def read_skype_sample(name_str='non-vpn-app', n=1):
data_path = '../input_data/Flow-Image-Features/skype-sub/all-%d' % n
data_path = '../input_data/Flow-Image-Features/%s-sub/all-%d' % (name_str, n)
train_images_file = '{}/{}pkts-subflow-{}-train-images-idx2-ubyte.gz'.format(data_path, n, name_str)
train_labels_file = '{}/{}pkts-subflow-{}-train-labels-idx1-ubyte.gz'.format(data_path, n, name_str)
test_images_file = '{}/{}pkts-subflow-{}-test-images-idx2-ubyte.gz'.format(data_path, n, name_str)
test_labels_file = '{}/{}pkts-subflow-{}-test-labels-idx1-ubyte.gz'.format(data_path, n, name_str)
# X_train, X_test = np.expand_dims(idx_reader.read_images(train_images_file), 1), np.expand_dims(
# idx_reader.read_images(test_images_file), 1)
X_train, X_test = idx_reader.read_images(train_images_file), idx_reader.read_images(test_images_file)
y_train, y_test = idx_reader.read_labels(train_labels_file), idx_reader.read_labels(test_labels_file)
# return X_train, y_train, X_test, y_test
train_output_file = '%s_%dpkts_train.csv' % (name_str, n)
with open(train_output_file, 'w') as fid_out:
(m, n) = X_train.shape
for row in range(m):
line = ''
for col in range(n):
line += str(X_train[row][col]) + ','
line += str(int(y_train[row])) + '\n'
fid_out.write(line)
test_output_file = '%s_%dpkts_test.csv' % (name_str, n)
with open(test_output_file, 'w') as fid_out:
(m, n) = X_test.shape
for row in range(m):
line = ''
for col in range(n):
line += str(X_test[row][col]) + ','
line += str(int(y_test[row])) + '\n'
fid_out.write(line)
return train_output_file, test_output_file
def read_skype_sample(name_str='facebook', n=784):
data_path = '../input_data/fixed-length-transport-layer-payload/session/{}'.format(name_str)
train_images_file = '{}/{}-byte-payload-per-flow-{}-train-images-idx2-ubyte.gz'.format(data_path, n, name_str)
train_labels_file = '{}/{}-byte-payload-per-flow-{}-train-labels-idx1-ubyte.gz'.format(data_path, n, name_str)
test_images_file = '{}/{}-byte-payload-per-flow-{}-test-images-idx2-ubyte.gz'.format(data_path, n, name_str)
test_labels_file = '{}/{}-byte-payload-per-flow-{}-test-labels-idx1-ubyte.gz'.format(data_path, n, name_str)
# X_train, X_test = np.expand_dims(idx_reader.read_images(train_images_file), 1), np.expand_dims(
# idx_reader.read_images(test_images_file), 1)
X_train, X_test = idx_reader.read_images(train_images_file), idx_reader.read_images(test_images_file)
y_train, y_test = idx_reader.read_labels(train_labels_file), idx_reader.read_labels(test_labels_file)
# return X_train, y_train, X_test, y_test
train_output_file = '%s_%dBytes_train.csv' % (name_str, n)
with open(train_output_file, 'w') as fid_out:
(m, n) = X_train.shape
for row in range(m):
line = ''
for col in range(n):
line += str(X_train[row][col]) + ','
line += str(int(y_train[row])) + '\n'
fid_out.write(line)
test_output_file = '%s_%dBytes_test.csv' % (name_str, n)
with open(test_output_file, 'w') as fid_out:
(m, n) = X_test.shape
for row in range(m):
line = ''
for col in range(n):
line += str(X_test[row][col]) + ','
line += str(int(y_test[row])) + '\n'
fid_out.write(line)
return train_output_file, test_output_file