n2v_graph = N2vGraph(csf_graph=training_graph, p=p, q=q)
# generate random walks
self.walk_length = 10
self.num_walks = 5
self.walks = self.n2v_graph.simulate_walks(num_walks=self.num_walks, walk_length=self.walk_length)
# walks is now a list of lists of ints
# build cbow model
self.cbow = ContinuousBagOfWordsWord2Vec(self.walks,
worddictionary=worddictionary,
reverse_worddictionary=reverse_worddictionary,
num_epochs=2)
self.cbow.train()
def karate_test(pos_train_file, pos_valid_file, pos_test_file, neg_train_file,
neg_valid_file, neg_test_file, embed_graph, p, q, walk_length,
num_walks, num_epochs, classifier, edge_embed_method,
skipValidation, output):
pos_train_graph = CSFGraph(pos_train_file)
pos_valid_graph = CSFGraph(pos_valid_file)
pos_test_graph = CSFGraph(pos_test_file)
neg_train_graph = CSFGraph(neg_train_file)
neg_valid_graph = CSFGraph(neg_valid_file)
neg_test_graph = CSFGraph(neg_test_file)
# Graph (node) embeding using SkipGram as the word2vec model, with 2 epochs.
graph = embiggen.random_walk_generator.N2vGraph(pos_train_graph, p, q)
walks = graph.simulate_walks(num_walks, walk_length)
worddictionary = pos_train_graph.get_node_to_index_map()
reverse_worddictionary = pos_train_graph.get_index_to_node_map()
model = SkipGramWord2Vec(walks,
worddictionary=worddictionary,
reverse_worddictionary=reverse_worddictionary,
num_epochs=num_epochs)
model.train()
write_embeddings(embed_graph, model.embedding, reverse_worddictionary)
# Link prediction on the pos/neg train/valid/test sets using RF classifier
lp = LinkPrediction(pos_train_graph, pos_valid_graph, pos_test_graph,
neg_train_graph, neg_valid_graph, neg_test_graph,
embed_graph, edge_embed_method, classifier,
skipValidation, output)
lp.prepare_edge_and_node_labels()
lp.predict_links()
lp.output_classifier_results()
def setUp(self):
data_dir = os.path.join(os.path.dirname(__file__), 'data')
edge_file = os.path.join(data_dir, 'small_het_graph_edges.tsv')
node_file = os.path.join(data_dir, 'small_het_graph_nodes.tsv')
g = CSFGraph(edge_file=edge_file, node_file=node_file)
self.graph = g
self.nodes = g.nodes()
self.g1index = self.__get_index('g1')
self.d1index = self.__get_index('d1')
def setUp(self):
input_file = os.path.join(os.path.dirname(__file__), 'data',
'unweighted_small_graph.txt')
g = CSFGraph(input_file)
self.g = g
return str(g)
def setUp(self):
data_dir = os.path.join(os.path.dirname(__file__), 'data')
# these pass the tests okay
node_file = os.path.join(data_dir, 'small_graph_nodes.tsv')
edge_file = os.path.join(data_dir, 'small_graph_edges.tsv')
g = CSFGraph(edge_file=edge_file, node_file=node_file)
self.graph = g
def setUp(self):
data_dir = os.path.join(os.path.dirname(__file__), 'data')
# files for canonical test graph
self.edge_file = os.path.join(data_dir, 'small_graph_edges.tsv')
self.node_file = os.path.join(data_dir, 'small_graph_nodes.tsv')
# legacy and non-standard test files
self.legacy_edge_file = os.path.join(data_dir,
'small_graph_LEGACY.txt')
self.tsv_no_subject = os.path.join(data_dir,
'small_graph_edges_NO_SUBJECT.tsv')
self.tsv_no_object = os.path.join(data_dir,
'small_graph_edges_NO_OBJECT.tsv')
self.node_file_missing_nodes = os.path.join(
data_dir, 'small_graph_nodes_MISSING_NODES.tsv')
g = CSFGraph(edge_file=self.edge_file)
self.g = g
str(g)
def setUp(self) -> None:
self.file_dir = 'tests/data/ppismall_with_validation/'
self.pos_train_graph = CSFGraph(
os.path.join(self.file_dir, 'pos_train_edges_max_comp_graph'))
self.pos_valid_graph = CSFGraph(
os.path.join(self.file_dir, 'pos_validation_edges_max_comp_graph'))
self.pos_test_graph = CSFGraph(
os.path.join(self.file_dir, 'pos_test_edges_max_comp_graph'))
self.neg_train_graph = CSFGraph(
os.path.join(self.file_dir, 'neg_train_edges_max_comp_graph'))
self.neg_valid_graph = CSFGraph(
os.path.join(self.file_dir, 'neg_validation_edges_max_comp_graph'))
self.neg_test_graph = CSFGraph(
os.path.join(self.file_dir, 'neg_test_edges_max_comp_graph'))
self.test_embeddings = os.path.join(self.file_dir, 'test.embeddings')
def read_graphs():
"""
Reads pos_train, pos_vslid, pos_test, neg_train train_valid and neg_test edges with CSFGraph
:return: pos_train, pos_valid, pos_test, neg_train, neg_valid and neg_test graphs in CSFGraph format
"""
start = time.time()
pos_train_graph = CSFGraph(args.pos_train)
pos_valid_graph = CSFGraph(args.pos_valid)
pos_test_graph = CSFGraph(args.pos_test)
neg_train_graph = CSFGraph(args.neg_train)
neg_valid_graph = CSFGraph(args.neg_valid)
neg_test_graph = CSFGraph(args.neg_test)
end = time.time()
logging.info("reading input edge lists files: {} seconds".format(end -
start))
return pos_train_graph, pos_valid_graph, pos_test_graph, neg_train_graph, neg_valid_graph, neg_test_graph
def test_nodetype2count_dictionary(self):
het_g = CSFGraph(edge_file=self.edge_file, node_file=self.node_file)
self.assertIsInstance(self.g.nodetype2count_dictionary, dict)
self.assertEqual(
self.g.nodetype2count_dictionary['biolink:NamedThing'], 11)
self.assertEqual(het_g.nodetype2count_dictionary['biolink:Disease'], 3)
def test_count_edges_legacy_edge_file(self):
g = CSFGraph(edge_file=self.legacy_edge_file)
self.assertEqual(6, g.edge_count())
def test_csfgraph_constructor_accepts_node_file(self):
g = CSFGraph(edge_file=self.edge_file, node_file=self.node_file)
def test_csfgraph_accepts_edge_file(self):
g = CSFGraph(edge_file=self.edge_file)
def test_csfgraph_checks_for_object_column(self):
with self.assertRaises(CSFGraphNoObjectColumnError) as context:
CSFGraph(
edge_file=self.tsv_no_object) # file doesn't have object col
def test_csfgraph_populates_nodetype_to_index_map(self):
het_g = CSFGraph(edge_file=self.edge_file, node_file=self.node_file)
self.assertEqual(het_g.nodetype_to_index_map['biolink:Disease'],
[0, 1, 2])
def test_csfgraph_populates_index_to_nodetype_map(self):
het_g = CSFGraph(edge_file=self.edge_file, node_file=self.node_file)
self.assertEqual(11, len(het_g.index_to_nodetype_map))
self.assertEqual(het_g.index_to_nodetype_map[0], 'biolink:Disease')
def setUp(self):
inputfile = os.path.join(os.path.dirname(__file__), 'data',
'small_graph.txt')
g = CSFGraph(inputfile)
self.g = g
str(g)
def test_csfgraph_tolerates_missing_node_info(self):
het_g = CSFGraph(edge_file=self.edge_file,
node_file=self.node_file_missing_nodes)
self.assertEqual(het_g.index_to_nodetype_map[2],
het_g.default_node_type)
def test_csfgraph_requires_arg(self):
with self.assertRaises(Exception) as context:
CSFGraph() # missing edge arg
self.assertTrue(str('missing' in context.exception))