Fix returned type in predict (#88)

* actions[0] -> actions.squeeze(0)

* Update changelog

* Update changelog

* Update version

Co-authored-by: Antonin Raffin <antonin.raffin@ensta.org>

docs/misc/changelog.rst

@@ -3,6 +3,22 @@
 Changelog
 ==========
 
+Release 1.6.1a0 (WIP)
+-------------------------------
+
+Breaking Changes:
+^^^^^^^^^^^^^^^^^
+- Fixed the issue that ``predict`` does not always return action as ``np.ndarray`` (@qgallouedec)
+
+New Features:
+^^^^^^^^^^^^^
+
+Bug Fixes:
+^^^^^^^^^^
+
+Deprecations:
+^^^^^^^^^^^^^
+
 Release 1.6.0 (2022-07-11)
 -------------------------------
 
@@ -276,4 +292,4 @@ Stable-Baselines3 is currently maintained by `Antonin Raffin`_ (aka `@araffin`_)
 Contributors:
 -------------
 
-@ku2482 @guyk1971 @minhlong94 @ayeright @kronion @glmcdona @cyprienc @sgillen @Gregwar @rnederstigt
+@ku2482 @guyk1971 @minhlong94 @ayeright @kronion @glmcdona @cyprienc @sgillen @Gregwar @rnederstigt @qgallouedec

sb3_contrib/common/maskable/policies.py

@@ -254,7 +254,7 @@ class MaskableActorCriticPolicy(BasePolicy):
         if not vectorized_env:
             if state is not None:
                 raise ValueError("Error: The environment must be vectorized when using recurrent policies.")
-            actions = actions[0]
+            actions = actions.squeeze(axis=0)
 
         return actions, None
 

sb3_contrib/common/recurrent/policies.py

@@ -416,7 +416,7 @@ class RecurrentActorCriticPolicy(ActorCriticPolicy):
 
         # Remove batch dimension if needed
         if not vectorized_env:
-            actions = actions[0]
+            actions = actions.squeeze(axis=0)
 
         return actions, states
 

sb3_contrib/version.txt

@@ -1 +1 @@
-1.6.0
+1.6.1a0