Merge pull request #2 from Farama-Foundation/main

`gymnasium==1.0` support (Farama-Foundation#211)
Kallinteris-Andreas · Feb 17, 2024 · 889ba7a · 889ba7a
2 parents 7638294 + 12a336a
commit 889ba7a
Show file tree

Hide file tree

Showing 25 changed files with 121 additions and 8 deletions.
diff --git a/docs/content/multi-goal_api.md b/docs/content/multi-goal_api.md
@@ -21,6 +21,9 @@ goal, e.g. state derived from the simulation.
 
 ```python
 import gymnasium as gym
+import gymnasium_robotics
+
+gym.register_envs(gymnasium_robotics)
 
 env = gym.make("FetchReach-v2")
 env.reset()

diff --git a/docs/envs/franka_kitchen/index.md b/docs/envs/franka_kitchen/index.md
@@ -16,6 +16,9 @@ The tasks can be selected when the environment is initialized passing a list of
 ```python
 
 import gymnasium as gym
+import gymnasium_robotics
+
+gym.register_envs(gymnasium_robotics)
 
 env = gym.make('FrankaKitchen-v1', tasks_to_complete=['microwave', 'kettle'])
 ```

diff --git a/docs/envs/shadow_dexterous_hand/index.md b/docs/envs/shadow_dexterous_hand/index.md
@@ -22,6 +22,9 @@ These environments are instanceated by adding the following strings to the Hand
 
 ```python
 import gymnasium as gym
+import gymnasium_robotics
+
+gym.register_envs(gymnasium_robotics)
 
 env = gym.make('HandManipulateEgg_BooleanTouchSensors-v1')
 ```

diff --git a/docs/index.md b/docs/index.md
@@ -52,6 +52,10 @@ The creation and interaction with the robotic environments follow the Gymnasium
 ```{code-block} python
 
 import gymnasium as gym
+import gymnasium_robotics
+
+gym.register_envs(gymnasium_robotics)
+
 env = gym.make("FetchPickAndPlace-v2", render_mode="human")
 observation, info = env.reset(seed=42)
 for _ in range(1000):

diff --git a/gymnasium_robotics/__init__.py b/gymnasium_robotics/__init__.py
@@ -7,6 +7,8 @@
 
 __version__ = "1.2.4"
 
+__version__ = "1.2.4"
+
 
 def register_robotics_envs():
     """Register all environment ID's to Gymnasium."""

diff --git a/gymnasium_robotics/envs/adroit_hand/adroit_door.py b/gymnasium_robotics/envs/adroit_hand/adroit_door.py
@@ -163,6 +163,9 @@ class AdroitHandDoorEnv(MujocoEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AdroitHandDoor-v1', max_episode_steps=400)
     ```

diff --git a/gymnasium_robotics/envs/adroit_hand/adroit_hammer.py b/gymnasium_robotics/envs/adroit_hand/adroit_hammer.py
@@ -172,6 +172,9 @@ class AdroitHandHammerEnv(MujocoEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AdroitHandHammer-v1', max_episode_steps=400)
     ```

diff --git a/gymnasium_robotics/envs/adroit_hand/adroit_pen.py b/gymnasium_robotics/envs/adroit_hand/adroit_pen.py
@@ -165,6 +165,9 @@ class AdroitHandPenEnv(MujocoEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AdroitHandPen-v1', max_episode_steps=400)
     ```

diff --git a/gymnasium_robotics/envs/adroit_hand/adroit_relocate.py b/gymnasium_robotics/envs/adroit_hand/adroit_relocate.py
@@ -165,6 +165,9 @@ class AdroitHandRelocateEnv(MujocoEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AdroitHandRelocate-v1', max_episode_steps=400)
     ```

diff --git a/gymnasium_robotics/envs/fetch/pick_and_place.py b/gymnasium_robotics/envs/fetch/pick_and_place.py
@@ -92,6 +92,9 @@ class MujocoFetchPickAndPlaceEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchPickAndPlaceDense-v2')
     ```
@@ -118,6 +121,9 @@ class MujocoFetchPickAndPlaceEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchPickAndPlace-v2', max_episode_steps=100)
     ```

diff --git a/gymnasium_robotics/envs/fetch/push.py b/gymnasium_robotics/envs/fetch/push.py
@@ -120,6 +120,9 @@ class MujocoFetchPushEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchPushDense-v2')
     ```
@@ -146,6 +149,9 @@ class MujocoFetchPushEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchPush-v2', max_episode_steps=100)
     ```

diff --git a/gymnasium_robotics/envs/fetch/reach.py b/gymnasium_robotics/envs/fetch/reach.py
@@ -82,6 +82,9 @@ class MujocoFetchReachEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchReachDense-v2')
     ```
@@ -104,6 +107,9 @@ class MujocoFetchReachEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchReach-v2', max_episode_steps=100)
     ```

diff --git a/gymnasium_robotics/envs/fetch/slide.py b/gymnasium_robotics/envs/fetch/slide.py
@@ -120,6 +120,9 @@ class MujocoFetchSlideEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchSlideDense-v2')
     ```
@@ -145,6 +148,9 @@ class MujocoFetchSlideEnv(MujocoFetchEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('FetchSlide-v2', max_episode_steps=100)
     ```

diff --git a/gymnasium_robotics/envs/franka_kitchen/kitchen_env.py b/gymnasium_robotics/envs/franka_kitchen/kitchen_env.py
@@ -64,6 +64,10 @@ class KitchenEnv(GoalEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
+
     env = gym.make('FrankaKitchen-v1', tasks_to_complete=['microwave', 'kettle'])
     ```
 

diff --git a/gymnasium_robotics/envs/maze/ant_maze.py b/gymnasium_robotics/envs/maze/ant_maze.py
@@ -122,3 +122,11 @@ def render(self):
     def close(self):
         super().close()
         self.ant_env.close()
+
+    @property
+    def model(self):
+        return self.ant_env.model
+
+    @property
+    def data(self):
+        return self.ant_env.data
diff --git a/gymnasium_robotics/envs/maze/ant_maze_v4.py b/gymnasium_robotics/envs/maze/ant_maze_v4.py
@@ -65,6 +65,9 @@ class AntMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     example_map = [[1, 1, 1, 1, 1],
            [1, C, 0, C, 1],
@@ -158,6 +161,9 @@ class AntMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AntMaze_UMaze-v4')
     ```
@@ -192,6 +198,9 @@ class AntMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('AntMaze_UMaze-v4', max_episode_steps=100)
     ```
@@ -319,3 +328,11 @@ def render(self):
     def close(self):
         super().close()
         self.ant_env.close()
+
+    @property
+    def model(self):
+        return self.ant_env.model
+
+    @property
+    def data(self):
+        return self.ant_env.data
diff --git a/gymnasium_robotics/envs/maze/point_maze.py b/gymnasium_robotics/envs/maze/point_maze.py
@@ -196,6 +196,9 @@ class PointMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     example_map = [[1, 1, 1, 1, 1],
            [1, C, 0, C, 1],
@@ -253,6 +256,9 @@ class PointMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('PointMaze_UMazeDense-v3')
     ```
@@ -286,6 +292,9 @@ class PointMazeEnv(MazeEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('PointMaze_UMaze-v3', max_episode_steps=100)
     ```
@@ -415,3 +424,11 @@ def render(self):
     def close(self):
         super().close()
         self.point_env.close()
+
+    @property
+    def model(self):
+        return self.point_env.model
+
+    @property
+    def data(self):
+        return self.point_env.data
diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_block.py b/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_block.py
@@ -166,6 +166,9 @@ class MujocoHandBlockEnv(MujocoManipulateEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulateBlock-v1')
     ```
@@ -193,6 +196,9 @@ class MujocoHandBlockEnv(MujocoManipulateEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulateBlock-v1', max_episode_steps=100)
     ```
@@ -203,7 +209,6 @@ class MujocoHandBlockEnv(MujocoManipulateEnv, EzPickle):
 
     * v1: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v0: the environment depends on `mujoco_py` which is no longer maintained.
-
     """
 
     def __init__(

diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_egg.py b/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_egg.py
@@ -169,6 +169,9 @@ class MujocoHandEggEnv(MujocoManipulateEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulateEgg-v1')
     ```
@@ -196,6 +199,9 @@ class MujocoHandEggEnv(MujocoManipulateEnv, EzPickle):
 
     ```
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulateEgg-v1', max_episode_steps=100)
     ```

diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_pen.py b/gymnasium_robotics/envs/shadow_dexterous_hand/manipulate_pen.py
@@ -168,6 +168,9 @@ class MujocoHandPenEnv(MujocoManipulateEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulatePen-v1')
     ```
@@ -195,6 +198,9 @@ class MujocoHandPenEnv(MujocoManipulateEnv, EzPickle):
 
     ```python
     import gymnasium as gym
+    import gymnasium_robotics
+
+    gym.register_envs(gymnasium_robotics)
 
     env = gym.make('HandManipulatePen-v1', max_episode_steps=100)
     ```
@@ -205,7 +211,6 @@ class MujocoHandPenEnv(MujocoManipulateEnv, EzPickle):
 
     * v1: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v0: the environment depends on `mujoco_py` which is no longer maintained.
-
     """
 
     def __init__(

diff --git a/tests/envs/franka_kitchen/test_kitchen_env.py b/tests/envs/franka_kitchen/test_kitchen_env.py
@@ -47,7 +47,7 @@ def test_task_completion(remove_task_when_completed, terminate_on_tasks_complete
         # Complete a task sequentially for each environment step
         for task in TASKS:
             # Force task to be achieved
-            env.data.qpos[OBS_ELEMENT_INDICES[task]] = OBS_ELEMENT_GOALS[task]
+            env.unwrapped.data.qpos[OBS_ELEMENT_INDICES[task]] = OBS_ELEMENT_GOALS[task]
             _, _, terminated, _, info = env.step(env.action_space.sample())
             completed_tasks.add(task)
 
@@ -91,7 +91,7 @@ def test_task_completion(remove_task_when_completed, terminate_on_tasks_complete
         # Complete a task sequentially for each environment step
         for task in TASKS:
             # Force task to be achieved
-            env.data.qpos[OBS_ELEMENT_INDICES[task]] = OBS_ELEMENT_GOALS[task]
+            env.unwrapped.data.qpos[OBS_ELEMENT_INDICES[task]] = OBS_ELEMENT_GOALS[task]
             completed_tasks.add(task)
 
         _, _, terminated, _, info = env.step(env.action_space.sample())

diff --git a/tests/envs/hand/test_manipulate.py b/tests/envs/hand/test_manipulate.py
@@ -16,7 +16,7 @@ def test_serialize_deserialize(environment_id):
     env1.reset()
     env2 = pickle.loads(pickle.dumps(env1))
 
-    assert env1.target_position == env2.target_position, (
+    assert env1.unwrapped.target_position == env2.unwrapped.target_position, (
         env1.target_position,
         env2.target_position,
     )
diff --git a/tests/envs/hand/test_manipulate_touch_sensors.py b/tests/envs/hand/test_manipulate_touch_sensors.py
@@ -16,7 +16,7 @@ def test_serialize_deserialize(environment_id):
     env1.reset()
     env2 = pickle.loads(pickle.dumps(env1))
 
-    assert env1.target_position == env2.target_position, (
+    assert env1.unwrapped.target_position == env2.unwrapped.target_position, (
         env1.target_position,
         env2.target_position,
     )
diff --git a/tests/envs/hand/test_reach.py b/tests/envs/hand/test_reach.py
@@ -8,7 +8,7 @@ def test_serialize_deserialize():
     env1.reset()
     env2 = pickle.loads(pickle.dumps(env1))
 
-    assert env1.distance_threshold == env2.distance_threshold, (
+    assert env1.unwrapped.distance_threshold == env2.unwrapped.distance_threshold, (
         env1.distance_threshold,
         env2.distance_threshold,
     )