def load_data(if_norm=True):
print('Loading data set ...')
load_time = time.time()
test_data = read_test('./data/test.csv')
test_data = np.array(test_data)
train_data = read_train('./data/train.csv')
train_data = np.array(train_data)
loaded_time = time.time() - load_time
print('test_data: {0}, train_data shape: {1}'.format(
test_data.shape, train_data.shape))
train_features, train_labels = train_data[:,
1:-1], train_data[:,
-1].astype(int)
test_features = test_data[:, 1:]
if if_norm is True:
test_min, test_max = test_features.min(), test_features.max()
test_features_norm = (test_features - test_min) / (test_max - test_min)
train_min, train_max = train_features.min(), train_features.max()
train_features_norm = (train_features - train_min) / (train_max -
train_min)
print('Data set loaded successfully in {0:.4f} seconds.'.format(
loaded_time))
return test_features_norm, train_features_norm, train_labels
else:
print('Data set loaded successfully in {0:.4f} seconds.'.format(
loaded_time))
return test_features, train_features, train_labels