Save latest auto rendered frame as cache (Farama-Foundation#1010)

gymnasium/wrappers/rendering.py

@@ -477,6 +477,7 @@ def __init__(self, env: gym.Env[ObsType, ActType], auto_rendering: bool = True):
         self.window = None  # Has to be initialized before asserts, as self.window is used in auto close
         self.clock = None
         self.auto_rendering = auto_rendering
+        self._latest_frame = None
 
         assert (
             self.env.render_mode in self.ACCEPTED_RENDER_MODES
@@ -498,7 +499,7 @@ def step(self, action: ActType) -> tuple[ObsType, SupportsFloat, bool, bool, dic
         """Perform a step in the base environment and render a frame to the screen."""
         result = super().step(action)
         if self.auto_rendering:
-            self._render_frame()
+            self._latest_frame = self._render_frame()
         return result
 
     def reset(
@@ -507,11 +508,15 @@ def reset(
         """Reset the base environment and render a frame to the screen."""
         result = super().reset(seed=seed, options=options)
         if self.auto_rendering:
-            self._render_frame()
+            self._latest_frame = self._render_frame()
         return result
 
     def render(self):
         """This method doesn't do much, actual rendering is usually performed in :meth:`step` and :meth:`reset`."""
+        if self.auto_rendering and self._latest_frame is not None:
+            frame = self._latest_frame
+            self._latest_frame = None
+            return frame
         return self._render_frame()
 
     def _render_frame(self):

gymnasium/wrappers/vector/rendering.py

@@ -65,7 +65,7 @@ def step(
         """Perform a step in the base environment and render a frame to the screen."""
         result = super().step(actions)
         if self.auto_rendering:
-            self._render_frame()
+            self._latest_frame = self._render_frame()
         return result
 
     def reset(
@@ -77,11 +77,15 @@ def reset(
         """Reset the base environment and render a frame to the screen."""
         result = super().reset(seed=seed, options=options)
         if self.auto_rendering:
-            self._render_frame()
+            self._latest_frame = self._render_frame()
         return result
 
     def render(self):
         """This method doesn't do much, actual rendering is usually performed in :meth:`step` and :meth:`reset`."""
+        if self.auto_rendering and self._latest_frame is not None:
+            frame = self._latest_frame
+            self._latest_frame = None
+            return frame
         return self._render_frame()
 
     def _render_frame(self):

tests/wrappers/test_human_rendering.py

@@ -37,20 +37,23 @@ def test_human_rendering():
 @pytest.mark.parametrize("env_id", ["CartPole-v1"])
 @pytest.mark.parametrize("num_envs", [1, 3, 9])
 @pytest.mark.parametrize("screen_size", [None])
-def test_human_rendering_manual(env_id, num_envs, screen_size):
+@pytest.mark.parametrize("auto_rendering", [False, True])
+def test_human_rendering_manual(env_id, num_envs, screen_size, auto_rendering):
     env = HumanRendering(
         gym.make("CartPole-v1", render_mode="rgb_array", disable_env_checker=True),
-        auto_rendering=False,
+        auto_rendering=auto_rendering,
     )
     assert env.render_mode == "human"
+    assert env.auto_rendering == auto_rendering
+
     env.reset()
 
     for _ in range(75):
         _, _, terminated, truncated, _ = env.step(env.action_space.sample())
         if terminated or truncated:
             env.reset()
-    rendering = env.render()
     # output should match mode
+    rendering = env.render()
     assert isinstance(rendering, np.ndarray)
 
     env.close()

tests/wrappers/vector/test_human_rendering.py

@@ -25,14 +25,21 @@ def test_num_envs_screen_size(env_id, num_envs, screen_size):
 
 
 def test_render_modes():
+    num_envs = 3
     envs = HumanRendering(
-        gym.make_vec("CartPole-v1", num_envs=3, render_mode="rgb_array_list")
+        gym.make_vec("CartPole-v1", num_envs=num_envs, render_mode="rgb_array_list")
     )
     assert envs.render_mode == "human"
 
     envs.reset()
     for _ in range(25):
         envs.step(envs.action_space.sample())
+        # output should match mode, list of environment rgb_arrays
+        rendering = envs.render()
+        assert isinstance(rendering, list)
+        assert len(rendering) == num_envs
+        assert isinstance(rendering[0], np.ndarray)
+
     envs.close()
 
     # HumanRenderer on human renderer should not work
@@ -48,19 +55,20 @@ def test_render_modes():
 @pytest.mark.parametrize("env_id", ["CartPole-v1"])
 @pytest.mark.parametrize("num_envs", [1, 3, 9])
 @pytest.mark.parametrize("screen_size", [None])
-def test_human_rendering_manual(env_id, num_envs, screen_size):
+@pytest.mark.parametrize("auto_rendering", [False, True])
+def test_human_rendering_manual(env_id, num_envs, screen_size, auto_rendering):
     envs = gym.make_vec(env_id, num_envs=num_envs, render_mode="rgb_array")
-    envs = HumanRendering(envs, screen_size=screen_size, auto_rendering=False)
+    envs = HumanRendering(envs, screen_size=screen_size, auto_rendering=auto_rendering)
 
     assert envs.render_mode == "human"
-    assert not envs.auto_rendering
+    assert envs.auto_rendering == auto_rendering
 
     envs.reset()
 
     # Test Manual render() call
     envs.step(envs.action_space.sample())
-    rendering = envs.render()
     # output should match mode, list of environment rgb_arrays
+    rendering = envs.render()
     assert isinstance(rendering, list)
     assert len(rendering) == num_envs
     assert isinstance(rendering[0], np.ndarray)