From ce6734e25824bb64a9d27044835667e63022d6fc Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Mon, 25 Nov 2024 23:11:27 -0500
Subject: [PATCH 1/8] update Fetch and Robotic environments to fix reset issue
 introduced in v1.3.0

---
 gymnasium_robotics/__init__.py             | 10 +++++-----
 gymnasium_robotics/envs/fetch/fetch_env.py |  6 ++++++
 gymnasium_robotics/envs/robot_env.py       |  8 ++++++++
 3 files changed, 19 insertions(+), 5 deletions(-)

diff --git a/gymnasium_robotics/__init__.py b/gymnasium_robotics/__init__.py
index feb748a6..c1481ddc 100644
--- a/gymnasium_robotics/__init__.py
+++ b/gymnasium_robotics/__init__.py
@@ -30,7 +30,7 @@ def _merge(a, b):
         )
 
         register(
-            id=f"FetchSlide{suffix}-v3",
+            id=f"FetchSlide{suffix}-v4",
             entry_point="gymnasium_robotics.envs.fetch.slide:MujocoFetchSlideEnv",
             kwargs=kwargs,
             max_episode_steps=50,
@@ -44,7 +44,7 @@ def _merge(a, b):
         )
 
         register(
-            id=f"FetchPickAndPlace{suffix}-v3",
+            id=f"FetchPickAndPlace{suffix}-v4",
             entry_point="gymnasium_robotics.envs.fetch.pick_and_place:MujocoFetchPickAndPlaceEnv",
             kwargs=kwargs,
             max_episode_steps=50,
@@ -58,7 +58,7 @@ def _merge(a, b):
         )
 
         register(
-            id=f"FetchReach{suffix}-v3",
+            id=f"FetchReach{suffix}-v4",
             entry_point="gymnasium_robotics.envs.fetch.reach:MujocoFetchReachEnv",
             kwargs=kwargs,
             max_episode_steps=50,
@@ -72,7 +72,7 @@ def _merge(a, b):
         )
 
         register(
-            id=f"FetchPush{suffix}-v3",
+            id=f"FetchPush{suffix}-v4",
             entry_point="gymnasium_robotics.envs.fetch.push:MujocoFetchPushEnv",
             kwargs=kwargs,
             max_episode_steps=50,
@@ -87,7 +87,7 @@ def _merge(a, b):
         )
 
         register(
-            id=f"HandReach{suffix}-v2",
+            id=f"HandReach{suffix}-v3",
             entry_point="gymnasium_robotics.envs.shadow_dexterous_hand.reach:MujocoHandReachEnv",
             kwargs=kwargs,
             max_episode_steps=50,
diff --git a/gymnasium_robotics/envs/fetch/fetch_env.py b/gymnasium_robotics/envs/fetch/fetch_env.py
index 1e90984c..5b59c2c6 100644
--- a/gymnasium_robotics/envs/fetch/fetch_env.py
+++ b/gymnasium_robotics/envs/fetch/fetch_env.py
@@ -376,6 +376,12 @@ def _reset_sim(self):
         # Reset buffers for joint states, actuators, warm-start, control buffers etc.
         self._mujoco.mj_resetData(self.model, self.data)
 
+        self.data.time = self.initial_time
+        self.data.qpos[:] = np.copy(self.initial_qpos)
+        self.data.qvel[:] = np.copy(self.initial_qvel)
+        if self.model.na != 0:
+            self.data.act[:] = None
+
         # Randomize start position of object.
         if self.has_object:
             object_xpos = self.initial_gripper_xpos[:2]
diff --git a/gymnasium_robotics/envs/robot_env.py b/gymnasium_robotics/envs/robot_env.py
index e482715d..9353873a 100644
--- a/gymnasium_robotics/envs/robot_env.py
+++ b/gymnasium_robotics/envs/robot_env.py
@@ -305,6 +305,14 @@ def _initialize_simulation(self):
     def _reset_sim(self):
         # Reset buffers for joint states, warm-start, control buffers etc.
         mujoco.mj_resetData(self.model, self.data)
+
+        self.data.time = self.initial_time
+        self.data.qpos[:] = np.copy(self.initial_qpos)
+        self.data.qvel[:] = np.copy(self.initial_qvel)
+        if self.model.na != 0:
+            self.data.act[:] = None
+
+        self._mujoco.mj_forward(self.model, self.data)
         return super()._reset_sim()
 
     def render(self):

From 0d0662e5cd54236a2b77a41c5cf26fbda0c38418 Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Sat, 30 Nov 2024 12:55:04 -0500
Subject: [PATCH 2/8] update test_envs

---
 tests/envs/hand/test_reach.py |  2 +-
 tests/test_envs.py            | 46 +++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+), 1 deletion(-)

diff --git a/tests/envs/hand/test_reach.py b/tests/envs/hand/test_reach.py
index 1b91afd9..52080b4d 100644
--- a/tests/envs/hand/test_reach.py
+++ b/tests/envs/hand/test_reach.py
@@ -8,7 +8,7 @@
 
 
 def test_serialize_deserialize():
-    env1 = gym.make("HandReach-v2", distance_threshold=1e-6)
+    env1 = gym.make("HandReach-v3", distance_threshold=1e-6)
     env1.reset()
     env2 = pickle.loads(pickle.dumps(env1))
 
diff --git a/tests/test_envs.py b/tests/test_envs.py
index f880b297..6e966998 100644
--- a/tests/test_envs.py
+++ b/tests/test_envs.py
@@ -1,6 +1,8 @@
 import pickle
 import warnings
 
+import numpy as np
+
 import gymnasium as gym
 import pytest
 from gymnasium.envs.mujoco.utils import check_mujoco_reset_state
@@ -163,3 +165,47 @@ def test_pickle_env(env_spec):
     data_equivalence(env.step(action), pickled_env.step(action))
     env.close()
     pickled_env.close()
+
+_test_robot_env_reset_list = ['Fetch', 'HandReach']
+
+@pytest.mark.parametrize(
+    "spec", 
+    [spec for spec in non_mujoco_py_env_specs if np.any([tar in spec.id for tar in _test_robot_env_reset_list])], 
+    ids=[spec.id for spec in non_mujoco_py_env_specs if np.any([tar in spec.id for tar in _test_robot_env_reset_list])]
+)
+def test_robot_env_reset(spec):
+    """Checks initial state of robotic environment, i.e. Fetch and Shadow Dexterous Hand Reach, 
+    whether their initial states align with the document."""
+
+    def _test_initial_states(env, seed=None):
+        diag_dict = {}
+
+        init_obs = env.reset(seed = seed)
+
+        if isinstance(init_obs[0], dict):
+            diag_dict.update(**init_obs[0])
+        elif isinstance(init_obs[0], np.ndarray):
+            diag_dict.update({'observation': init_obs[0]})
+        diag_dict.update({
+            'qpos': env.unwrapped.data.qpos,
+            'qvel': env.unwrapped.data.qvel,
+            'init_qpos': env.unwrapped.initial_qpos,
+            'init_qvel': env.unwrapped.initial_qvel
+        })
+
+        # exclude object location from environments
+        # if spec.id[:-3] in ['FetchPush', 'FetchPickAndPlace', 'FetchSlide',]:
+        if np.any([tar in spec.id for tar in ['FetchPush', 'FetchPickAndPlace', 'FetchSlide',]]):
+            diag_dict['qpos'] = np.delete(diag_dict['qpos'], np.s_[-7:-5])
+            diag_dict['init_qpos'] = np.delete(diag_dict['init_qpos'], np.s_[-7:-5])
+
+        # testing
+        assert np.allclose(diag_dict['qpos'], diag_dict['init_qpos'])
+        assert np.allclose(diag_dict['qvel'], diag_dict['init_qvel'])
+        return diag_dict
+    
+    cur_env: gym.Env = spec.make()
+
+    _test_initial_states(cur_env, seed=24)
+    _test_initial_states(cur_env, seed=10)
+    return

From 2e6010f8abd74c45a6f6853a19c20024bcd08918 Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Sat, 30 Nov 2024 13:23:13 -0500
Subject: [PATCH 3/8] delete commented unused code

---
 tests/test_envs.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/tests/test_envs.py b/tests/test_envs.py
index 6e966998..e27524ae 100644
--- a/tests/test_envs.py
+++ b/tests/test_envs.py
@@ -194,7 +194,6 @@ def _test_initial_states(env, seed=None):
         })
 
         # exclude object location from environments
-        # if spec.id[:-3] in ['FetchPush', 'FetchPickAndPlace', 'FetchSlide',]:
         if np.any([tar in spec.id for tar in ['FetchPush', 'FetchPickAndPlace', 'FetchSlide',]]):
             diag_dict['qpos'] = np.delete(diag_dict['qpos'], np.s_[-7:-5])
             diag_dict['init_qpos'] = np.delete(diag_dict['init_qpos'], np.s_[-7:-5])

From ed9c7372a2c6bec0ab152004e6ddf3ed2f4925fe Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Wed, 11 Dec 2024 12:45:37 -0500
Subject: [PATCH 4/8] update changelog for Fetch environment and HandReach of
 Shadow Dexterous Hand

---
 gymnasium_robotics/envs/fetch/pick_and_place.py        | 1 +
 gymnasium_robotics/envs/fetch/push.py                  | 2 ++
 gymnasium_robotics/envs/fetch/reach.py                 | 2 ++
 gymnasium_robotics/envs/fetch/slide.py                 | 2 ++
 gymnasium_robotics/envs/shadow_dexterous_hand/reach.py | 1 +
 5 files changed, 8 insertions(+)

diff --git a/gymnasium_robotics/envs/fetch/pick_and_place.py b/gymnasium_robotics/envs/fetch/pick_and_place.py
index 37f9a75b..32124700 100644
--- a/gymnasium_robotics/envs/fetch/pick_and_place.py
+++ b/gymnasium_robotics/envs/fetch/pick_and_place.py
@@ -130,6 +130,7 @@ class MujocoFetchPickAndPlaceEnv(MujocoFetchEnv, EzPickle):
 
     ## Version History
 
+    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/push.py b/gymnasium_robotics/envs/fetch/push.py
index 5420be0b..994cce95 100644
--- a/gymnasium_robotics/envs/fetch/push.py
+++ b/gymnasium_robotics/envs/fetch/push.py
@@ -157,6 +157,8 @@ class MujocoFetchPushEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
+    
+    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/reach.py b/gymnasium_robotics/envs/fetch/reach.py
index cb227f9f..0a778592 100644
--- a/gymnasium_robotics/envs/fetch/reach.py
+++ b/gymnasium_robotics/envs/fetch/reach.py
@@ -115,6 +115,8 @@ class MujocoFetchReachEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
+    
+    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/slide.py b/gymnasium_robotics/envs/fetch/slide.py
index d4898aed..87df28b4 100644
--- a/gymnasium_robotics/envs/fetch/slide.py
+++ b/gymnasium_robotics/envs/fetch/slide.py
@@ -156,6 +156,8 @@ class MujocoFetchSlideEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
+    
+    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
index 19665424..aafaa3c2 100644
--- a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
+++ b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
@@ -387,6 +387,7 @@ class MujocoHandReachEnv(get_base_hand_reanch_env(MujocoHandEnv)):
     ```
 
     ## Version History
+    * v3: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v2: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * 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.

From 35baa79fedf5b7bed6056a5c87e81d0645877ca7 Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Wed, 11 Dec 2024 12:47:10 -0500
Subject: [PATCH 5/8] correct HandReach changelog description

---
 gymnasium_robotics/envs/shadow_dexterous_hand/reach.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
index aafaa3c2..de338213 100644
--- a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
+++ b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
@@ -387,7 +387,7 @@ class MujocoHandReachEnv(get_base_hand_reanch_env(MujocoHandEnv)):
     ```
 
     ## Version History
-    * v3: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+    * v3: Fixed bug: the initial state not matching the initial state description in the documentation. Hand Reach environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
     * v2: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * 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.

From 83b4a5f7039a011438df526ba0d46522563ea26d Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Fri, 13 Dec 2024 17:14:22 -0500
Subject: [PATCH 6/8] update changelog description and fix test_envs.py
 formatting

---
 .../envs/fetch/pick_and_place.py              |  2 +-
 gymnasium_robotics/envs/fetch/push.py         |  4 +-
 gymnasium_robotics/envs/fetch/reach.py        |  4 +-
 gymnasium_robotics/envs/fetch/slide.py        |  4 +-
 .../envs/shadow_dexterous_hand/reach.py       |  3 +-
 tests/test_envs.py                            | 64 ++++++++++++-------
 6 files changed, 51 insertions(+), 30 deletions(-)

diff --git a/gymnasium_robotics/envs/fetch/pick_and_place.py b/gymnasium_robotics/envs/fetch/pick_and_place.py
index 32124700..187e3451 100644
--- a/gymnasium_robotics/envs/fetch/pick_and_place.py
+++ b/gymnasium_robotics/envs/fetch/pick_and_place.py
@@ -130,7 +130,7 @@ class MujocoFetchPickAndPlaceEnv(MujocoFetchEnv, EzPickle):
 
     ## Version History
 
-    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+    * v4: Fixed bug where initial state did not match initial state description in documentation. Fetch environments' initial states after reset now match the documentation (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251)).
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/push.py b/gymnasium_robotics/envs/fetch/push.py
index 994cce95..8ef1a0d8 100644
--- a/gymnasium_robotics/envs/fetch/push.py
+++ b/gymnasium_robotics/envs/fetch/push.py
@@ -157,8 +157,8 @@ class MujocoFetchPushEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
-    
-    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+
+    * v4: Fixed bug where initial state did not match initial state description in documentation. Fetch environments' initial states after reset now match the documentation (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251)).
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/reach.py b/gymnasium_robotics/envs/fetch/reach.py
index 0a778592..7a95146e 100644
--- a/gymnasium_robotics/envs/fetch/reach.py
+++ b/gymnasium_robotics/envs/fetch/reach.py
@@ -115,8 +115,8 @@ class MujocoFetchReachEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
-    
-    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+
+    * v4: Fixed bug where initial state did not match initial state description in documentation. Fetch environments' initial states after reset now match the documentation (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251)).
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/fetch/slide.py b/gymnasium_robotics/envs/fetch/slide.py
index 87df28b4..686b854c 100644
--- a/gymnasium_robotics/envs/fetch/slide.py
+++ b/gymnasium_robotics/envs/fetch/slide.py
@@ -156,8 +156,8 @@ class MujocoFetchSlideEnv(MujocoFetchEnv, EzPickle):
     ```
 
     ## Version History
-    
-    * v4: Fixed bug: the initial state not matching the initial state description in the documentation. Fetch environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+
+    * v4: Fixed bug where initial state did not match initial state description in documentation. Fetch environments' initial states after reset now match the documentation (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251)).
     * v3: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * v2: the environment depends on the newest [mujoco python bindings](https://mujoco.readthedocs.io/en/latest/python.html) maintained by the MuJoCo team in Deepmind.
     * v1: the environment depends on `mujoco_py` which is no longer maintained.
diff --git a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
index de338213..8d99d5b6 100644
--- a/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
+++ b/gymnasium_robotics/envs/shadow_dexterous_hand/reach.py
@@ -387,7 +387,8 @@ class MujocoHandReachEnv(get_base_hand_reanch_env(MujocoHandEnv)):
     ```
 
     ## Version History
-    * v3: Fixed bug: the initial state not matching the initial state description in the documentation. Hand Reach environments' initial states after reset now align with the documentation.(related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251))
+
+    * v3: Fixed bug where initial state did not match initial state description in documentation. Hand Reach environments' initial states after reset now match the documentation (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/251)).
     * v2: Fixed bug: `env.reset()` not properly resetting the internal state. Fetch environments now properly reset their state (related [GitHub issue](https://github.com/Farama-Foundation/Gymnasium-Robotics/issues/207)).
     * 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.
diff --git a/tests/test_envs.py b/tests/test_envs.py
index e27524ae..ccc0bf32 100644
--- a/tests/test_envs.py
+++ b/tests/test_envs.py
@@ -1,9 +1,8 @@
 import pickle
 import warnings
 
-import numpy as np
-
 import gymnasium as gym
+import numpy as np
 import pytest
 from gymnasium.envs.mujoco.utils import check_mujoco_reset_state
 from gymnasium.envs.registration import EnvSpec
@@ -166,43 +165,64 @@ def test_pickle_env(env_spec):
     env.close()
     pickled_env.close()
 
-_test_robot_env_reset_list = ['Fetch', 'HandReach']
+
+_test_robot_env_reset_list = ["Fetch", "HandReach"]
+
 
 @pytest.mark.parametrize(
-    "spec", 
-    [spec for spec in non_mujoco_py_env_specs if np.any([tar in spec.id for tar in _test_robot_env_reset_list])], 
-    ids=[spec.id for spec in non_mujoco_py_env_specs if np.any([tar in spec.id for tar in _test_robot_env_reset_list])]
+    "spec",
+    [
+        spec
+        for spec in non_mujoco_py_env_specs
+        if np.any([tar in spec.id for tar in _test_robot_env_reset_list])
+    ],
+    ids=[
+        spec.id
+        for spec in non_mujoco_py_env_specs
+        if np.any([tar in spec.id for tar in _test_robot_env_reset_list])
+    ],
 )
 def test_robot_env_reset(spec):
-    """Checks initial state of robotic environment, i.e. Fetch and Shadow Dexterous Hand Reach, 
-    whether their initial states align with the document."""
+    """Check initial state of robotic environment, i.e. Fetch and Shadow Dexterous Hand Reach,
+    whether their initial states match the description in the documentation."""
 
     def _test_initial_states(env, seed=None):
         diag_dict = {}
 
-        init_obs = env.reset(seed = seed)
+        init_obs = env.reset(seed=seed)
 
         if isinstance(init_obs[0], dict):
             diag_dict.update(**init_obs[0])
         elif isinstance(init_obs[0], np.ndarray):
-            diag_dict.update({'observation': init_obs[0]})
-        diag_dict.update({
-            'qpos': env.unwrapped.data.qpos,
-            'qvel': env.unwrapped.data.qvel,
-            'init_qpos': env.unwrapped.initial_qpos,
-            'init_qvel': env.unwrapped.initial_qvel
-        })
+            diag_dict.update({"observation": init_obs[0]})
+        diag_dict.update(
+            {
+                "qpos": env.unwrapped.data.qpos,
+                "qvel": env.unwrapped.data.qvel,
+                "init_qpos": env.unwrapped.initial_qpos,
+                "init_qvel": env.unwrapped.initial_qvel,
+            }
+        )
 
         # exclude object location from environments
-        if np.any([tar in spec.id for tar in ['FetchPush', 'FetchPickAndPlace', 'FetchSlide',]]):
-            diag_dict['qpos'] = np.delete(diag_dict['qpos'], np.s_[-7:-5])
-            diag_dict['init_qpos'] = np.delete(diag_dict['init_qpos'], np.s_[-7:-5])
+        if np.any(
+            [
+                tar in spec.id
+                for tar in [
+                    "FetchPush",
+                    "FetchPickAndPlace",
+                    "FetchSlide",
+                ]
+            ]
+        ):
+            diag_dict["qpos"] = np.delete(diag_dict["qpos"], np.s_[-7:-5])
+            diag_dict["init_qpos"] = np.delete(diag_dict["init_qpos"], np.s_[-7:-5])
 
         # testing
-        assert np.allclose(diag_dict['qpos'], diag_dict['init_qpos'])
-        assert np.allclose(diag_dict['qvel'], diag_dict['init_qvel'])
+        assert np.allclose(diag_dict["qpos"], diag_dict["init_qpos"])
+        assert np.allclose(diag_dict["qvel"], diag_dict["init_qvel"])
         return diag_dict
-    
+
     cur_env: gym.Env = spec.make()
 
     _test_initial_states(cur_env, seed=24)

From fba389283ff8f79c1448f28bbd35fc23b0acebf0 Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Sat, 14 Dec 2024 01:20:10 -0500
Subject: [PATCH 7/8] update test_envs.py

---
 tests/test_envs.py | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/tests/test_envs.py b/tests/test_envs.py
index ccc0bf32..64466c03 100644
--- a/tests/test_envs.py
+++ b/tests/test_envs.py
@@ -189,12 +189,8 @@ def test_robot_env_reset(spec):
     def _test_initial_states(env, seed=None):
         diag_dict = {}
 
-        init_obs = env.reset(seed=seed)
+        env.reset(seed=seed)
 
-        if isinstance(init_obs[0], dict):
-            diag_dict.update(**init_obs[0])
-        elif isinstance(init_obs[0], np.ndarray):
-            diag_dict.update({"observation": init_obs[0]})
         diag_dict.update(
             {
                 "qpos": env.unwrapped.data.qpos,
@@ -227,4 +223,3 @@ def _test_initial_states(env, seed=None):
 
     _test_initial_states(cur_env, seed=24)
     _test_initial_states(cur_env, seed=10)
-    return

From 7a3b71af7c346f5575f11344992450a10ef54129 Mon Sep 17 00:00:00 2001
From: wjxgeorge <12130815d@connect.polyu.hk>
Date: Tue, 17 Dec 2024 12:58:03 -0500
Subject: [PATCH 8/8] update assertion to equal instead of allclose

---
 tests/test_envs.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tests/test_envs.py b/tests/test_envs.py
index 64466c03..397aa8a5 100644
--- a/tests/test_envs.py
+++ b/tests/test_envs.py
@@ -215,8 +215,8 @@ def _test_initial_states(env, seed=None):
             diag_dict["init_qpos"] = np.delete(diag_dict["init_qpos"], np.s_[-7:-5])
 
         # testing
-        assert np.allclose(diag_dict["qpos"], diag_dict["init_qpos"])
-        assert np.allclose(diag_dict["qvel"], diag_dict["init_qvel"])
+        assert np.all(diag_dict["qpos"] == diag_dict["init_qpos"])
+        assert np.all(diag_dict["qvel"] == diag_dict["init_qvel"])
         return diag_dict
 
     cur_env: gym.Env = spec.make()