def gen_training_data(train_graph, node_feature, filename):
print('=========Generating training data=========', file=sys.stderr)
# sample negative samples
neg_edge_num = train_graph.number_of_edges()
neg_edges = sample_negative_edges(train_graph, neg_edge_num)
train_pairs = train_graph.edges()
train_pairs.extend(neg_edges)
random.shuffle(train_pairs)
train_labels = gen_label_mapping(train_graph.edges(), 1)
train_labels.update(gen_label_mapping(neg_edges, -1))
# extract topplogical features
pair_feature = feature_extraction(train_graph, train_pairs, node_feature)
# if you want to output dummy variable, comment the next line
node_feature = None
outfile = open(filename, 'w')
cf.convert_to_svm_format(train_pairs,
node_feature,
pair_feature,
outfile,
testing_ans=train_labels)
outfile.close()
def gen_testing_data(all_graph, test_pairs, test_labels,
lender_feature, loan_feature, filename):
# extract topological features
pair_feature = feature_extraction(all_graph, test_pairs)
outfile = open(filename, 'w')
cf.convert_to_svm_format(test_pairs, lender_feature, loan_feature,
pair_feature, outfile, testing_ans = test_labels)
outfile.close()
def gen_testing_data(all_graph, test_pairs, test_labels,
node_feature, filename):
print('=========Generating testing data=========', file=sys.stderr)
# extract topological features
pair_feature = feature_extraction(all_graph, test_pairs, node_feature)
# if you want to output dummy variable, comment the next line
node_feature = None
outfile = open(filename, 'w')
cf.convert_to_svm_format(test_pairs, node_feature,
pair_feature, outfile, testing_ans = test_labels)
outfile.close()
def gen_testing_data(all_graph, test_pairs, test_labels,
node_feature, filename):
print('=========Generating testing data=========', file=sys.stderr)
# extract topological features
pair_feature = feature_extraction(all_graph, test_pairs, node_feature)
# if you want to output dummy variable, comment the next line
node_feature = None
outfile = open(filename, 'w')
cf.convert_to_svm_format(test_pairs, node_feature,
pair_feature, outfile, testing_ans = test_labels)
outfile.close()
def gen_training_data(train_graph, lender_feature, loan_feature, filename):
# sample negative samples
neg_edge_num = train_graph.number_of_edges()
neg_edges = sample_negative_edges(train_graph, neg_edge_num)
train_pairs = train_graph.edges()
train_pairs.extend(neg_edges)
train_labels = gen_label_mapping(train_graph.edges(), 1)
train_labels.update(gen_label_mapping(neg_edges, 0))
# extract topplogical features
pair_feature = feature_extraction(train_graph, train_pairs)
outfile = open(filename, 'w')
cf.convert_to_svm_format(train_pairs, lender_feature, loan_feature,
pair_feature, outfile, testing_ans = train_labels)
outfile.close()
def gen_training_data(train_graph, node_feature, filename):
print('=========Generating training data=========', file=sys.stderr)
# sample negative samples
neg_edge_num = train_graph.number_of_edges()
neg_edges = sample_negative_edges(train_graph, neg_edge_num)
train_pairs = train_graph.edges()
train_pairs.extend(neg_edges)
random.shuffle(train_pairs)
train_labels = gen_label_mapping(train_graph.edges(), 1)
train_labels.update(gen_label_mapping(neg_edges, -1))
# extract topplogical features
pair_feature = feature_extraction(train_graph, train_pairs, node_feature)
# if you want to output dummy variable, comment the next line
node_feature = None
outfile = open(filename, 'w')
cf.convert_to_svm_format(train_pairs, node_feature,
pair_feature, outfile, testing_ans = train_labels)
outfile.close()