gym-examples 2.0.151__py3-none-any.whl → 2.0.152__py3-none-any.whl

gym_examples/__init__.py

@@ -6,4 +6,4 @@ register(
      max_episode_steps=50,
 )
 
-__version__ = "2.0.151"
+__version__ = "2.0.152"

gym_examples/envs/wsn_env.py

@@ -68,14 +68,6 @@ class WSNRoutingEnv(gym.Env):
         
 
     def reset(self):
-        # print individual rewards
-        
-        # if self.rewards_individual != []:
-        #     print("\n=================================================")
-        #     print(f"Episode: {self.episode_count}")
-        #     print(f"Rewards: {self.rewards_individual}")
-        #     print("=================================================\n")
-        # Initialize remaining energy of each sensor to initial_energy joule
         self.sensor_positions = np.random.rand(self.n_sensors, 2) * (upper_bound - lower_bound) + lower_bound
         self.distance_to_base = np.linalg.norm(self.sensor_positions - base_station_position, axis=1)
         self.remaining_energy = np.ones(self.n_sensors) * initial_energy
@@ -85,7 +77,6 @@ class WSNRoutingEnv(gym.Env):
 
     def step(self, actions):
         rewards = [0] * self.n_sensors
-        # self.rewards_individual = [0] * self.n_sensors
         dones = [False] * self.n_sensors
         for i, action in enumerate(actions):
             
@@ -131,17 +122,11 @@ class WSNRoutingEnv(gym.Env):
             # Calculate final reward
             rewards_individual = torch.tensor(rewards[i], dtype=torch.double)                
             final_reward = net(rewards_individual)
-            # self.rewards_individual[i] = rewards[i]
             # final_reward = np.sum(rewards[i]) 
             # weights = np.ones(self.n_sensors, dtype=int)
             # final_reward = np.sum(reward * weight for reward, weight in zip(rewards[i], weights))
             rewards[i] = final_reward
-        # rewards = [0.5 * r + 0.5 * (1/self.n_sensors) * (self.network_reward_consumption_energy() + self.network_reward_dispersion_remaining_energy()) for r in rewards]
-        # rewards = [0.5 * r + 0.5 * (self.network_reward_consumption_energy() + self.network_reward_dispersion_remaining_energy()) for r in rewards]
-        # Only proceed if network consumption energy is not zero to avoid unnecessary list comprehension
-        # self.rewards_individual = [r for r in self.rewards_individual if ((r != 0) and (r[len(r) -1] < 1))]
-        # rewards = np.mean(rewards)
-        # self.rewards_individual = [{"ind": r, "net_consumption_energy": self.network_reward_consumption_energy(), "net_dispersion_energy": self.network_reward_dispersion_remaining_energy()} for r in self.rewards_individual if ((r != 0) and (self.network_reward_consumption_energy() != 0))]
+        rewards = np.mean(rewards)
         for i in range(self.n_sensors):
             if (self.remaining_energy[i] <= 0) or (self.number_of_packets[i] <= 0):
                 dones[i] = True
@@ -242,16 +227,6 @@ class WSNRoutingEnv(gym.Env):
 
         return np.clip(1 - normalized_total_energy, 0, 1)
     
-    # def compute_reward_dispersion_remaining_energy(self,i):
-    #     '''
-    #     Compute the reward based on the difference between the remaining energy of the sensor i and the mean remaining energy of all sensors
-    #     '''
-    #     difference = np.abs(self.remaining_energy[i] - np.mean(self.remaining_energy))
-    #     # Normalize the difference
-    #     normalized_difference = difference / initial_energy
-
-    #     return np.clip(1 - normalized_difference, 0, 1)
-
     def compute_reward_dispersion_remaining_energy(self):
         '''
         Compute the reward based on the standard deviation of the remaining energy
@@ -286,10 +261,7 @@ class WSNRoutingEnv(gym.Env):
         reward_number_of_packets = self.compute_reward_number_of_packets(action)
 
         return [reward_angle, reward_consumption_energy, reward_dispersion_remaining_energy, reward_number_of_packets]
-        # return [reward_angle, reward_distance, reward_consumption_energy, reward_number_of_packets]
-        # return [reward_angle, reward_distance, reward_dispersion_remaining_energy, reward_number_of_packets]
-        # return [reward_angle, reward_distance, reward_consumption_energy, reward_dispersion_remaining_energy]
-
+        
     def network_reward_dispersion_remaining_energy(self):
         '''
         Compute the reward based on the standard deviation of the remaining energy at the network level

{gym_examples-2.0.151.dist-info → gym_examples-2.0.152.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-examples
-Version: 2.0.151
+Version: 2.0.152
 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-2.0.152.dist-info/RECORD

@@ -0,0 +1,7 @@
+gym_examples/__init__.py,sha256=KwIBREa7Yw-YuBVfuQuM2O8nQxiVChcpRIv4Zmh3Sbo,194
+gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
+gym_examples/envs/wsn_env.py,sha256=SjIZcBWTm6Pfc2_omEPqvCqdhWDK57OdzgNhGUekiVY,15185
+gym_examples-2.0.152.dist-info/METADATA,sha256=q951TYn4MDujO4Qv9RGsbgaXGR65k_0mFMEstR-jIik,412
+gym_examples-2.0.152.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+gym_examples-2.0.152.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
+gym_examples-2.0.152.dist-info/RECORD,,

gym_examples-2.0.151.dist-info/RECORD

@@ -1,7 +0,0 @@
-gym_examples/__init__.py,sha256=I77PF4BWS2EhkDtblBd1gJeEf5eG2xexbSBTloWiVOo,194
-gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
-gym_examples/envs/wsn_env.py,sha256=q1MTuPzqUkyp5Gl9iP7a412pLyB01DosWeSgpF7PP8g,17380
-gym_examples-2.0.151.dist-info/METADATA,sha256=p5pEMVQeeUi9I9w-N5Cp38F_waraaLuUfy5ALTLCUSY,412
-gym_examples-2.0.151.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-gym_examples-2.0.151.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
-gym_examples-2.0.151.dist-info/RECORD,,