docstring fix for compute_reward (#2380)

openai · Sep 1, 2021 · ee0a568 · ee0a568
1 parent 7573c57
commit ee0a568
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/gym/core.py b/gym/core.py
@@ -204,7 +204,7 @@ def compute_reward(self, achieved_goal, desired_goal, info):
             goal. Note that the following should always hold true:
 
                 ob, reward, done, info = env.step()
-                assert reward == env.compute_reward(ob['achieved_goal'], ob['goal'], info)
+                assert reward == env.compute_reward(ob['achieved_goal'], ob['desired_goal'], info)
         """
         raise NotImplementedError