evolib 0.2.0b3.dev3__tar.gz → 0.2.0b3.dev4__tar.gz

{evolib-0.2.0b3.dev3/evolib.egg-info → evolib-0.2.0b3.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: evolib
-Version: 0.2.0b3.dev3
+Version: 0.2.0b3.dev4
 Summary: A modular framework for evolutionary strategies and neuroevolution.
 Author-email: EvoLib <evolib@dismail.de>
 License: MIT License

{evolib-0.2.0b3.dev3 → evolib-0.2.0b3.dev4}/evolib/envs/gym_wrapper.py

@@ -31,7 +31,6 @@ class GymEnv:
         self,
         env_name: str,
         max_steps: int = 500,
-        normalize: bool = True,
         **env_kwargs: Any,
     ):
         """
@@ -45,23 +44,10 @@ class GymEnv:
         """
         self.env_name = env_name
         self.max_steps = max_steps
-        self.normalize = normalize
         self.env_kwargs = env_kwargs
         # headless env (no Render)
         self.env = gym.make(env_name, **env_kwargs)
 
-    def normalize_obs(self, obs: Any, clip_value: float = 5.0) -> np.ndarray:
-        """
-        Normalize observations to [-1, 1].
-
-        This helps stabilize evolution by keeping all input features within a consistent
-        numeric range.
-        """
-        obs = np.array(obs, dtype=np.float32)
-        obs = np.clip(obs, -clip_value, clip_value)
-        obs /= clip_value
-        return obs
-
     def evaluate(
         self,
         indiv: Individual,
@@ -84,9 +70,6 @@ class GymEnv:
         for _ in range(episodes):
             obs, _ = self.env.reset()
 
-            if self.normalize:
-                obs = self.normalize_obs(obs)
-
             ep_reward = 0.0
 
             for _ in range(self.max_steps):
@@ -105,12 +88,7 @@ class GymEnv:
                 else:
                     action = np.array(action, dtype=np.float32)
 
-                next_obs, reward, terminated, truncated, _ = self.env.step(action)
-
-                if self.normalize:
-                    next_obs = self.normalize_obs(next_obs)
-
-                obs = next_obs
+                obs, reward, terminated, truncated, _ = self.env.step(action)
 
                 ep_reward += float(reward)
 
@@ -151,9 +129,6 @@ class GymEnv:
         )
         obs, _ = env.reset()
 
-        if self.normalize:
-            obs = self.normalize_obs(obs)
-
         RenderFrame = Union[np.ndarray, list[np.ndarray], None]
         frames: list[np.ndarray] = []
 
@@ -172,12 +147,7 @@ class GymEnv:
             else:
                 action = np.array(action, dtype=np.float32)
 
-            next_obs, reward, terminated, truncated, _ = env.step(action)
-
-            if self.normalize:
-                next_obs = self.normalize_obs(next_obs)
-
-            obs = next_obs
+            obs, reward, terminated, truncated, _ = env.step(action)
 
             frame: RenderFrame = env.render()
             if isinstance(frame, np.ndarray):

{evolib-0.2.0b3.dev3 → evolib-0.2.0b3.dev4/evolib.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: evolib
-Version: 0.2.0b3.dev3
+Version: 0.2.0b3.dev4
 Summary: A modular framework for evolutionary strategies and neuroevolution.
 Author-email: EvoLib <evolib@dismail.de>
 License: MIT License

{evolib-0.2.0b3.dev3 → evolib-0.2.0b3.dev4}/examples/08_gym/05_bipedal_walker.py

@@ -27,9 +27,10 @@ gym_env = GymEnv("BipedalWalker-v3", max_steps=MAX_STEPS)
 
 def eval_walker_fitness(indiv: Individual) -> None:
     """Run a single episode in a fresh environment to avoid Box2D race conditions."""
-    local_env = GymEnv("BipedalWalker-v3", max_steps=MAX_STEPS)
-    fitness = local_env.evaluate(indiv, module="brain", episodes=3)
-    indiv.fitness = -fitness
+    #local_env = GymEnv("BipedalWalker-v3", max_steps=MAX_STEPS)
+    reward = gym_env.evaluate(indiv, module="brain", episodes=1)
+    #print(f"Reward: {reward}")
+    indiv.fitness = -reward
 
 def checkpoint(pop: Population) -> None:                                                 
     save_checkpoint(pop, run_name=RUN_NAME)
@@ -37,7 +38,8 @@ def checkpoint(pop: Population) -> None:
 def on_generation_end(pop: Population) -> None:
     """Visualize the best individual every 10 generations (and at the end)."""
 
-    checkpoint(pop)
+    #checkpoint(pop)
+
     best = pop.best()
     print(best.para["brain"].evo_params.mutation_strength)
 

{evolib-0.2.0b3.dev3 → evolib-0.2.0b3.dev4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "evolib"
-version = "0.2.0b3dev3"
+version = "0.2.0b3dev4"
 description = "A modular framework for evolutionary strategies and neuroevolution."
 authors = [
     { name = "EvoLib", email = "evolib@dismail.de" }