Improve comments

examples/gym/train_a3c_gym.py

@@ -151,7 +151,8 @@ def make_env(process_idx, test):
         if args.monitor and process_idx == 0:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render and process_idx == 0 and not test:
             misc.env_modifiers.make_rendered(env)

examples/gym/train_acer_gym.py

@@ -95,7 +95,8 @@ def make_env(process_idx, test):
         if args.monitor and process_idx == 0:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render and process_idx == 0 and not test:
             misc.env_modifiers.make_rendered(env)

examples/gym/train_categorical_dqn_gym.py

@@ -84,7 +84,8 @@ def make_env(test):
         if args.monitor:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if ((args.render_eval and test) or
                 (args.render_train and not test)):

examples/gym/train_ddpg_gym.py

@@ -90,7 +90,8 @@ def make_env(test):
         if isinstance(env.action_space, spaces.Box):
             misc.env_modifiers.make_action_filtered(env, clip_action_filter)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render and not test:
             misc.env_modifiers.make_rendered(env)

examples/gym/train_dqn_gym.py

@@ -100,7 +100,8 @@ def make_env(test):
         if isinstance(env.action_space, spaces.Box):
             misc.env_modifiers.make_action_filtered(env, clip_action_filter)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if ((args.render_eval and test) or
                 (args.render_train and not test)):

examples/gym/train_pcl_gym.py

@@ -110,7 +110,8 @@ def make_env(process_idx, test):
         if args.monitor and process_idx == 0:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render and process_idx == 0 and not test:
             misc.env_modifiers.make_rendered(env)

examples/gym/train_ppo_gym.py

@@ -142,7 +142,8 @@ def make_env(test):
         if args.monitor:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render:
             misc.env_modifiers.make_rendered(env)

examples/gym/train_reinforce_gym.py

@@ -72,7 +72,8 @@ def make_env(test):
         if args.monitor:
             env = gym.wrappers.Monitor(env, args.outdir)
         if not test:
-            # Scale rewards observed by an agent to facilitate training
+            # Scale rewards (and thus returns) to a reasonable range so that
+            # training is easier
             env = chainerrl.wrappers.ScaleReward(env, args.reward_scale_factor)
         if args.render and not test:
             misc.env_modifiers.make_rendered(env)