Remove truncated from termination criteria

Author: pockerman

rest_api/gymnasium_envs/classic_control/acrobot_env_api.py

@@ -147,9 +147,12 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation = [float(val) for val in observation]
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
+
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/classic_control/cart_pole_env_api.py

@@ -143,9 +143,12 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation = [float(val) for val in observation]
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
+
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/classic_control/mountain_car_env_api.py

@@ -137,8 +137,12 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation = [float(val) for val in observation]
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
+
+            if info is not None:
+                info['truncated'] = truncated
+
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/classic_control/pendulum_env_api.py

@@ -152,9 +152,12 @@ async def step(action: float = Body(...), cidx: int = Body(...)) -> JSONResponse
             observation = [float(val) for val in observation]
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
+
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/classic_control/v/acrobot_v_env_api.py

@@ -14,7 +14,6 @@
 
 acrobot_v_router = APIRouter(prefix="/gymnasium/acrobot-env/v", tags=["Acrobot Vector env API"])
 
-
 NUM_COPIES = 0
 ENV_NAME = "Acrobot"
 VECTORIZATION_MODE = 'sync'
@@ -24,7 +23,6 @@
     0: None
 }
 
-
 # actions that the environment accepts
 ACTIONS_SPACE = {0: "apply -1 torque to the actuated joint",
                  1: "apply 0 torque to the actuated joint",
@@ -164,7 +162,6 @@ async def reset(seed: int = Body(default=42), cidx: int = Body(...),
 
 @acrobot_v_router.post("/step")
 async def step(action: dict[str, list[int]] = Body(title='actions'), cidx: int = Body(...)) -> JSONResponse:
-
     global NUM_COPIES
 
     actions = action['actions']
@@ -191,17 +188,20 @@ async def step(action: dict[str, list[int]] = Body(title='actions'), cidx: int =
 
             # if we truncate or terminate
             # set the environment step type to finished
-            for i, tr in enumerate(truncates):
-                if tr:
-                    step_types[i] = TimeStepType.LAST
+            for i, tr in enumerate(terminates):
+                # if tr:
+                #     step_types[i] = TimeStepType.LAST
 
                 if terminates[i]:
                     step_types[i] = TimeStepType.LAST
 
+            for i in truncates:
+                infos[i]['truncated'] = truncates[i]
+
             step = TimeStepV(observations=observations_ar,
                              rewards=rewards,
                              step_types=step_types,
-                             infos=[],
+                             infos=infos,
                              discounts=[1.0] * NUM_COPIES)
 
             logger.info(f'Step in environment {ENV_NAME} and index {cidx}')

rest_api/gymnasium_envs/toy_text/black_jack_env_api.py

@@ -133,9 +133,11 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation, reward, terminated, truncated, info = envs[cidx].step(action)
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/toy_text/cliffwalking_env_api.py

@@ -128,8 +128,11 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
                 observation, reward, terminated, truncated, info = envs[cidx].step(action)
 
                 step_type = TimeStepType.MID
-                if terminated or truncated:
+                if terminated:
                     step_type = TimeStepType.LAST
+
+                if info is not None:
+                    info['truncated'] = truncated
                 step = TimeStep(observation=observation,
                                 reward=reward,
                                 step_type=step_type,

rest_api/gymnasium_envs/toy_text/frozenlake_env_api.py

@@ -134,9 +134,12 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation, reward, terminated, truncated, info = envs[cidx].step(action)
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
+
             step = TimeStep(observation=observation,
                             reward=reward,
                             step_type=step_type,

rest_api/gymnasium_envs/toy_text/taxi_env_api.py

@@ -131,9 +131,12 @@ async def step(action: int = Body(...), cidx: int = Body(...)) -> JSONResponse:
             observation, reward, terminated, truncated, info = envs[cidx].step(action)
 
             step_type = TimeStepType.MID
-            if terminated or truncated:
+            if terminated:
                 step_type = TimeStepType.LAST
 
+            if info is not None:
+                info['truncated'] = truncated
+
             action_mask = info['action_mask']
             step = TimeStep(observation=observation,
                             reward=reward,