gym-examples 2.0.24__py3-none-any.whl → 2.0.25__py3-none-any.whl

gym_examples/__init__.py

@@ -6,4 +6,4 @@ register(
      max_episode_steps=50,
 )
 
-__version__ = "2.0.24"
+__version__ = "2.0.25"

gym_examples/envs/wsn_env.py

@@ -118,7 +118,8 @@ class WSNRoutingEnv(gym.Env):
             # Calculate final reward
             # rewards_individual = torch.tensor(rewards[i], dtype=torch.double)                
             # final_reward = net(rewards_individual)
-            final_reward = sum(rewards[i])
+            # final_reward = sum(rewards[i])
+            final_reward = np.mean(rewards[i])
             rewards[i] = final_reward
 
         for i in range(self.n_sensors):
@@ -249,10 +250,12 @@ class WSNRoutingEnv(gym.Env):
     
     def compute_reward_number_of_packets(self, i, action, neighbors_i):
         '''
-        Compute the reward based on the number of packets to transmit
+        Compute the reward based on the number of packets of each sensor in the neighborhood
         '''
-        if len(neighbors_i) == 1:
+        if len(neighbors_i) == 1 or action == self.n_sensors:
             return 1
+        elif self.n_sensors in neighbors_i.keys():
+            return 0
         else:
             total_number_of_packets = np.sum([self.number_of_packets[x] for x in neighbors_i])
             normalized_number_of_packets = self.number_of_packets[action] / total_number_of_packets
@@ -269,8 +272,9 @@ class WSNRoutingEnv(gym.Env):
         reward_dispersion_remaining_energy = self.compute_reward_dispersion_remaining_energy(i, action, neighbors_i, remaining_energy_before)
         reward_number_of_packets = self.compute_reward_number_of_packets(i, action, neighbors_i)
 
-        return [reward_angle, reward_distance, reward_consumption_energy, reward_dispersion_remaining_energy, reward_number_of_packets]
-    
+        # return [reward_angle, reward_distance, reward_consumption_energy, reward_dispersion_remaining_energy, reward_number_of_packets]
+        return [reward_angle, reward_distance, reward_consumption_energy, reward_dispersion_remaining_energy]
+
     def integrate_mobility(self):
         '''
         Integrate the mobility of the sensors after each step

{gym_examples-2.0.24.dist-info → gym_examples-2.0.25.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-examples
-Version: 2.0.24
+Version: 2.0.25
 Summary: A custom environment for multi-agent reinforcement learning focused on WSN routing.
 Home-page: https://github.com/gedji/CODES.git
 Author: Georges Djimefo

gym_examples-2.0.25.dist-info/RECORD

@@ -0,0 +1,7 @@
+gym_examples/__init__.py,sha256=M32_6YSJZauKwn2-uBxRdUz47YkLA2k9Kba-ATqlmjY,193
+gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
+gym_examples/envs/wsn_env.py,sha256=ZyQ0DZ0SUVV8yj2TE0_8pbwPVfuzKzidcsymlQvLpU8,16041
+gym_examples-2.0.25.dist-info/METADATA,sha256=bJhOsljL-MgvqCHj9v74RQ0_6qQCr7hTCsxLOjiebFM,411
+gym_examples-2.0.25.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+gym_examples-2.0.25.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
+gym_examples-2.0.25.dist-info/RECORD,,

gym_examples-2.0.24.dist-info/RECORD

@@ -1,7 +0,0 @@
-gym_examples/__init__.py,sha256=cfrZU0rlt-uBEIIfos7Z0TRaI2NgwTZmM5YVG4ojpTE,193
-gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
-gym_examples/envs/wsn_env.py,sha256=HXDG3mR-GXlkwlPAf8xOvzbAk-Gj_sf5nFpsG-z7SEs,15757
-gym_examples-2.0.24.dist-info/METADATA,sha256=p7X6WjQuMnFSqIr0oO4U_Iexvi3pRG_15678VqCXPl8,411
-gym_examples-2.0.24.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-gym_examples-2.0.24.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
-gym_examples-2.0.24.dist-info/RECORD,,