(v3.3.8) - Observation normalization bug Fix (again), negative values…

… in obs_rms.var (#422)

* Evl Callback: Using argument train_env instyead oh inhereted training environment

* Evl Callback: Fixed mean and var normalization calibration set (now it is applied correctly)

* Normalization Wrapper: Deleted RecordConstructorArgs

* Normalization wrapper: Deleted RecordConstructorArgs inherit

* Normalize Wrapper: Fixed var property bug (returning mean again instead of var)

* Updated Sinergym version from 3.3.7 to 3.3.8

sinergym/utils/callbacks.py

@@ -359,7 +359,7 @@ def _on_step(self) -> bool:
             self._is_success_buffer = []
 
             # We close training env before to start the evaluation
-            self.training_env.close()
+            self.train_env.close()
 
             self._sync_envs()
 
@@ -375,7 +375,7 @@ def _on_step(self) -> bool:
 
             # We close evaluation env and starts training env again
             self.eval_env.close()
-            self.training_env.reset()
+            self.train_env.reset()
 
             if self.log_path is not None:
                 for key, value in episodes_data.items():
@@ -502,5 +502,7 @@ def _sync_envs(self):
                 self.eval_env,
                 NormalizeObservation):
             self.eval_env.get_wrapper_attr('deactivate_update')()
-            self.eval_env.obs_rms = deepcopy(
-                self.train_env.get_wrapper_attr('obs_rms'))
+            self.eval_env.get_wrapper_attr('set_mean')(
+                self.train_env.get_wrapper_attr('mean'))
+            self.eval_env.get_wrapper_attr('set_var')(
+                self.train_env.get_wrapper_attr('var'))

sinergym/utils/wrappers.py

@@ -56,7 +56,7 @@ def step(self, action: Union[int, np.ndarray]) -> Tuple[
         return obs, reward_vector, terminated, truncated, info
 
 
-class NormalizeObservation(gym.Wrapper, gym.utils.RecordConstructorArgs):
+class NormalizeObservation(gym.Wrapper):
 
     logger = Logger().getLogger(name='WRAPPER NormalizeObservation',
                                 level=LOG_WRAPPERS_LEVEL)
@@ -82,10 +82,6 @@ def __init__(self,
         mean = self._check_and_update_metric(mean, 'mean')
         var = self._check_and_update_metric(var, 'var')
 
-        # Save normalization configuration for whole python process
-        gym.utils.RecordConstructorArgs.__init__(
-            self, epsilon=epsilon, mean=mean, var=var)
-
         self.num_envs = 1
         self.is_vector_env = False
         self.automatic_update = automatic_update
@@ -196,7 +192,7 @@ def mean(self) -> Optional[np.float64]:
     def var(self) -> Optional[np.float64]:
         """Returns the variance value of the observations."""
         if hasattr(self, 'obs_rms'):
-            return self.obs_rms.mean
+            return self.obs_rms.var
         else:
             return None
 

sinergym/version.txt

@@ -1 +1 @@
-3.3.7
+3.3.8