Create plot_utility_Trainer.py

plot_utility_Trainer.py

@@ -0,0 +1,86 @@
+import matplotlib.pyplot as plt
+import numpy as np
+import os
+import time
+
+def smooth_curve(points, factor=0.9):
+
+    smoothed_points = []
+    if points:
+        smoothed_points.append(points[0])
+        for i in range(1, len(points)):
+            smoothed_points.append(smoothed_points[-1] * factor + points[i] * (1 - factor))
+    return smoothed_points
+
+def plot_rewards(rewards_history, log_interval, save_dir, filename="rewards_plot.png", show_plot=True):
+   
+    os.makedirs(save_dir, exist_ok=True)
+    
+    plt.figure(figsize=(12, 6))
+    episodes = [i * log_interval for i in range(1, len(rewards_history) + 1)]
+    plt.plot(episodes, rewards_history, label='Average Reward')
+    plt.xlabel('Episodes')
+    plt.ylabel('Average Reward')
+    plt.title('PPO Training Progress (Average Reward per Episode)')
+    plt.grid(True)
+    plt.legend()
+    plt.tight_layout()
+    
+    save_path = os.path.join(save_dir, filename)
+    plt.savefig(save_path)
+    print(f"Plot saved to: {os.path.abspath(save_path)}")
+    
+    if show_plot:
+        plt.show()
+
+def init_live_plot(save_dir, filename="live_rewards_plot.png"):
+
+    plt.ion() # Turn on interactive mode
+    fig, ax = plt.subplots(figsize=(12, 6))
+    line, = ax.plot([], [], label='Smoothed Average Reward')
+    ax.set_xlabel('Episodes')
+    ax.set_ylabel('Average Reward')
+    ax.set_title('Live PPO Training Progress')
+    ax.grid(True)
+    ax.legend()
+    plt.tight_layout()
+    
+    ax._save_path_final = os.path.join(save_dir, filename) 
+    
+    return fig, ax, line
+
+def update_live_plot(fig, ax, line, episodes, smoothed_rewards, current_timestep=None, total_timesteps=None):
+    """
+    Updates the live plot with new data.
+    """
+    if not episodes or not smoothed_rewards:
+        return
+
+    line.set_data(episodes, smoothed_rewards)
+    
+    ax.set_xlim(0, max(episodes) * 1.05 if episodes else 1) 
+    
+    min_y = min(smoothed_rewards) * 0.9 if smoothed_rewards else -1
+    max_y = max(smoothed_rewards) * 1.1 if smoothed_rewards else 1
+
+    if abs(max_y - min_y) < 0.1: 
+        min_y -= 0.05
+        max_y += 0.05
+    ax.set_ylim(min_y, max_y)
+
+    if current_timestep is not None and total_timesteps is not None:
+        ax.set_title(f'Live PPO Training Progress (Timestep: {current_timestep:,}/{total_timesteps:,})')
+
+    fig.canvas.draw()
+    fig.canvas.flush_events()
+    time.sleep(0.01)
+
+def save_live_plot_final(fig, ax):
+    
+    plt.ioff()
+    save_path = getattr(ax, '_save_path_final', None)
+    if save_path:
+        plt.savefig(save_path)
+        print(f"Final live plot saved to: {os.path.abspath(save_path)}")
+    plt.close(fig)
+    plt.show()