gym-examples 3.0.209__py3-none-any.whl → 3.0.211__py3-none-any.whl

gym_examples/__init__.py

@@ -6,4 +6,4 @@ register(
      max_episode_steps=50,
 )
 
-__version__ = "3.0.209"
+__version__ = "3.0.211"

gym_examples/envs/wsn_env.py

@@ -130,7 +130,7 @@ class WSNRoutingEnv(gym.Env):
     def step(self, actions):
         actions = [actions[i] for i in range(self.n_agents)] # We want to go back from the MultiDiscrete action space to a tuple of tuple of Discrete action spaces
         self.steps += 1 
-        rewards = [0] * self.n_sensors
+        rewards = [-100] * self.n_sensors
         dones = [False] * self.n_sensors
         for i, action in enumerate(actions):
             if action not in range(self.n_sensors + 1):
@@ -163,8 +163,8 @@ class WSNRoutingEnv(gym.Env):
                 self.total_latency += self.packet_latency[i] + latency_per_hop
                 self.packet_latency[i] = 0
 
-                rewards[i] = self.compute_individual_rewards(i, action)
-                # rewards[i] = np.ones(input_dim)  # Reward for transmitting data to the base station
+                # rewards[i] = self.compute_individual_rewards(i, action)
+                rewards[i] = np.ones(input_dim) * 100 # Reward for transmitting data to the base station
                 dones[i] = True       
             else:
                 distance = np.linalg.norm(self.sensor_positions[i] - self.sensor_positions[action])
@@ -211,6 +211,7 @@ class WSNRoutingEnv(gym.Env):
         rewards = [r.item() if isinstance(r, torch.Tensor) else r for r in rewards] # Convert the rewards to a list of floats
         # rewards = np.sum(rewards)  # Sum the rewards of all agents
         rewards = np.mean(rewards)  # Average the rewards of all agents
+        # rewards = np.mean(self.compute_network_rewards())  # Average the rewards of all agents
         # print(f"Step: {self.steps}, Rewards: {rewards}, Done: {dones}")
         dones = all(dones)  # Done if all agents are done
         
@@ -396,7 +397,8 @@ class WSNRoutingEnv(gym.Env):
         reward_packet_delivery_ratio = self.compute_reward_packet_delivery_ratio()        
         rewards_performance = np.array([reward_latency, reward_network_throughput, reward_packet_delivery_ratio])
 
-        return np.concatenate((rewards_energy, rewards_performance))
+        # return np.concatenate((rewards_energy, rewards_performance))
+        return rewards_energy
 
     def network_reward_dispersion_remaining_energy(self):
         '''

{gym_examples-3.0.209.dist-info → gym_examples-3.0.211.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gym-examples
-Version: 3.0.209
+Version: 3.0.211
 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-3.0.211.dist-info/RECORD

@@ -0,0 +1,7 @@
+gym_examples/__init__.py,sha256=sHByS6d2-26uT2bg-8T_Fr7OAllomtmEs8lwcIUGJKs,194
+gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
+gym_examples/envs/wsn_env.py,sha256=FvQq-Y7S-yOKE1p0fU4bubGP0eWUiciMtsc4kUHmjoo,26057
+gym_examples-3.0.211.dist-info/METADATA,sha256=FOUYKkcwfztVdHqK-bvY_oRf1BWzMy945Ih8fhnQ2kQ,412
+gym_examples-3.0.211.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
+gym_examples-3.0.211.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
+gym_examples-3.0.211.dist-info/RECORD,,

gym_examples-3.0.209.dist-info/RECORD

@@ -1,7 +0,0 @@
-gym_examples/__init__.py,sha256=W8b-NCyzEIs31EMwPSeDN61nvcQN2IFTagqv0stxTZE,194
-gym_examples/envs/__init__.py,sha256=lgMe4pyOuUTgTBUddM0iwMlETsYTwFShny6ifm8PGM8,53
-gym_examples/envs/wsn_env.py,sha256=PFcrAnbBjdhC11_qfnSyAcyXe2IlwdW37Ru0nNXwIYI,25918
-gym_examples-3.0.209.dist-info/METADATA,sha256=OP2QbV15anG0gxnQEVqdbv01H0JX2d242hp6TsiprAE,412
-gym_examples-3.0.209.dist-info/WHEEL,sha256=2wepM1nk4DS4eFpYrW1TTqPcoGNfHhhO_i5m4cOimbo,92
-gym_examples-3.0.209.dist-info/top_level.txt,sha256=rJRksoAF32M6lTLBEwYzRdo4PgtejceaNnnZ3HeY_Rk,13
-gym_examples-3.0.209.dist-info/RECORD,,