from gym import spaces from gym_minigrid.minigrid import * from gym_minigrid.register import register class GoToDoorEnv(MiniGridEnv): """ Environment in which the agent is instructed to go to a given object named using an English text string """ def __init__( self, size=5 ): assert size >= 5 super().__init__(gridSize=size, maxSteps=10*size) self.observation_space = spaces.Dict({ 'image': self.observation_space }) self.reward_range = (-1, 1) def _genGrid(self, width, height): # Create the grid grid = Grid(width, height) # Randomly vary the room width and height width = self._randInt(5, width+1) height = self._randInt(5, height+1) # Generate the surrounding walls for i in range(0, width): grid.set(i, 0, Wall()) grid.set(i, height-1, Wall()) for j in range(0, height): grid.set(0, j, Wall()) grid.set(width-1, j, Wall()) # Randomize the player start position and orientation self.startPos = self._randPos( 1, width-1, 1, height-1 ) self.startDir = self._randInt(0, 4) # Generate the 4 doors at random positions doorPos = [] doorPos.append((self._randInt(2, width-2), 0)) doorPos.append((self._randInt(2, width-2), height-1)) doorPos.append((0, self._randInt(2, height-2))) doorPos.append((width-1, self._randInt(2, height-2))) # Generate the door colors doorColors = [] while len(doorColors) < len(doorPos): color = self._randElem(COLOR_NAMES) if color in doorColors: continue doorColors.append(color) # Place the doors in the grid for idx, pos in enumerate(doorPos): color = doorColors[idx] grid.set(*pos, Door(color)) # Select a random target door doorIdx = self._randInt(0, len(doorPos)) self.targetPos = doorPos[doorIdx] self.targetColor = doorColors[doorIdx] # Generate the mission string self.mission = 'go to the %s door' % self.targetColor #print(self.mission) return grid def _observation(self, obs): """ Encode observations """ obs = { 'image': obs, 'mission': self.mission } return obs def _reset(self): obs = MiniGridEnv._reset(self) return self._observation(obs) def _step(self, action): obs, reward, done, info = MiniGridEnv._step(self, action) ax, ay = self.agentPos tx, ty = self.targetPos # Don't let the agent open any of the doors if action == self.actions.toggle: done = True # Reward waiting in front of the target door if action == self.actions.wait: if (ax == tx and abs(ay - ty) == 1) or (ay == ty and abs(ax - tx) == 1): reward = 1 done = True obs = self._observation(obs) return obs, reward, done, info class GoToDoor8x8Env(GoToDoorEnv): def __init__(self): super().__init__(size=8) class GoToDoor6x6Env(GoToDoorEnv): def __init__(self): super().__init__(size=6) register( id='MiniGrid-GoToDoor-5x5-v0', entry_point='gym_minigrid.envs:GoToDoorEnv' ) register( id='MiniGrid-GoToDoor-6x6-v0', entry_point='gym_minigrid.envs:GoToDoor6x6Env' ) register( id='MiniGrid-GoToDoor-8x8-v0', entry_point='gym_minigrid.envs:GoToDoor8x8Env' )