Fix check_env for Sequence observation space (#1690)

* Fix Sequence obs env_checker

* Fix Sequence obs env_checker

* Add test : env_checker for Sequence obs

* Add test : env_checker for Sequence obs

* Cleanup and improve env checker messages

---------

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

docs/misc/changelog.rst

@@ -3,7 +3,7 @@
 Changelog
 ==========
 
-Release 2.2.0a3 (WIP)
+Release 2.2.0a4 (WIP)
 --------------------------
 
 Breaking Changes:
@@ -12,16 +12,7 @@ Breaking Changes:
 
 New Features:
 ^^^^^^^^^^^^^
-
-`SB3-Contrib`_
-^^^^^^^^^^^^^^
-
-`RL Zoo`_
-^^^^^^^^^
-
-`SBX`_
-^^^^^^^^^
-- Added ``DDPG`` and ``TD3``
+- Improved error message of the ``env_checker`` for env wrongly detected as GoalEnv (``compute_reward()`` is defined)
 
 Bug Fixes:
 ^^^^^^^^^^
@@ -33,7 +24,17 @@ Bug Fixes:
 - Fixed replay buffer device after loading in OffPolicyAlgorithm (@PatrickHelm)
 - Fixed ``render_mode`` which was not properly loaded when using ``VecNormalize.load()``
 - Fixed success reward dtype in ``SimpleMultiObsEnv`` (@NixGD)
+- Fixed check_env for Sequence observation space (@corentinlger)
 
+`SB3-Contrib`_
+^^^^^^^^^^^^^^
+
+`RL Zoo`_
+^^^^^^^^^
+
+`SBX`_
+^^^^^^^^^
+- Added ``DDPG`` and ``TD3``
 
 Deprecations:
 ^^^^^^^^^^^^^
@@ -1459,4 +1460,4 @@ And all the contributors:
 @Melanol @qgallouedec @francescoluciano @jlp-ue @burakdmb @timothe-chaumont @honglu2875
 @anand-bala @hughperkins @sidney-tio @AlexPasqua @dominicgkerr @Akhilez @Rocamonde @tobirohrer @ZikangXiong
 @DavyMorgan @luizapozzobon @Bonifatius94 @theSquaredError @harveybellini @DavyMorgan @FieteO @jonasreiher @npit @WeberSamuel @troiganto
-@lutogniew @lbergmann1 @lukashass @BertrandDecoster @pseudo-rnd-thoughts @stefanbschneider @kyle-he @PatrickHelm
+@lutogniew @lbergmann1 @lukashass @BertrandDecoster @pseudo-rnd-thoughts @stefanbschneider @kyle-he @PatrickHelm @corentinlger

stable_baselines3/common/env_checker.py

@@ -80,7 +80,7 @@ def _check_unsupported_spaces(env: gym.Env, observation_space: spaces.Space, act
 
     if isinstance(observation_space, spaces.Tuple):
         warnings.warn(
-            "The observation space is a Tuple,"
+            "The observation space is a Tuple, "
             "this is currently not supported by Stable Baselines3. "
             "However, you can convert it to a Dict observation space "
             "(cf. https://gymnasium.farama.org/api/spaces/composite/#dict). "
@@ -93,6 +93,13 @@ def _check_unsupported_spaces(env: gym.Env, observation_space: spaces.Space, act
             "You can use a wrapper or update your observation space."
         )
 
+    if isinstance(observation_space, spaces.Sequence):
+        warnings.warn(
+            "Sequence observation space is not supported by Stable-Baselines3. "
+            "You can pad your observation to have a fixed size instead.\n"
+            "Note: The checks for returned values are skipped."
+        )
+
     if isinstance(action_space, spaces.Discrete) and action_space.start != 0:
         warnings.warn(
             "Discrete action space with a non-zero start is not supported by Stable-Baselines3. "
@@ -347,9 +354,15 @@ def _check_spaces(env: gym.Env) -> None:
     assert isinstance(env.action_space, spaces.Space), f"The action space must inherit from gymnasium.spaces ({gym_spaces})"
 
     if _is_goal_env(env):
-        assert isinstance(
-            env.observation_space, spaces.Dict
-        ), "Goal conditioned envs (previously gym.GoalEnv) require the observation space to be gymnasium.spaces.Dict"
+        print(
+            "We detected your env to be a GoalEnv because `env.compute_reward()` was defined.\n"
+            "If it's not the case, please rename `env.compute_reward()` to something else to avoid False positives."
+        )
+        assert isinstance(env.observation_space, spaces.Dict), (
+            "Goal conditioned envs (previously gym.GoalEnv) require the observation space to be gymnasium.spaces.Dict.\n"
+            "Note: if your env is not a GoalEnv, please rename `env.compute_reward()` "
+            "to something else to avoid False positive."
+        )
 
 
 # Check render cannot be covered by CI
@@ -440,6 +453,10 @@ def check_env(env: gym.Env, warn: bool = True, skip_render_check: bool = True) -
                 f"Your action space has dtype {action_space.dtype}, we recommend using np.float32 to avoid cast errors."
             )
 
+    # If Sequence observation space, do not check the observation any further
+    if isinstance(observation_space, spaces.Sequence):
+        return
+
     # ============ Check the returned values ===============
     _check_returned_values(env, observation_space, action_space)
 

stable_baselines3/version.txt

@@ -1 +1 @@
-2.2.0a3
+2.2.0a4

tests/test_env_checker.py

@@ -37,6 +37,28 @@ def test_check_env_dict_action():
         check_env(env=test_env, warn=True)
 
 
+class SequenceObservationEnv(gym.Env):
+    metadata = {"render_modes": [], "render_fps": 2}
+
+    def __init__(self, render_mode=None):
+        self.observation_space = spaces.Sequence(spaces.Discrete(8))
+        self.action_space = spaces.Discrete(4)
+
+    def reset(self, seed=None, options=None):
+        super().reset(seed=seed)
+        return self.observation_space.sample(), {}
+
+    def step(self, action):
+        return self.observation_space.sample(), 1.0, False, False, {}
+
+
+def test_check_env_sequence_obs():
+    test_env = SequenceObservationEnv()
+
+    with pytest.warns(Warning, match="Sequence.*not supported"):
+        check_env(env=test_env, warn=True)
+
+
 @pytest.mark.parametrize(
     "obs_tuple",
     [