Allow removal of parameters with different values in RemoveFixed tran…

…sform (#2779)

Summary:
Pull Request resolved: #2779

Previously, the `RemoveFixed.transform_observation_features` transform checked that the parameter being removed had the same value as the fixed parameter in the search space used to initialize the transform.
This diff removes this check to allow removal of any observation parameter with the same name. This will allow the transform to operate on observations from two similar but non-identical search spaces.

Reviewed By: susanxia1006

Differential Revision: D63322396

fbshipit-source-id: 2648a1cac171a871229101fcc1d44c4200f58d1d

ax/modelbridge/transforms/remove_fixed.py

@@ -48,14 +48,8 @@ def transform_observation_features(
         self, observation_features: list[ObservationFeatures]
     ) -> list[ObservationFeatures]:
         for obsf in observation_features:
-            for p_name, fixed_p in self.fixed_parameters.items():
-                if p_name in obsf.parameters:
-                    if obsf.parameters[p_name] != fixed_p.value:
-                        raise ValueError(
-                            f"Fixed parameter {p_name} with out of design value: "
-                            f"{obsf.parameters[p_name]} passed to `RemoveFixed`."
-                        )
-                    obsf.parameters.pop(p_name)
+            for p_name in self.fixed_parameters:
+                obsf.parameters.pop(p_name, None)
         return observation_features
 
     def _transform_search_space(self, search_space: SearchSpace) -> SearchSpace:

ax/modelbridge/transforms/tests/test_remove_fixed_transform.py

@@ -58,12 +58,15 @@ def test_TransformObservationFeatures(self) -> None:
         observation_features = [
             ObservationFeatures(parameters={"a": 2.2, "b": "b", "c": "a"})
         ]
-        observation_features_invalid = [
+        observation_features_different = [
             ObservationFeatures(parameters={"a": 2.2, "b": "b", "c": "b"})
         ]
-        # Fixed parameter out of design!
-        with self.assertRaises(ValueError):
-            self.t.transform_observation_features(observation_features_invalid)
+        # Fixed parameter is out of design. It will still get removed.
+        t_obs = self.t.transform_observation_features(observation_features)
+        t_obs_different = self.t.transform_observation_features(
+            observation_features_different
+        )
+        self.assertEqual(t_obs, t_obs_different)
 
     def test_TransformSearchSpace(self) -> None:
         ss2 = self.search_space.clone()