Bug Fix for the RGBImgPartialObsWrapper view size issue

minigrid/minigrid_env.py

@@ -649,19 +649,32 @@ def gen_obs(self):
 
         return obs
 
-    def get_pov_render(self, tile_size):
+    def get_pov_render(self, tile_size, agent_view_size=None):
         """
         Render an agent's POV observation for visualization
         """
-        grid, vis_mask = self.gen_obs_grid()
-
+        grid, vis_mask = self.gen_obs_grid(agent_view_size)
+        '''
+        Bug Fix for https://github.com/Farama-Foundation/Minigrid/issues/419
+        - Added the agent_view_size field and used it if provided without
+        breaking existing code
+        '''
+
         # Render the whole grid
-        img = grid.render(
-            tile_size,
-            agent_pos=(self.agent_view_size // 2, self.agent_view_size - 1),
-            agent_dir=3,
-            highlight_mask=vis_mask,
-        )
+        if agent_view_size is None:            
+            img = grid.render(
+                tile_size,
+                agent_pos=(self.agent_view_size // 2, self.agent_view_size - 1),
+                agent_dir=3,
+                highlight_mask=vis_mask,
+            )
+        else:
+            img = grid.render(
+                tile_size,
+                agent_pos=(agent_view_size // 2, agent_view_size - 1),
+                agent_dir=3,
+                highlight_mask=vis_mask,
+            )       
 
         return img
 

minigrid/wrappers.py

@@ -364,6 +364,7 @@ def __init__(self, env, tile_size=8):
         self.tile_size = tile_size
 
         obs_shape = env.observation_space.spaces["image"].shape
+
         new_image_space = spaces.Box(
             low=0,
             high=255,
@@ -376,7 +377,17 @@ def __init__(self, env, tile_size=8):
         )
 
     def observation(self, obs):
-        rgb_img_partial = self.get_frame(tile_size=self.tile_size, agent_pov=True)
+        '''
+        Bug Fix for https://github.com/Farama-Foundation/Minigrid/issues/419
+
+        - Bypass self.get_frame() with self.get_pov_render(). The former method 
+        was not necessary in this context. 
+        - Add an "agent_view_size" field to the self.get_pov_render() method
+        '''
+        rgb_img_partial = self.get_pov_render(
+            tile_size=self.tile_size,
+            agent_view_size=self.agent_view_size
+        )
 
         return {**obs, "image": rgb_img_partial}