train_file = sys.argv[1]
test_file = sys.argv[2]
test_ans_file = sys.argv[3]
user_profile_file = sys.argv[4]
config_file = sys.argv[5]
out_train_file = sys.argv[6]
out_test_file = sys.argv[7]
# read in data
train_graph = file_io.read_graph(train_file)
config = file_io.read_config(config_file)
user_feature = None
(user_feature,
feature_name) = file_io.read_feature_column_major(user_profile_file,
config)
#normalize features
for column in user_feature:
if column.type == 'numerical':
cf.normalize_column(column)
elif column.type == 'categorical':
cf.convert_to_dummy_variable(column)
test_pair = file_io.read_data(test_file)
train_graph = update_nodes_from_test_data(train_graph, test_pair)
test_ans = gen_label_mapping(test_pair, file_io.read_ans(test_ans_file))
gen_training_data(train_graph, user_feature, out_train_file)
gen_testing_data(train_graph, test_pair, test_ans, user_feature,
out_test_file)
# arguments
train_file = sys.argv[1]
test_file = sys.argv[2]
test_ans_file = sys.argv[3]
user_profile_file = sys.argv[4]
config_file = sys.argv[5]
out_train_file = sys.argv[6]
out_test_file = sys.argv[7]
# read in data
train_graph = file_io.read_graph(train_file)
config = file_io.read_config(config_file)
user_feature = None
(user_feature, feature_name) = file_io.read_feature_column_major(user_profile_file, config)
#normalize features
for column in user_feature:
if column.type == 'numerical':
cf.normalize_column(column)
elif column.type == 'categorical':
cf.convert_to_dummy_variable(column)
test_pair = file_io.read_data(test_file)
train_graph = update_nodes_from_test_data(train_graph, test_pair)
test_ans = gen_label_mapping(test_pair, file_io.read_ans(test_ans_file))
gen_training_data(train_graph, user_feature, out_train_file)
gen_testing_data(train_graph, test_pair, test_ans,
user_feature, out_test_file)