def test_model_to_disk(self):
np.random.seed(0)
tf.random.set_seed(0)
data, discrete = generate_data(self._vars['batch_size'])
model = CTGANSynthesizer(batch_size=self._vars['batch_size'],
pac=self._vars['pac'])
self.assertIsNotNone(model)
model.train(data, discrete, epochs=1)
model_path = os.path.join(self._current_dir, 'model_test.joblib')
model.dump(model_path, overwrite=True)
loaded_model = CTGANSynthesizer(file_path=model_path)
self.assertIsNotNone(loaded_model)
for attr, value in loaded_model.__dict__.items():
self.assertTrue(attr in model.__dict__)
if type(value) in [int, float, tuple]:
self.assertEqual(value, model.__dict__[attr])
np.testing.assert_equal(loaded_model._cond_generator.__dict__,
model._cond_generator.__dict__)
for attr, value in loaded_model._transformer.__dict__.items():
if isinstance(value, pd.Series):
pd.testing.assert_series_equal(
value, model._transformer.__dict__[attr])
elif isinstance(value, list) and isinstance(value[0], tf.Tensor):
tf.assert_equal(value, model._transformer.__dict__[attr])
else:
np.testing.assert_equal(value,
model._transformer.__dict__[attr])
np.testing.assert_equal(loaded_model._generator.get_weights(),
model._generator.get_weights())
def test_inverse_transform(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
transformed_data = transformer.transform(data)
inverse_data = transformer.inverse_transform(transformed_data)
pd.testing.assert_frame_equal(data, inverse_data)
def test_sample(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
train_data = transformer.transform(data)
cond_gen = ConditionalGenerator(
train_data, transformer.output_info, True)
output = cond_gen.sample(self._vars['batch_size'])
self.assertIsNotNone(output)
c, m, col, opt = output
expected_c = np.array([
[1., 0., 0., 0.],
[0., 0., 1., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 0., 0., 1.],
[0., 1., 0., 0.]], dtype=np.float32)
expected_m = np.array([[1.],
[1.],
[1.],
[1.],
[1.],
[1.],
[1.],
[1.],
[1.],
[1.]], dtype=np.float32)
expected_col = np.array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0])
expected_opt = np.array([0, 2, 0, 0, 0, 0, 0, 1, 3, 1])
np.testing.assert_equal(c, expected_c)
np.testing.assert_equal(m, expected_m)
np.testing.assert_equal(col, expected_col)
np.testing.assert_equal(opt, expected_opt)
output = cond_gen.sample_zero(self._vars['batch_size'])
self.assertIsNotNone(output)
print(output)
expected_output = [
[0., 0., 1., 0.],
[0., 1., 0., 0.],
[0., 1., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[1., 0., 0., 0.],
[0., 0., 1., 0.],
[1., 0., 0., 0.],
[0., 1., 0., 0.]]
np.testing.assert_equal(output, expected_output)
def test_fit(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
expected_info = [(1, 'tanh', 1), (1, 'softmax', 1), (4, 'softmax', 0)]
expected_dimensions = 6
np.testing.assert_equal(transformer.output_info, expected_info)
np.testing.assert_equal(transformer.output_dimensions,
expected_dimensions)
def test_sample_none(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, [])
train_data = transformer.transform(data)
sampler = DataSampler(train_data, transformer.output_info)
output = sampler.sample(1, None, None)
expected_output = np.array([[0.46909913, 1., 0.08564136, 0., 1.]])
np.testing.assert_almost_equal(output, expected_output,
self._vars['decimal'])
def test_sample(self):
np.random.seed(0)
tf.random.set_seed(0)
data, discrete = generate_data(self._vars['batch_size'])
model = CTGANSynthesizer(
batch_size=self._vars['batch_size'], pac=self._vars['pac'])
self.assertIsNotNone(model)
model.train(data, discrete, epochs=1)
output = model.sample(self._n_samples).values
expected_output = np.array([[0.4139329, 3.0]])
np.testing.assert_almost_equal(
output, expected_output, decimal=self._vars['decimal'])
def test_sample_none(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, [])
train_data = transformer.transform(data)
cond_gen = ConditionalGenerator(train_data, transformer.output_info,
True)
output = cond_gen.sample(self._vars['batch_size'])
self.assertIsNone(output)
output = cond_gen.sample_zero(self._vars['batch_size'])
self.assertIsNone(output)
def test_tensors(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
transformer.generate_tensors()
expected_info = [
tf.constant([0, 1, 0], dtype=tf.int32),
tf.constant([1, 2, 1], dtype=tf.int32),
tf.constant([2, 6, 1], dtype=tf.int32)
]
expected_cond = [tf.constant([2, 6, 0, 4, 0], dtype=tf.int32)]
tf.assert_equal(expected_info, transformer.output_tensor)
tf.assert_equal(expected_cond, transformer.cond_tensor)
def test_sample(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
train_data = transformer.transform(data)
sampler = DataSampler(train_data, transformer.output_info)
output = sampler.sample(1, [0, 0], [0, 0])
expected_output = np.array([[0.3721639, 1., 1., 0., 0., 0.],
[-0.31326372, 1., 1., 0., 0., 0.]])
np.testing.assert_almost_equal(output,
expected_output,
decimal=self._vars['decimal'])
def test_transform(self):
np.random.seed(0)
data, discrete = generate_data(self._vars['batch_size'])
transformer = DataTransformer()
transformer.fit(data, discrete)
transformed_data = transformer.transform(data)
expected_data = np.array([[-0.09027826, 1., 1., 0., 0., 0.],
[0.1340608, 1., 0., 1., 0., 0.],
[-0.01753295, 1., 0., 1., 0., 0.],
[-0.09557786, 1., 1., 0., 0., 0.],
[-0.25904065, 1., 0., 1., 0., 0.],
[0.04062398, 1., 0., 0., 0., 1.],
[-0.24025436, 1., 0., 0., 1., 0.],
[0.3721639, 1., 1., 0., 0., 0.],
[0.46909913, 1., 0., 0., 1., 0.],
[-0.31326372, 1., 1., 0., 0., 0.]])
np.testing.assert_almost_equal(transformed_data,
expected_data,
decimal=self._vars['decimal'])
def test_train(self):
np.random.seed(0)
tf.random.set_seed(0)
data, discrete = generate_data(self._vars['batch_size'])
self._assert_train_equal(data, [])
self._assert_train_equal(data, discrete)