Merge pull request #2 from kngwyu/troom

TRoom

mujoco_maze/agent_model.py

@@ -17,6 +17,13 @@ class AgentModel(ABC, MujocoEnv, EzPickle):
         MujocoEnv.__init__(self, file_path, frame_skip)
         EzPickle.__init__(self)
 
+    def close(self):
+        if self.viewer is not None and hasattr(self.viewer, "window"):
+            import glfw
+
+            glfw.destroy_window(self.viewer.window)
+        super().close()
+
     @abstractmethod
     def _get_obs(self) -> np.ndarray:
         """Returns the observation from the model.

mujoco_maze/maze_env.py

@@ -32,10 +32,11 @@ class MazeEnv(gym.Env):
         maze_size_scaling: float = 4.0,
         inner_reward_scaling: float = 1.0,
         restitution_coef: float = 0.8,
+        task_kwargs: dict = {},
         *args,
         **kwargs,
     ) -> None:
-        self._task = maze_task(maze_size_scaling)
+        self._task = maze_task(maze_size_scaling, **task_kwargs)
 
         xml_path = os.path.join(MODEL_DIR, model_cls.FILE)
         tree = ET.parse(xml_path)
@@ -452,3 +453,6 @@ class MazeEnv(gym.Env):
         done = self._task.termination(next_obs)
         info["position"] = self.wrapped_env.get_xy()
         return next_obs, inner_reward + outer_reward, done, info
+
+    def close(self):
+        self.wrapped_env.close()

mujoco_maze/maze_task.py

@@ -2,7 +2,7 @@
 """
 
 from abc import ABC, abstractmethod
-from typing import Dict, List, NamedTuple, Type
+from typing import Dict, List, NamedTuple, Tuple, Type
 
 import numpy as np
 
@@ -237,6 +237,44 @@ class SubGoal4Rooms(GoalReward4Rooms):
         ]
 
 
+class GoalRewardTRoom(MazeTask):
+    REWARD_THRESHOLD: float = 0.9
+    MAZE_SIZE_SCALING: Scaling = Scaling(4.0, 4.0)
+
+    def __init__(
+        self,
+        scale: float,
+        goals: List[Tuple[float, float]] = [(2.0, -4.0)],
+    ) -> None:
+        super().__init__(scale)
+        self.goals = []
+        for x, y in goals:
+            self.goals.append(MazeGoal(np.array([x * scale, y * scale])))
+
+    def reward(self, obs: np.ndarray) -> float:
+        for goal in self.goals:
+            if goal.neighbor(obs):
+                return goal.reward_scale
+        return -0.0001
+
+    @staticmethod
+    def create_maze() -> List[List[MazeCell]]:
+        E, B, R = MazeCell.EMPTY, MazeCell.BLOCK, MazeCell.ROBOT
+        return [
+            [B, B, B, B, B, B, B],
+            [B, E, E, E, E, E, B],
+            [B, E, E, B, E, E, B],
+            [B, E, E, B, E, E, B],
+            [B, E, B, B, B, E, B],
+            [B, E, E, R, E, E, B],
+            [B, B, B, B, B, B, B],
+        ]
+
+
+class DistRewardTRoom(GoalRewardTRoom, DistRewardMixIn):
+    pass
+
+
 class TaskRegistry:
     REGISTRY: Dict[str, List[Type[MazeTask]]] = {
         "UMaze": [DistRewardUMaze, GoalRewardUMaze],
@@ -244,6 +282,7 @@ class TaskRegistry:
         "Fall": [DistRewardFall, GoalRewardFall],
         "2Rooms": [DistReward2Rooms, GoalReward2Rooms, SubGoal2Rooms],
         "4Rooms": [DistReward4Rooms, GoalReward4Rooms, SubGoal4Rooms],
+        "TRoom": [DistRewardTRoom, GoalRewardTRoom],
     }
 
     @staticmethod

tests/test_envs.py

@@ -20,3 +20,11 @@ def test_point_maze(maze_id):
         assert env.reset().shape == (7,)
         s, _, _, _ = env.step(env.action_space.sample())
         assert s.shape == (7,)
+
+
+@pytest.mark.parametrize("v", [0, 1])
+def test_maze_args(v):
+    env = gym.make(f"PointTRoom-v{v}", task_kwargs={"goals": [(-2.0, -4.0)]})
+    assert env.reset().shape == (7,)
+    s, _, _, _ = env.step(env.action_space.sample())
+    assert s.shape == (7,)