gym-examples 2.0.55__py3-none-any.whl → 2.0.57__py3-none-any.whl

gym_examples/__init__.py

@@ -6,4 +6,4 @@ register(
      max_episode_steps=50,
 )
 
-__version__ = "2.0.55"
+__version__ = "2.0.57"

gym_examples/envs/wsn_env.py

@@ -57,6 +57,7 @@ class WSNRoutingEnv(gym.Env):
         self.episode_count = 0
         self.scale_displacement = 0.01 * (upper_bound - lower_bound) # scale of the random displacement of the sensors
         self.epsilon = 1e-10 # small value to avoid division by zero
+        self.rewards_individual = [0] * self.n_sensors
 
         # Define observation space
         self.observation_space = Tuple(
@@ -67,7 +68,10 @@ class WSNRoutingEnv(gym.Env):
         self.reset()
 
     def reset(self):
-
+        # print individual rewards
+        print("\n=================================================")
+        print(f"Rewards: {[[self.rewards_individual[i], self.network_reward_consumption_energy(), self.network_reward_dispersion_remaining_energy()] for i in range(self.n_sensors)]}")
+        print("==================================================\n")
         # Initialize the position of the sensors randomly
         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)
@@ -80,6 +84,7 @@ 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):
             
@@ -115,12 +120,13 @@ 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]) 
             rewards[i] = final_reward
-        print("\n=================================================")
-        print(f"Rewards: {[[rewards[i], self.network_reward_consumption_energy(), self.network_reward_dispersion_remaining_energy()] for i in range(self.n_sensors)]}")
-        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]
-        print("==================================================\n")
+
+        # 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]
+        self.rewards_individual = [[r, self.network_reward_consumption_energy(), self.network_reward_dispersion_remaining_energy()] for r in self.rewards_individual]
         for i in range(self.n_sensors):
             if (self.remaining_energy[i] <= 0) or (self.number_of_packets[i] <= 0):
                 dones[i] = True
@@ -212,14 +218,14 @@ class WSNRoutingEnv(gym.Env):
             distance = np.linalg.norm(self.sensor_positions[i] - self.sensor_positions[action])
             transmission_energy = self.transmission_energy(self.number_of_packets[i], distance)
             reception_energy = self.reception_energy(self.number_of_packets[i])
-            # total_energy = transmission_energy + reception_energy
-            total_energy = transmission_energy
+            total_energy = transmission_energy + reception_energy
+            # total_energy = transmission_energy
 
         # Normalize the total energy consumption
         max_transmission_energy = self.transmission_energy(self.n_sensors * initial_number_of_packets, self.coverage_radius)
         max_reception_energy = self.reception_energy(self.n_sensors * initial_number_of_packets)
-        # max_total_energy = max_transmission_energy + max_reception_energy
-        max_total_energy = max_transmission_energy
+        max_total_energy = max_transmission_energy + max_reception_energy
+        # max_total_energy = max_transmission_energy
         normalized_total_energy = total_energy / max_total_energy
 
         return np.clip(1 - normalized_total_energy, 0, 1)

{gym_examples-2.0.55.dist-info → gym_examples-2.0.57.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-examples
-Version: 2.0.55
+Version: 2.0.57
 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-2.0.57.dist-info/RECORD

@@ -0,0 +1,7 @@
+gym_examples/__init__.py,sha256=01VDdzN_SbYr_5dKx66uCyvGnM5ZLNPARRNolC5y5io,193
+gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
+gym_examples/envs/wsn_env.py,sha256=GWO7toTzhMb8_knk9IqylWYmKdMGIu-_vjUDaOy93wE,15946
+gym_examples-2.0.57.dist-info/METADATA,sha256=xx2bmn5KT-dwZioEANxTHjglopFVesbUevoM6RPHzKk,411
+gym_examples-2.0.57.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+gym_examples-2.0.57.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
+gym_examples-2.0.57.dist-info/RECORD,,

gym_examples-2.0.55.dist-info/RECORD

@@ -1,7 +0,0 @@
-gym_examples/__init__.py,sha256=vnWhhwCczJ-SFYD9e9tkVw0v8gya25yTvJHPjBeSm-U,193
-gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
-gym_examples/envs/wsn_env.py,sha256=aoL7il1bOsPoyBp7HchTRRpumVMIiBAD7kDD53jIck0,15411
-gym_examples-2.0.55.dist-info/METADATA,sha256=BAqN3wGi55dKSDpBVPjBbSXbCMccK7M8jIO2Ry1AnXY,411
-gym_examples-2.0.55.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-gym_examples-2.0.55.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
-gym_examples-2.0.55.dist-info/RECORD,,