Make all generators know the dim (#513)

Summary:
Pull Request resolved: #513

Some of the generators know the dimensionality of the search space but others don't. Unified the API such that every generator has a dim attribute. This allows other classes working with generators to be able to always access the dimensionality (even if the bounds are not available).

Reviewed By: crasanders

Differential Revision: D67997208

fbshipit-source-id: 3a2aff9945fbc2cbb70e74e92cc52c2b8dc027a8

aepsych/generators/acqf_thompson_sampler_generator.py

@@ -67,6 +67,7 @@ def __init__(
         self.stimuli_per_trial = stimuli_per_trial
         self.lb = lb
         self.ub = ub
+        self.dim = len(lb)
 
     def _instantiate_acquisition_fn(self, model: ModelProtocol) -> AcquisitionFunction:
         """Instantiate the acquisition function with the model and any extra arguments.

aepsych/generators/base.py

@@ -44,6 +44,7 @@ class AEPsychGenerator(abc.ABC, Generic[AEPsychModelType]):
 
     acqf: AcquisitionFunction
     acqf_kwargs: Dict[str, Any]
+    dim: int
 
     def __init__(
         self,

aepsych/generators/epsilon_greedy_generator.py

@@ -34,6 +34,7 @@ def __init__(
         self.epsilon = epsilon
         self.lb = lb
         self.ub = ub
+        self.dim = len(lb)
 
     @classmethod
     def from_config(cls, config: Config) -> "EpsilonGreedyGenerator":

aepsych/generators/monotonic_rejection_generator.py

@@ -68,7 +68,7 @@ def __init__(
         self.acqf_kwargs = acqf_kwargs
         self.model_gen_options = model_gen_options
         self.explore_features = explore_features
-        self.lb, self.ub, _ = _process_bounds(lb, ub, None)
+        self.lb, self.ub, self.dim = _process_bounds(lb, ub, None)
         self.bounds = torch.stack((self.lb, self.ub))
 
     def _instantiate_acquisition_fn(

aepsych/generators/monotonic_thompson_sampler_generator.py

@@ -30,6 +30,7 @@ def __init__(
         num_ts_points: int,
         target_value: float,
         objective: MCAcquisitionObjective,
+        dim: int,
         explore_features: Optional[List[Type[int]]] = None,
     ) -> None:
         """Initialize MonotonicMCAcquisition
@@ -41,6 +42,7 @@ def __init__(
             target_value (float): target value that is being looked for
             objective (MCAcquisitionObjective): Objective transform of the GP output
                 before evaluating the acquisition. Defaults to identity transform.
+            dim (int): Dimensionality of the model.
             explore_features (List[Type[int]], optional): List of features that will be selected randomly and then
                 fixed for acquisition fn optimization. Defaults to None.
         """
@@ -50,6 +52,7 @@ def __init__(
         self.target_value = target_value
         self.objective = objective()
         self.explore_features = explore_features
+        self.dim = dim
 
     def gen(
         self,

aepsych/generators/optimize_acqf_generator.py

@@ -71,6 +71,7 @@ def __init__(
         self.stimuli_per_trial = stimuli_per_trial
         self.lb = lb
         self.ub = ub
+        self.dim = len(lb)
 
     def _instantiate_acquisition_fn(self, model: ModelProtocol) -> AcquisitionFunction:
         """