[Breaking] Differentiate between velocity-cartpole and position-cartp…

…ole (#26)

* Release v1.0.3

* Rename cartpole and pendulum envs to avoid confusion. Also add velocity cartpole env

examples/play_cartpole.py

@@ -1,7 +1,7 @@
-from popgym.envs.stateless_cartpole import StatelessCartPole
+from popgym.envs.stateless_cartpole import PositionOnlyCartPole
 
 if __name__ == "__main__":
-    game = StatelessCartPole()
+    game = PositionOnlyCartPole()
     done = False
     obs, info = game.reset(return_info=True)
     reward = -float("inf")

examples/play_minesweeper.py

@@ -1,15 +1,18 @@
-from popgym.envs.minesweeper import MineSweeperHard
+from popgym.envs.minesweeper import MineSweeperEasy
 
 if __name__ == "__main__":
-    game = MineSweeperHard()
+    game = MineSweeperEasy()
     done = False
-    obs, info = game.reset(return_info=True)
+    obs, info = game.reset()
     reward = -float("inf")
     game.render()
+    done = False
 
     while not done:
         action = input("input index:").split(",")
         action_int = (int(action[0]), int(action[1]))
-        obs, reward, done, info = game.step(action_int)
-        game.render()
+        obs, reward, truncated, terminated, info = game.step(action_int)
+        done = truncated or terminated
+        print(obs)
+        # game.render()
         print("reward:", reward)

examples/play_pendulum.py

@@ -1,7 +1,7 @@
-from popgym.envs.stateless_pendulum import StatelessPendulum
+from popgym.envs.stateless_pendulum import PositionOnlyPendulum
 
 if __name__ == "__main__":
-    game = StatelessPendulum()
+    game = PositionOnlyPendulum()
     done = False
     obs, info = game.reset(return_info=True)
     reward = -float("inf")

popgym/envs/__init__.py

@@ -49,17 +49,29 @@
     MultiarmedBanditHard,
     MultiarmedBanditMedium,
 )
-from popgym.envs.noisy_stateless_cartpole import (
-    NoisyStatelessCartPole,
-    NoisyStatelessCartPoleEasy,
-    NoisyStatelessCartPoleHard,
-    NoisyStatelessCartPoleMedium,
+from popgym.envs.noisy_position_only_cartpole import (
+    NoisyPositionOnlyCartPole,
+    NoisyPositionOnlyCartPoleEasy,
+    NoisyPositionOnlyCartPoleHard,
+    NoisyPositionOnlyCartPoleMedium,
 )
 from popgym.envs.noisy_stateless_pendulum import (
-    NoisyStatelessPendulum,
-    NoisyStatelessPendulumEasy,
-    NoisyStatelessPendulumHard,
-    NoisyStatelessPendulumMedium,
+    NoisyPositionOnlyPendulum,
+    NoisyPositionOnlyPendulumEasy,
+    NoisyPositionOnlyPendulumHard,
+    NoisyPositionOnlyPendulumMedium,
+)
+from popgym.envs.position_only_cartpole import (
+    PositionOnlyCartPole,
+    PositionOnlyCartPoleEasy,
+    PositionOnlyCartPoleHard,
+    PositionOnlyCartPoleMedium,
+)
+from popgym.envs.position_only_pendulum import (
+    PositionOnlyPendulum,
+    PositionOnlyPendulumEasy,
+    PositionOnlyPendulumHard,
+    PositionOnlyPendulumMedium,
 )
 from popgym.envs.repeat_first import (
     RepeatFirst,
@@ -73,17 +85,11 @@
     RepeatPreviousHard,
     RepeatPreviousMedium,
 )
-from popgym.envs.stateless_cartpole import (
-    StatelessCartPole,
-    StatelessCartPoleEasy,
-    StatelessCartPoleHard,
-    StatelessCartPoleMedium,
-)
-from popgym.envs.stateless_pendulum import (
-    StatelessPendulum,
-    StatelessPendulumEasy,
-    StatelessPendulumHard,
-    StatelessPendulumMedium,
+from popgym.envs.velocity_only_cartpole import (
+    VelocityOnlyCartPole,
+    VelocityOnlyCartPoleEasy,
+    VelocityOnlyCartPoleHard,
+    VelocityOnlyCartPoleMedium,
 )
 
 #
@@ -123,34 +129,46 @@
 # Control envs
 #
 CONTROL: Dict[gym.Env, Dict[str, Any]] = {
-    StatelessCartPole: {"id": "popgym-StatelessCartPole-v0"},
-    StatelessPendulum: {
-        "id": "popgym-StatelessPendulum-v0",
+    PositionOnlyCartPole: {"id": "popgym-PositionOnlyCartPole-v0"},
+    PositionOnlyPendulum: {
+        "id": "popgym-PositionOnlyPendulum-v0",
     },
+    VelocityOnlyCartPole: {
+        "id": "popgym-VelocityOnlyCartpole-v0",
+    }
     # BipedalWalker: {"id": "popgym-BipedalWalker-v0"},
 }
 
 CONTROL_EASY: Dict[gym.Env, Dict[str, Any]] = {
-    StatelessCartPoleEasy: {"id": "popgym-StatelessCartPoleEasy-v0"},
-    StatelessPendulumEasy: {
-        "id": "popgym-StatelessPendulumEasy-v0",
+    PositionOnlyCartPoleEasy: {"id": "popgym-PositionOnlyCartPoleEasy-v0"},
+    PositionOnlyPendulumEasy: {
+        "id": "popgym-PositionOnlyPendulumEasy-v0",
     },
+    VelocityOnlyCartPoleEasy: {
+        "id": "popgym-VelocityOnlyCartpoleEasy-v0",
+    }
     # BipedalWalkerEasy: {"id": "popgym-BipedalWalkerEasy-v0"},
 }
 
 CONTROL_MEDIUM: Dict[gym.Env, Dict[str, Any]] = {
-    StatelessCartPoleMedium: {"id": "popgym-StatelessCartPoleMedium-v0"},
-    StatelessPendulumMedium: {
-        "id": "popgym-StatelessPendulumMedium-v0",
+    PositionOnlyCartPoleMedium: {"id": "popgym-PositionOnlyCartPoleMedium-v0"},
+    PositionOnlyPendulumMedium: {
+        "id": "popgym-PositionOnlyPendulumMedium-v0",
     },
+    VelocityOnlyCartPoleMedium: {
+        "id": "popgym-VelocityOnlyCartpoleMedium-v0",
+    }
     # BipedalWalkerMedium: {"id": "popgym-BipedalWalkerMedium-v0"},
 }
 
 CONTROL_HARD: Dict[gym.Env, Dict[str, Any]] = {
-    StatelessCartPoleHard: {"id": "popgym-StatelessCartPoleHard-v0"},
-    StatelessPendulumHard: {
-        "id": "popgym-StatelessPendulumHard-v0",
+    PositionOnlyCartPoleHard: {"id": "popgym-PositionOnlyCartPoleHard-v0"},
+    PositionOnlyPendulumHard: {
+        "id": "popgym-PositionOnlyPendulumHard-v0",
     },
+    VelocityOnlyCartPoleHard: {
+        "id": "popgym-VelocityOnlyCartpoleHard-v0",
+    }
     # BipedalWalkerHard: {"id": "popgym-BipedalWalkerHard-v0"},
 }
 
@@ -160,36 +178,38 @@
 # Noisy envs
 #
 NOISY: Dict[gym.Env, Dict[str, Any]] = {
-    NoisyStatelessCartPole: {"id": "popgym-NoisyStatelessCartPole-v0"},
-    NoisyStatelessPendulum: {
-        "id": "popgym-NoisyStatelessPendulum-v0",
+    NoisyPositionOnlyCartPole: {"id": "popgym-NoisyPositionOnlyCartPole-v0"},
+    NoisyPositionOnlyPendulum: {
+        "id": "popgym-NoisyPositionOnlyPendulum-v0",
     },
     MultiarmedBandit: {"id": "popgym-MultiarmedBandit-v0"},
     # NonstationaryBandit: {"id": "popgym-NonstationaryBandit-v0"},
 }
 
 NOISY_EASY: Dict[gym.Env, Dict[str, Any]] = {
-    NoisyStatelessCartPoleEasy: {"id": "popgym-NoisyStatelessCartPoleEasy-v0"},
-    NoisyStatelessPendulumEasy: {
-        "id": "popgym-NoisyStatelessPendulumEasy-v0",
+    NoisyPositionOnlyCartPoleEasy: {"id": "popgym-NoisyPositionOnlyCartPoleEasy-v0"},
+    NoisyPositionOnlyPendulumEasy: {
+        "id": "popgym-NoisyPositionOnlyPendulumEasy-v0",
     },
     MultiarmedBanditEasy: {"id": "popgym-MultiarmedBanditEasy-v0"},
     # NonstationaryBanditEasy: {"id": "popgym-NonstationaryBanditEasy-v0"},
 }
 
 NOISY_MEDIUM: Dict[gym.Env, Dict[str, Any]] = {
-    NoisyStatelessCartPoleMedium: {"id": "popgym-NoisyStatelessCartPoleMedium-v0"},
-    NoisyStatelessPendulumMedium: {
-        "id": "popgym-NoisyStatelessPendulumMedium-v0",
+    NoisyPositionOnlyCartPoleMedium: {
+        "id": "popgym-NoisyPositionOnlyCartPoleMedium-v0"
+    },
+    NoisyPositionOnlyPendulumMedium: {
+        "id": "popgym-NoisyPositionOnlyPendulumMedium-v0",
     },
     MultiarmedBanditMedium: {"id": "popgym-MultiarmedBanditMedium-v0"},
     # NonstationaryBanditMedium: {"id": "popgym-NonstationaryBanditMedium-v0"},
 }
 
 NOISY_HARD: Dict[gym.Env, Dict[str, Any]] = {
-    NoisyStatelessCartPoleHard: {"id": "popgym-NoisyStatelessCartPoleHard-v0"},
-    NoisyStatelessPendulumHard: {
-        "id": "popgym-NoisyStatelessPendulumHard-v0",
+    NoisyPositionOnlyCartPoleHard: {"id": "popgym-NoisyPositionOnlyCartPoleHard-v0"},
+    NoisyPositionOnlyPendulumHard: {
+        "id": "popgym-NoisyPositionOnlyPendulumHard-v0",
     },
     MultiarmedBanditHard: {"id": "popgym-MultiarmedBanditHard-v0"},
     # NonstationaryBanditHard: {"id": "popgym-NonstationaryBanditHard-v0"},

popgym/envs/noisy_stateless_cartpole.py → popgym/envs/noisy_position_only_cartpole.py

@@ -4,10 +4,10 @@
 import numpy as np
 from gymnasium.core import ActType, ObsType
 
-from popgym.envs.stateless_cartpole import StatelessCartPole
+from popgym.envs.position_only_cartpole import PositionOnlyCartPole
 
 
-class NoisyStatelessCartPole(StatelessCartPole):
+class NoisyPositionOnlyCartPole(PositionOnlyCartPole):
     def __init__(self, *args, **kwargs):
         noise_sigma = kwargs.pop("noise_sigma", 0.1)
         super().__init__(*args, **kwargs)
@@ -33,16 +33,16 @@ def reset(
         return init_obs, info
 
 
-class NoisyStatelessCartPoleEasy(NoisyStatelessCartPole):
+class NoisyPositionOnlyCartPoleEasy(NoisyPositionOnlyCartPole):
     def __init__(self):
         super().__init__(noise_sigma=0.1)
 
 
-class NoisyStatelessCartPoleMedium(NoisyStatelessCartPole):
+class NoisyPositionOnlyCartPoleMedium(NoisyPositionOnlyCartPole):
     def __init__(self):
         super().__init__(noise_sigma=0.2)
 
 
-class NoisyStatelessCartPoleHard(NoisyStatelessCartPole):
+class NoisyPositionOnlyCartPoleHard(NoisyPositionOnlyCartPole):
     def __init__(self):
         super().__init__(noise_sigma=0.3)

popgym/envs/noisy_stateless_pendulum.py

@@ -4,10 +4,10 @@
 import numpy as np
 from gymnasium.core import ActType, ObsType
 
-from popgym.envs.stateless_pendulum import StatelessPendulum
+from popgym.envs.position_only_pendulum import PositionOnlyPendulum
 
 
-class NoisyStatelessPendulum(StatelessPendulum):
+class NoisyPositionOnlyPendulum(PositionOnlyPendulum):
     def __init__(self, *args, **kwargs):
         noise_sigma = kwargs.pop("noise_sigma", 0.1)
         super().__init__(*args, **kwargs)
@@ -32,16 +32,16 @@ def reset(
         return init_obs, info
 
 
-class NoisyStatelessPendulumEasy(NoisyStatelessPendulum):
+class NoisyPositionOnlyPendulumEasy(NoisyPositionOnlyPendulum):
     def __init__(self):
         super().__init__(noise_sigma=0.1)
 
 
-class NoisyStatelessPendulumMedium(NoisyStatelessPendulum):
+class NoisyPositionOnlyPendulumMedium(NoisyPositionOnlyPendulum):
     def __init__(self):
         super().__init__(noise_sigma=0.2)
 
 
-class NoisyStatelessPendulumHard(NoisyStatelessPendulum):
+class NoisyPositionOnlyPendulumHard(NoisyPositionOnlyPendulum):
     def __init__(self):
         super().__init__(noise_sigma=0.3)

popgym/envs/stateless_cartpole.py → popgym/envs/position_only_cartpole.py

@@ -13,7 +13,7 @@
 from popgym.core.env import POPGymEnv
 
 
-class StatelessCartPole(CartPoleEnv, POPGymEnv):
+class PositionOnlyCartPole(CartPoleEnv, POPGymEnv):
     """Partially observable variant of the CartPole gym environment.
 
     https://github.com/openai/gym/blob/master/gym/envs/classic_control/
@@ -73,15 +73,15 @@ def reset(
         return init_obs, info
 
 
-class StatelessCartPoleEasy(StatelessCartPole):
+class PositionOnlyCartPoleEasy(PositionOnlyCartPole):
     pass
 
 
-class StatelessCartPoleMedium(StatelessCartPole):
+class PositionOnlyCartPoleMedium(PositionOnlyCartPole):
     def __init__(self, *args, **kwargs):
         super().__init__(max_episode_length=400)
 
 
-class StatelessCartPoleHard(StatelessCartPole):
+class PositionOnlyCartPoleHard(PositionOnlyCartPole):
     def __init__(self, *args, **kwargs):
         super().__init__(max_episode_length=600)

popgym/envs/stateless_pendulum.py → popgym/envs/position_only_pendulum.py

@@ -12,7 +12,7 @@
 from popgym.core.env import POPGymEnv
 
 
-class StatelessPendulum(PendulumEnv, POPGymEnv):
+class PositionOnlyPendulum(PendulumEnv, POPGymEnv):
     """Partially observable variant of the Pendulum gym environment.
 
     https://github.com/openai/gym/blob/master/gym/envs/classic_control/
@@ -64,15 +64,15 @@ def reset(
         return init_obs[:-1], info
 
 
-class StatelessPendulumEasy(StatelessPendulum):
+class PositionOnlyPendulumEasy(PositionOnlyPendulum):
     pass
 
 
-class StatelessPendulumMedium(StatelessPendulum):
+class PositionOnlyPendulumMedium(PositionOnlyPendulum):
     def __init__(self):
         super().__init__(max_episode_length=150)
 
 
-class StatelessPendulumHard(StatelessPendulum):
+class PositionOnlyPendulumHard(PositionOnlyPendulum):
     def __init__(self):
         super().__init__(max_episode_length=100)