Gymnasium dep (#814)

.github/ISSUE_TEMPLATE/bug.md

@@ -14,7 +14,7 @@ Save time, add code.
 
 **System Info**
 Describe the characteristic of your environment:
- * Describe how Gym was installed (pip, docker, source, ...)
+ * Describe how Gymnasium was installed (pip, docker, source, ...)
  * Operating system: 
  * Python version:
 

README.md

@@ -58,7 +58,7 @@ In certain environments, it's a valid to assume that agents take their actions a
 
 ## SuperSuit
 
-SuperSuit is a library that includes all commonly used wrappers in RL (frame stacking, observation, normalization, etc.) for PettingZoo and Gym environments with a nice API. We developed it in lieu of wrappers built into PettingZoo. https://github.com/Farama-Foundation/SuperSuit
+SuperSuit is a library that includes all commonly used wrappers in RL (frame stacking, observation, normalization, etc.) for PettingZoo and Gymnasium environments with a nice API. We developed it in lieu of wrappers built into PettingZoo. https://github.com/Farama-Foundation/SuperSuit
 
 ## Environment Versioning
 

docs/api/supersuit_wrappers.md

@@ -38,7 +38,7 @@ Supersuit includes the following wrappers:
 
 * `flatten_v0(env)` flattens observations into a 1D array.
 
-* `frame_skip_v0(env, num_frames)` skips `num_frames` number of frames by reapplying old actions over and over. Observations skipped over are ignored. Rewards skipped over are accumulated. Like Gym Atari's frameskip parameter, `num_frames` can also be a tuple `(min_skip, max_skip)`, which indicates a range of possible skip lengths which are randomly chosen from (in single agent environments only).
+* `frame_skip_v0(env, num_frames)` skips `num_frames` number of frames by reapplying old actions over and over. Observations skipped over are ignored. Rewards skipped over are accumulated. Like Gymnasium Atari's frameskip parameter, `num_frames` can also be a tuple `(min_skip, max_skip)`, which indicates a range of possible skip lengths which are randomly chosen from (in single agent environments only).
 
 * `delay_observations_v0(env, delay)` Delays observation by `delay` frames. Before `delay` frames have been executed, the observation is all zeros. Along with frame_skip, this is the preferred way to implement reaction time for high FPS games.
 
@@ -103,7 +103,7 @@ Supersuit includes the following wrappers:
 [//]: # (The following function performs this conversion.)
 
 [//]: # ()
-[//]: # (* `pettingzoo_env_to_vec_env_v0(env)`: Takes a PettingZoo ParallelEnv with the following assumptions: no agent death or generation, homogeneous action and observation spaces. Returns a gym vector environment where each "environment" in the vector represents one agent. An arbitrary PettingZoo parallel environment can be enforced to have these assumptions by wrapping it with the pad_action_space, pad_observations, and the black_death wrapper). This conversion to a vector environment can be used to train appropriate pettingzoo environments with standard single agent RL methods such as stable baselines's A2C out of box (example below).)
+[//]: # (* `pettingzoo_env_to_vec_env_v0(env)`: Takes a PettingZoo ParallelEnv with the following assumptions: no agent death or generation, homogeneous action and observation spaces. Returns a gymnasium vector environment where each "environment" in the vector represents one agent. An arbitrary PettingZoo parallel environment can be enforced to have these assumptions by wrapping it with the pad_action_space, pad_observations, and the black_death wrapper). This conversion to a vector environment can be used to train appropriate pettingzoo environments with standard single agent RL methods such as stable baselines's A2C out of box (example below).)
 
 [//]: # ()
 [//]: # (You can also use the `concat_vec_envs_v0` functionality to train on several vector environments in parallel, forming a vector which looks like)
@@ -190,7 +190,7 @@ Supersuit includes the following wrappers:
 [//]: # (* `observation_lambda_v0(env, observation_fn, observation_space_fn)` allows you to define arbitrary changes to the via `observation_fn(observation, obs_space) : observation`, and `observation_space_fn(obs_space) : obs_space`. For Box-Box transformations the space transformation will be inferred from `change_observation_fn` if `change_obs_space_fn=None` by passing the `high` and `low` bounds through the `observation_space_fn`. In multi-agent environments only, the lambda functions can optionally accept an `agent` parameter, which lets you know the agent name of the observation/observation space, e.g. `observation_fn(observation, obs_space, agent) : observation`.)
 
 [//]: # ()
-[//]: # (* `reward_lambda_v0(env, change_reward_fn)` allows you to make arbitrary changes to rewards by passing in a `change_reward_fn(reward) : reward` function. For Gym environments this is called every step to transform the returned reward. For AECEnv, this function is used to change each element in the rewards dictionary every step.)
+[//]: # (* `reward_lambda_v0(env, change_reward_fn)` allows you to make arbitrary changes to rewards by passing in a `change_reward_fn(reward) : reward` function. For Gymnasium environments this is called every step to transform the returned reward. For AECEnv, this function is used to change each element in the rewards dictionary every step.)
 
 [//]: # ()
 [//]: # (### Lambda Function Examples)

docs/code_examples/aec_rps.py

@@ -1,8 +1,8 @@
 import functools
 
-import gym
+import gymnasium
 import numpy as np
-from gym.spaces import Discrete
+from gymnasium.spaces import Discrete
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import agent_selector, wrappers
@@ -47,7 +47,7 @@ def env(render_mode=None):
 
 class raw_env(AECEnv):
     """
-    The metadata holds environment constants. From gym, we inherit the "render_modes",
+    The metadata holds environment constants. From gymnasium, we inherit the "render_modes",
     metadata which specifies which modes can be put into the render() method.
     At least human mode should be supported.
     The "name" metadata allows the environment to be pretty printed.
@@ -69,7 +69,7 @@ def __init__(self, render_mode=None):
             zip(self.possible_agents, list(range(len(self.possible_agents))))
         )
 
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         self._action_spaces = {agent: Discrete(3) for agent in self.possible_agents}
         self._observation_spaces = {
             agent: Discrete(4) for agent in self.possible_agents
@@ -80,7 +80,7 @@ def __init__(self, render_mode=None):
     # allows action space seeding to work as expected
     @functools.lru_cache(maxsize=None)
     def observation_space(self, agent):
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         return Discrete(4)
 
     @functools.lru_cache(maxsize=None)
@@ -93,7 +93,7 @@ def render(self):
         up a graphical window, or open up some other display that a human can see and understand.
         """
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

docs/code_examples/parallel_rps.py

@@ -1,7 +1,7 @@
 import functools
 
-import gym
-from gym.spaces import Discrete
+import gymnasium
+from gymnasium.spaces import Discrete
 
 from pettingzoo import ParallelEnv
 from pettingzoo.utils import parallel_to_aec, wrappers
@@ -75,7 +75,7 @@ def __init__(self, render_mode=None):
     # allows action space seeding to work as expected
     @functools.lru_cache(maxsize=None)
     def observation_space(self, agent):
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         return Discrete(4)
 
     @functools.lru_cache(maxsize=None)
@@ -88,7 +88,7 @@ def render(self):
         up a graphical window, or open up some other display that a human can see and understand.
         """
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

docs/content/code-examples/aec_rps.py

@@ -1,7 +1,7 @@
 import functools
 
 import numpy as np
-from gym.spaces import Discrete
+from gymnasium.spaces import Discrete
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import agent_selector, wrappers
@@ -44,7 +44,7 @@ def env():
 
 class raw_env(AECEnv):
     """
-    The metadata holds environment constants. From gym, we inherit the "render_modes",
+    The metadata holds environment constants. From gymnasium, we inherit the "render_modes",
     metadata which specifies which modes can be put into the render() method.
     At least human mode should be supported.
     The "name" metadata allows the environment to be pretty printed.
@@ -66,7 +66,7 @@ def __init__(self):
             zip(self.possible_agents, list(range(len(self.possible_agents))))
         )
 
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         self._action_spaces = {agent: Discrete(3) for agent in self.possible_agents}
         self._observation_spaces = {
             agent: Discrete(4) for agent in self.possible_agents
@@ -76,7 +76,7 @@ def __init__(self):
     # allows action space seeding to work as expected
     @functools.lru_cache(maxsize=None)
     def observation_space(self, agent):
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         return Discrete(4)
 
     @functools.lru_cache(maxsize=None)

docs/content/code-examples/parallel_rps.py

@@ -1,6 +1,6 @@
 import functools
 
-from gym.spaces import Discrete
+from gymnasium.spaces import Discrete
 
 from pettingzoo import ParallelEnv
 from pettingzoo.utils import parallel_to_aec, wrappers
@@ -71,7 +71,7 @@ def __init__(self):
     # allows action space seeding to work as expected
     @functools.lru_cache(maxsize=None)
     def observation_space(self, agent):
-        # Gym spaces are defined and documented here: https://gym.openai.com/docs/#spaces
+        # gymnasium spaces are defined and documented here: https://gymnasium.openai.com/docs/#spaces
         return Discrete(4)
 
     @functools.lru_cache(maxsize=None)

pettingzoo/__init__.py

@@ -13,4 +13,4 @@
 
 os.environ["PYGAME_HIDE_SUPPORT_PROMPT"] = "hide"
 
-__version__ = "1.21.0"
+__version__ = "1.22.0"

pettingzoo/atari/base_atari_env.py

@@ -1,10 +1,10 @@
 from pathlib import Path
 
-import gym
+import gymnasium
 import multi_agent_ale_py
 import numpy as np
-from gym import spaces
-from gym.utils import EzPickle, seeding
+from gymnasium import spaces
+from gymnasium.utils import EzPickle, seeding
 
 from pettingzoo.utils import wrappers
 from pettingzoo.utils.conversions import (  # noqa: F401
@@ -132,7 +132,7 @@ def __init__(
         self.action_mapping = action_mapping
 
         if obs_type == "ram":
-            observation_space = gym.spaces.Box(
+            observation_space = gymnasium.spaces.Box(
                 low=0, high=255, dtype=np.uint8, shape=(128,)
             )
         else:
@@ -153,7 +153,8 @@ def __init__(
         self.possible_agents = self.agents[:]
 
         self.action_spaces = {
-            agent: gym.spaces.Discrete(action_size) for agent in self.possible_agents
+            agent: gymnasium.spaces.Discrete(action_size)
+            for agent in self.possible_agents
         }
         self.observation_spaces = {
             agent: observation_space for agent in self.possible_agents
@@ -240,7 +241,7 @@ def step(self, action_dict):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

pettingzoo/butterfly/cooperative_pong/cooperative_pong.py

@@ -73,10 +73,10 @@
 
 """
 
-import gym
+import gymnasium
 import numpy as np
 import pygame
-from gym.utils import EzPickle, seeding
+from gymnasium.utils import EzPickle, seeding
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import wrappers
@@ -169,19 +169,21 @@ def __init__(
         self.off_screen_penalty = off_screen_penalty
 
         # define action and observation spaces
-        self.action_space = [gym.spaces.Discrete(3) for _ in range(self.num_agents)]
+        self.action_space = [
+            gymnasium.spaces.Discrete(3) for _ in range(self.num_agents)
+        ]
         original_shape = original_obs_shape(
             self.s_width, self.s_height, kernel_window_length=kernel_window_length
         )
         original_color_shape = (original_shape[0], original_shape[1], 3)
         self.observation_space = [
-            gym.spaces.Box(
+            gymnasium.spaces.Box(
                 low=0, high=255, shape=(original_color_shape), dtype=np.uint8
             )
             for _ in range(self.num_agents)
         ]
         # define the global space of the environment or state
-        self.state_space = gym.spaces.Box(
+        self.state_space = gymnasium.spaces.Box(
             low=0, high=255, shape=((self.s_height, self.s_width, 3)), dtype=np.uint8
         )
 
@@ -262,7 +264,7 @@ def enable_render(self):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

pettingzoo/butterfly/knights_archers_zombies/knights_archers_zombies.py

@@ -184,12 +184,12 @@
 import sys
 from itertools import repeat
 
-import gym
+import gymnasium
 import numpy as np
 import pygame
 import pygame.gfxdraw
-from gym.spaces import Box, Discrete
-from gym.utils import EzPickle, seeding
+from gymnasium.spaces import Box, Discrete
+from gymnasium.utils import EzPickle, seeding
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import agent_selector, wrappers
@@ -797,7 +797,7 @@ def draw(self):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

pettingzoo/butterfly/pistonball/pistonball.py

@@ -85,12 +85,12 @@
 
 import math
 
-import gym
+import gymnasium
 import numpy as np
 import pygame
 import pymunk
 import pymunk.pygame_util
-from gym.utils import EzPickle, seeding
+from gymnasium.utils import EzPickle, seeding
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import agent_selector, wrappers
@@ -191,7 +191,7 @@ def __init__(
             zip(
                 self.agents,
                 [
-                    gym.spaces.Box(
+                    gymnasium.spaces.Box(
                         low=0,
                         high=255,
                         shape=(obs_height, self.piston_width * 3, 3),
@@ -206,14 +206,14 @@ def __init__(
             self.action_spaces = dict(
                 zip(
                     self.agents,
-                    [gym.spaces.Box(low=-1, high=1, shape=(1,))] * self.n_pistons,
+                    [gymnasium.spaces.Box(low=-1, high=1, shape=(1,))] * self.n_pistons,
                 )
             )
         else:
             self.action_spaces = dict(
-                zip(self.agents, [gym.spaces.Discrete(3)] * self.n_pistons)
+                zip(self.agents, [gymnasium.spaces.Discrete(3)] * self.n_pistons)
             )
-        self.state_space = gym.spaces.Box(
+        self.state_space = gymnasium.spaces.Box(
             low=0,
             high=255,
             shape=(self.screen_height, self.screen_width, 3),
@@ -598,7 +598,7 @@ def get_local_reward(self, prev_position, curr_position):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return

pettingzoo/classic/chess/chess.py

@@ -89,9 +89,9 @@
 """
 
 import chess
-import gym
+import gymnasium
 import numpy as np
-from gym import spaces
+from gymnasium import spaces
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import wrappers
@@ -246,7 +246,7 @@ def step(self, action):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
         else:

pettingzoo/classic/connect_four/connect_four.py

@@ -64,10 +64,10 @@
 
 import os
 
-import gym
+import gymnasium
 import numpy as np
 import pygame
-from gym import spaces
+from gymnasium import spaces
 
 from pettingzoo import AECEnv
 from pettingzoo.utils import wrappers
@@ -224,7 +224,7 @@ def reset(self, seed=None, return_info=False, options=None):
 
     def render(self):
         if self.render_mode is None:
-            gym.logger.WARN(
+            gymnasium.logger.WARN(
                 "You are calling render method without specifying any render mode."
             )
             return