update tests, fix a bug and address typos

xopt/generators/bayesian/bayesian_generator.py

@@ -411,10 +411,18 @@ def propose_candidates(self, model, n_candidates=1):
         # get initial candidates to start acquisition function optimization
         initial_points = self._get_initial_conditions(n_candidates)
 
-        # get candidates
-        candidates = self.numerical_optimizer.optimize(
-            acq_funct, bounds, n_candidates, batch_initial_conditions=initial_points
-        )
+        # get candidates -- grid optimizer does not support batch_initial_conditions
+        if isinstance(self.numerical_optimizer, GridOptimizer):
+            candidates = self.numerical_optimizer.optimize(
+                acq_funct, bounds, n_candidates
+            )
+        else:
+            candidates = self.numerical_optimizer.optimize(
+                acq_funct,
+                bounds,
+                n_candidates,
+                batch_initial_conditions=initial_points
+            )
         return candidates
 
     def get_training_data(self, data: pd.DataFrame) -> pd.DataFrame:
@@ -799,7 +807,7 @@ def _get_scaled_data(self):
         weights = set_botorch_weights(self.vocs).to(**self._tkwargs)[
             : self.vocs.n_objectives
         ]
-        return variable_data, objective_data * weights
+        return variable_data, objective_data * weights, weights
 
     def calculate_hypervolume(self):
         """compute hypervolume given data"""
@@ -814,7 +822,7 @@ def calculate_hypervolume(self):
 
     def get_pareto_front(self):
         """compute the pareto front x/y values given data"""
-        variable_data, objective_data = self._get_scaled_data()
+        variable_data, objective_data, weights = self._get_scaled_data()
         obj_data = torch.vstack(
             (self.torch_reference_point.unsqueeze(0), objective_data)
         )
@@ -826,10 +834,6 @@ def get_pareto_front(self):
         )
         non_dominated = is_non_dominated(obj_data)
 
-        weights = set_botorch_weights(self.vocs).to(**self._tkwargs)[
-            : self.vocs.n_objectives
-        ]
-
         # note need to undo weights for real number output
         # only return values if non nan values exist
         if torch.all(torch.isnan(var_data[non_dominated])):

xopt/generators/bayesian/mobo.py

@@ -21,7 +21,7 @@ class MOBOGenerator(MultiObjectiveBayesianGenerator):
     supports_batch_generation: bool = True
     use_pf_as_initial_points: bool = Field(
         False,
-        description="flag to specify if pf front points are to be used during "
+        description="flag to specify if pareto front points are to be used during "
         "optimization of the acquisition function",
     )
     __doc__ = """Implements Multi-Objective Bayesian Optimization using the Expected

xopt/numerical_optimizer.py

@@ -70,7 +70,7 @@ class GridOptimizer(NumericalOptimizer):
         10, description="number of grid points per axis used for optimization"
     )
 
-    def optimize(self, function, bounds, n_candidates=1, **kwargs):
+    def optimize(self, function, bounds, n_candidates=1):
         assert isinstance(bounds, Tensor)
         # create mesh
         if len(bounds) != 2:

xopt/tests/generators/bayesian/test_mobo.py

@@ -168,6 +168,7 @@ def test_initial_conditions(self):
             reference_point=reference_point,
             use_pf_as_initial_points=True,
         )
+        gen.numerical_optimizer.max_time = 1.0
         gen.add_data(test_data)
         initial_points = gen._get_initial_conditions()
 
@@ -196,11 +197,11 @@ def test_initial_conditions(self):
         vocs.constraints = {"c1": ["GREATER_THAN", 0.5]}
         test_data = pd.DataFrame(
             {
-                "x1": [0.1, 0.2, 0.4, 0.4],
-                "x2": [0.1, 0.2, 0.3, 0.2],
-                "y1": [1.0, 2.0, 1.0, 0.0],
-                "y2": [0.5, 0.1, 1.0, 1.5],
-                "c1": [1.0, 1.0, 1.0, 1.0],
+                "x1": [0.1, 0.2, 0.4, 0.4, 0.15],
+                "x2": [0.1, 0.2, 0.3, 0.2, 0.15],
+                "y1": [1.0, 2.0, 1.0, 0.0, 1.5],
+                "y2": [0.5, 0.1, 1.0, 1.5, 0.25],
+                "c1": [1.0, 1.0, 1.0, 1.0, 0.0],
             }
         )
         gen = MOBOGenerator(
@@ -209,7 +210,17 @@ def test_initial_conditions(self):
             use_pf_as_initial_points=True,
         )
         gen.add_data(test_data)
-        gen._get_initial_conditions()
+        gen.numerical_optimizer.max_time = 1.0
+
+        # make sure that no invalid points make it into the initial conditions
+        ic = gen._get_initial_conditions()
+        assert not torch.allclose(
+            ic[:4],
+            torch.tensor(
+                ((0.1, 0.1), (0.2, 0.2), (0.4, 0.2), (0.15, 0.15))
+            ).reshape(4, 1, 2).double()
+        )
+
         gen.generate(1)
 
     def test_log_mobo(self):

xopt/vocs.py

@@ -561,12 +561,12 @@ def extract_data(self, data: pd.DataFrame, return_raw=False, return_valid=False)
         observable_data = self.observable_data(data, "")
 
         if return_valid:
-            feasable_status = self.feasibility_data(data)["feasible"]
+            feasible_status = self.feasibility_data(data)["feasible"]
             return (
-                variable_data[feasable_status],
-                objective_data[feasable_status],
-                constraint_data[feasable_status],
-                observable_data[feasable_status],
+                variable_data[feasible_status],
+                objective_data[feasible_status],
+                constraint_data[feasible_status],
+                observable_data[feasible_status],
             )
 
         return variable_data, objective_data, constraint_data, observable_data