def compute_model_dependencies( self, surrogate: Surrogate, bounds: List[Tuple[float, float]], objective_weights: Tensor, target_fidelities: Optional[Dict[int, float]] = None, pending_observations: Optional[List[Tensor]] = None, outcome_constraints: Optional[Tuple[Tensor, Tensor]] = None, linear_constraints: Optional[Tuple[Tensor, Tensor]] = None, fixed_features: Optional[Dict[int, float]] = None, options: Optional[Dict[str, Any]] = None, ) -> Dict[str, Any]: # Compute generic multi-fidelity dependencies first dependencies = super().compute_model_dependencies( surrogate=surrogate, bounds=bounds, objective_weights=objective_weights, pending_observations=pending_observations, outcome_constraints=outcome_constraints, linear_constraints=linear_constraints, fixed_features=fixed_features, target_fidelities=target_fidelities, options=options, ) _, best_point_acqf_value = surrogate.best_in_sample_point( bounds=bounds, objective_weights=objective_weights, outcome_constraints=outcome_constraints, linear_constraints=linear_constraints, fixed_features=fixed_features, ) dependencies.update({Keys.CURRENT_VALUE: best_point_acqf_value}) return dependencies
class SurrogateTest(TestCase): def setUp(self): self.botorch_model_class = SingleTaskGP self.mll_class = ExactMarginalLogLikelihood self.device = torch.device("cpu") self.dtype = torch.float self.X = torch.tensor([[1.0, 2.0, 3.0], [2.0, 3.0, 4.0]], dtype=self.dtype, device=self.device) self.Y = torch.tensor([[3.0], [4.0]], dtype=self.dtype, device=self.device) self.Yvar = torch.tensor([[0.0], [2.0]], dtype=self.dtype, device=self.device) self.training_data = TrainingData(X=self.X, Y=self.Y, Yvar=self.Yvar) self.surrogate_kwargs = self.botorch_model_class.construct_inputs( self.training_data) self.surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=self.mll_class) self.bounds = [(0.0, 1.0), (1.0, 4.0), (2.0, 5.0)] self.task_features = [] self.feature_names = ["x1", "x2", "x3"] self.metric_names = ["y"] self.fidelity_features = [] self.target_fidelities = {1: 1.0} self.fixed_features = {1: 2.0} self.refit = True self.objective_weights = torch.tensor([-1.0, 1.0], dtype=self.dtype, device=self.device) self.outcome_constraints = (torch.tensor([[1.0]]), torch.tensor([[0.5] ])) self.linear_constraints = ( torch.tensor([[0.0, 0.0, 0.0], [0.0, 1.0, 0.0]]), torch.tensor([[0.5], [1.0]]), ) self.options = {} @patch(f"{CURRENT_PATH}.Kernel") @patch(f"{CURRENT_PATH}.Likelihood") def test_init(self, mock_Likelihood, mock_Kernel): self.assertEqual(self.surrogate.botorch_model_class, self.botorch_model_class) self.assertEqual(self.surrogate.mll_class, self.mll_class) with self.assertRaisesRegex(NotImplementedError, "Customizing likelihood"): Surrogate(botorch_model_class=self.botorch_model_class, likelihood=Likelihood()) with self.assertRaisesRegex(NotImplementedError, "Customizing kernel"): Surrogate(botorch_model_class=self.botorch_model_class, kernel_class=Kernel()) def test_model_property(self): with self.assertRaisesRegex( ValueError, "BoTorch `Model` has not yet been constructed."): self.surrogate.model def test_training_data_property(self): with self.assertRaisesRegex( ValueError, "Underlying BoTorch `Model` has not yet received its training_data.", ): self.surrogate.training_data def test_dtype_property(self): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) self.assertEqual(self.dtype, self.surrogate.dtype) def test_device_property(self): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) self.assertEqual(self.device, self.surrogate.device) def test_from_BoTorch(self): surrogate = Surrogate.from_BoTorch( self.botorch_model_class(**self.surrogate_kwargs)) self.assertIsInstance(surrogate.model, self.botorch_model_class) self.assertFalse(surrogate._should_reconstruct) @patch(f"{CURRENT_PATH}.SingleTaskGP.__init__", return_value=None) def test_construct(self, mock_GP): base_surrogate = Surrogate(botorch_model_class=Model) with self.assertRaisesRegex(TypeError, "Cannot construct an abstract model."): base_surrogate.construct( training_data=self.training_data, fidelity_features=self.fidelity_features, ) self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) mock_GP.assert_called_with(train_X=self.X, train_Y=self.Y) @patch(f"{CURRENT_PATH}.SingleTaskGP.load_state_dict", return_value=None) @patch(f"{CURRENT_PATH}.ExactMarginalLogLikelihood") @patch(f"{SURROGATE_PATH}.fit_gpytorch_model") def test_fit(self, mock_fit_gpytorch, mock_MLL, mock_state_dict): surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=ExactMarginalLogLikelihood, ) # Checking that model is None before `fit` (and `construct`) calls. self.assertIsNone(surrogate._model) # Should instantiate mll and `fit_gpytorch_model` when `state_dict` # is `None`. surrogate.fit( training_data=self.training_data, bounds=self.bounds, task_features=self.task_features, feature_names=self.feature_names, metric_names=self.metric_names, fidelity_features=self.fidelity_features, target_fidelities=self.target_fidelities, refit=self.refit, ) mock_state_dict.assert_not_called() mock_MLL.assert_called_once() mock_fit_gpytorch.assert_called_once() mock_state_dict.reset_mock() mock_MLL.reset_mock() mock_fit_gpytorch.reset_mock() # Should `load_state_dict` when `state_dict` is not `None` # and `refit` is `False`. state_dict = {} surrogate.fit( training_data=self.training_data, bounds=self.bounds, task_features=self.task_features, feature_names=self.feature_names, metric_names=self.metric_names, fidelity_features=self.fidelity_features, target_fidelities=self.target_fidelities, refit=False, state_dict=state_dict, ) mock_state_dict.assert_called_once() mock_MLL.assert_not_called() mock_fit_gpytorch.assert_not_called() @patch(f"{SURROGATE_PATH}.predict_from_model") def test_predict(self, mock_predict): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) self.surrogate.predict(X=self.X) mock_predict.assert_called_with(model=self.surrogate.model, X=self.X) def test_best_in_sample_point(self): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) # `best_in_sample_point` requires `objective_weights` with patch(f"{SURROGATE_PATH}.best_in_sample_point", return_value=None) as mock_best_in_sample: with self.assertRaisesRegex(ValueError, "Could not obtain"): self.surrogate.best_in_sample_point(bounds=self.bounds, objective_weights=None) with patch(f"{SURROGATE_PATH}.best_in_sample_point", return_value=(self.X, 0.0)) as mock_best_in_sample: best_point, observed_value = self.surrogate.best_in_sample_point( bounds=self.bounds, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=self.fixed_features, options=self.options, ) mock_best_in_sample.assert_called_with( Xs=[self.training_data.X], model=self.surrogate, bounds=self.bounds, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=self.fixed_features, options=self.options, ) @patch(f"{ACQUISITION_PATH}.Acquisition.__init__", return_value=None) @patch( f"{ACQUISITION_PATH}.Acquisition.optimize", return_value=([torch.tensor([0.0])], [torch.tensor([1.0])]), ) @patch( f"{SURROGATE_PATH}.pick_best_out_of_sample_point_acqf_class", return_value=(qSimpleRegret, { Keys.SAMPLER: SobolQMCNormalSampler }), ) def test_best_out_of_sample_point(self, mock_best_point_util, mock_acqf_optimize, mock_acqf_init): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) # currently cannot use function with fixed features with self.assertRaisesRegex(NotImplementedError, "Fixed features"): self.surrogate.best_out_of_sample_point( bounds=self.bounds, objective_weights=self.objective_weights, fixed_features=self.fixed_features, ) candidate, acqf_value = self.surrogate.best_out_of_sample_point( bounds=self.bounds, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fidelity_features=self.fidelity_features, target_fidelities=self.target_fidelities, options=self.options, ) mock_acqf_init.assert_called_with( surrogate=self.surrogate, botorch_acqf_class=qSimpleRegret, bounds=self.bounds, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=None, target_fidelities=self.target_fidelities, options={Keys.SAMPLER: SobolQMCNormalSampler}, ) self.assertTrue(torch.equal(candidate, torch.tensor([0.0]))) self.assertTrue(torch.equal(acqf_value, torch.tensor([1.0]))) @patch(f"{SURROGATE_PATH}.Surrogate.fit") def test_update(self, mock_fit): self.surrogate.construct(training_data=self.training_data, fidelity_features=self.fidelity_features) # Call `fit` by default self.surrogate.update( training_data=self.training_data, bounds=self.bounds, task_features=self.task_features, feature_names=self.feature_names, metric_names=self.metric_names, fidelity_features=self.fidelity_features, refit=self.refit, ) mock_fit.assert_called_with( training_data=self.training_data, bounds=self.bounds, task_features=self.task_features, feature_names=self.feature_names, metric_names=self.metric_names, fidelity_features=self.fidelity_features, candidate_metadata=None, state_dict=self.surrogate.model.state_dict, refit=self.refit, ) # If should not be reconstructed, raise Error self.surrogate._should_reconstruct = False with self.assertRaisesRegex( NotImplementedError, ".* models that should not be re-constructed"): self.surrogate.update( training_data=self.training_data, bounds=self.bounds, task_features=self.task_features, feature_names=self.feature_names, metric_names=self.metric_names, fidelity_features=self.fidelity_features, refit=self.refit, )
class SurrogateTest(TestCase): def setUp(self): self.botorch_model_class = SingleTaskGP self.mll_class = ExactMarginalLogLikelihood self.device = torch.device("cpu") self.dtype = torch.float self.Xs, self.Ys, self.Yvars, self.bounds, _, _, _ = get_torch_test_data( dtype=self.dtype) self.training_data = TrainingData.from_block_design(X=self.Xs[0], Y=self.Ys[0], Yvar=self.Yvars[0]) self.surrogate_kwargs = self.botorch_model_class.construct_inputs( self.training_data) self.surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=self.mll_class) self.search_space_digest = SearchSpaceDigest( feature_names=["x1", "x2"], bounds=self.bounds, target_fidelities={1: 1.0}, ) self.metric_names = ["y"] self.fixed_features = {1: 2.0} self.refit = True self.objective_weights = torch.tensor([-1.0, 1.0], dtype=self.dtype, device=self.device) self.outcome_constraints = (torch.tensor([[1.0]]), torch.tensor([[0.5] ])) self.linear_constraints = ( torch.tensor([[0.0, 0.0, 0.0], [0.0, 1.0, 0.0]]), torch.tensor([[0.5], [1.0]]), ) self.options = {} @patch(f"{CURRENT_PATH}.Kernel") @patch(f"{CURRENT_PATH}.Likelihood") def test_init(self, mock_Likelihood, mock_Kernel): self.assertEqual(self.surrogate.botorch_model_class, self.botorch_model_class) self.assertEqual(self.surrogate.mll_class, self.mll_class) with self.assertRaisesRegex(NotImplementedError, "Customizing likelihood"): Surrogate(botorch_model_class=self.botorch_model_class, likelihood=Likelihood()) with self.assertRaisesRegex(NotImplementedError, "Customizing kernel"): Surrogate(botorch_model_class=self.botorch_model_class, kernel_class=Kernel()) @patch(f"{SURROGATE_PATH}.fit_gpytorch_model") def test_mll_options(self, _): mock_mll = MagicMock(self.mll_class) surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=mock_mll, mll_options={"some_option": "some_value"}, ) surrogate.fit( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=self.refit, ) self.assertEqual(mock_mll.call_args[1]["some_option"], "some_value") def test_model_property(self): with self.assertRaisesRegex( ValueError, "BoTorch `Model` has not yet been constructed."): self.surrogate.model def test_training_data_property(self): with self.assertRaisesRegex( ValueError, "Underlying BoTorch `Model` has not yet received its training_data.", ): self.surrogate.training_data def test_dtype_property(self): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) self.assertEqual(self.dtype, self.surrogate.dtype) def test_device_property(self): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) self.assertEqual(self.device, self.surrogate.device) def test_from_botorch(self): surrogate = Surrogate.from_botorch( self.botorch_model_class(**self.surrogate_kwargs)) self.assertIsInstance(surrogate.model, self.botorch_model_class) self.assertTrue(surrogate._constructed_manually) @patch(f"{CURRENT_PATH}.SingleTaskGP.__init__", return_value=None) def test_construct(self, mock_GP): with self.assertRaises(NotImplementedError): # Base `Model` does not implement `construct_inputs`. Surrogate(botorch_model_class=Model).construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) mock_GP.assert_called_once() call_kwargs = mock_GP.call_args[1] self.assertTrue(torch.equal(call_kwargs["train_X"], self.Xs[0])) self.assertTrue(torch.equal(call_kwargs["train_Y"], self.Ys[0])) self.assertFalse(self.surrogate._constructed_manually) # Check that `model_options` passed to the `Surrogate` constructor are # properly propagated. with patch.object( SingleTaskGP, "construct_inputs", wraps=SingleTaskGP.construct_inputs) as mock_construct_inputs: surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=self.mll_class, model_options={"some_option": "some_value"}, ) surrogate.construct(self.training_data) mock_construct_inputs.assert_called_with( training_data=self.training_data, some_option="some_value") @patch(f"{CURRENT_PATH}.SingleTaskGP.load_state_dict", return_value=None) @patch(f"{CURRENT_PATH}.ExactMarginalLogLikelihood") @patch(f"{SURROGATE_PATH}.fit_gpytorch_model") def test_fit(self, mock_fit_gpytorch, mock_MLL, mock_state_dict): surrogate = Surrogate( botorch_model_class=self.botorch_model_class, mll_class=ExactMarginalLogLikelihood, ) # Checking that model is None before `fit` (and `construct`) calls. self.assertIsNone(surrogate._model) # Should instantiate mll and `fit_gpytorch_model` when `state_dict` # is `None`. surrogate.fit( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=self.refit, ) # Check that training data is correctly passed through to the # BoTorch `Model`. self.assertTrue( torch.equal( surrogate.model.train_inputs[0], self.surrogate_kwargs.get("train_X"), )) self.assertTrue( torch.equal( surrogate.model.train_targets, self.surrogate_kwargs.get("train_Y").squeeze(1), )) mock_state_dict.assert_not_called() mock_MLL.assert_called_once() mock_fit_gpytorch.assert_called_once() mock_state_dict.reset_mock() mock_MLL.reset_mock() mock_fit_gpytorch.reset_mock() # Should `load_state_dict` when `state_dict` is not `None` # and `refit` is `False`. state_dict = {"state_attribute": "value"} surrogate.fit( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=False, state_dict=state_dict, ) mock_state_dict.assert_called_once() mock_MLL.assert_not_called() mock_fit_gpytorch.assert_not_called() @patch(f"{SURROGATE_PATH}.predict_from_model") def test_predict(self, mock_predict): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) self.surrogate.predict(X=self.Xs[0]) mock_predict.assert_called_with(model=self.surrogate.model, X=self.Xs[0]) def test_best_in_sample_point(self): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) # `best_in_sample_point` requires `objective_weights` with patch(f"{SURROGATE_PATH}.best_in_sample_point", return_value=None) as mock_best_in_sample: with self.assertRaisesRegex(ValueError, "Could not obtain"): self.surrogate.best_in_sample_point( search_space_digest=self.search_space_digest, objective_weights=None) with patch(f"{SURROGATE_PATH}.best_in_sample_point", return_value=(self.Xs[0], 0.0)) as mock_best_in_sample: best_point, observed_value = self.surrogate.best_in_sample_point( search_space_digest=self.search_space_digest, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=self.fixed_features, options=self.options, ) mock_best_in_sample.assert_called_with( Xs=[self.training_data.X], model=self.surrogate, bounds=self.search_space_digest.bounds, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=self.fixed_features, options=self.options, ) @patch(f"{ACQUISITION_PATH}.Acquisition.__init__", return_value=None) @patch( f"{ACQUISITION_PATH}.Acquisition.optimize", return_value=([torch.tensor([0.0])], [torch.tensor([1.0])]), ) @patch( f"{SURROGATE_PATH}.pick_best_out_of_sample_point_acqf_class", return_value=(qSimpleRegret, { Keys.SAMPLER: SobolQMCNormalSampler }), ) def test_best_out_of_sample_point(self, mock_best_point_util, mock_acqf_optimize, mock_acqf_init): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) # currently cannot use function with fixed features with self.assertRaisesRegex(NotImplementedError, "Fixed features"): self.surrogate.best_out_of_sample_point( search_space_digest=self.search_space_digest, objective_weights=self.objective_weights, fixed_features=self.fixed_features, ) candidate, acqf_value = self.surrogate.best_out_of_sample_point( search_space_digest=self.search_space_digest, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, options=self.options, ) mock_acqf_init.assert_called_with( surrogate=self.surrogate, botorch_acqf_class=qSimpleRegret, search_space_digest=self.search_space_digest, objective_weights=self.objective_weights, outcome_constraints=self.outcome_constraints, linear_constraints=self.linear_constraints, fixed_features=None, options={Keys.SAMPLER: SobolQMCNormalSampler}, ) self.assertTrue(torch.equal(candidate, torch.tensor([0.0]))) self.assertTrue(torch.equal(acqf_value, torch.tensor([1.0]))) @patch(f"{CURRENT_PATH}.SingleTaskGP.load_state_dict", return_value=None) @patch(f"{CURRENT_PATH}.ExactMarginalLogLikelihood") @patch(f"{SURROGATE_PATH}.fit_gpytorch_model") def test_update(self, mock_fit_gpytorch, mock_MLL, mock_state_dict): self.surrogate.construct( training_data=self.training_data, fidelity_features=self.search_space_digest.fidelity_features, ) # Check that correct arguments are passed to `fit`. with patch(f"{SURROGATE_PATH}.Surrogate.fit") as mock_fit: # Call `fit` by default self.surrogate.update( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=self.refit, state_dict={"key": "val"}, ) mock_fit.assert_called_with( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, candidate_metadata=None, refit=self.refit, state_dict={"key": "val"}, ) # Check that the training data is correctly passed through to the # BoTorch `Model`. Xs, Ys, Yvars, bounds, _, _, _ = get_torch_test_data(dtype=self.dtype, offset=1.0) training_data = TrainingData.from_block_design(X=Xs[0], Y=Ys[0], Yvar=Yvars[0]) surrogate_kwargs = self.botorch_model_class.construct_inputs( training_data) self.surrogate.update( training_data=training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=self.refit, state_dict={"key": "val"}, ) self.assertTrue( torch.equal( self.surrogate.model.train_inputs[0], surrogate_kwargs.get("train_X"), )) self.assertTrue( torch.equal( self.surrogate.model.train_targets, surrogate_kwargs.get("train_Y").squeeze(1), )) # If should not be reconstructed, check that error is raised. self.surrogate._constructed_manually = True with self.assertRaisesRegex(NotImplementedError, ".* constructed manually"): self.surrogate.update( training_data=self.training_data, search_space_digest=self.search_space_digest, metric_names=self.metric_names, refit=self.refit, ) def test_serialize_attributes_as_kwargs(self): expected = self.surrogate.__dict__ self.assertEqual(self.surrogate._serialize_attributes_as_kwargs(), expected)