Update app.py

app.py

@@ -6,7 +6,7 @@ import torch.optim as optim
 from collections import deque
 import random
 from pathlib import Path
-from typing import Dict, Tuple, Any, List
+from typing import Dict, Tuple, Any, List, Optional
 import plotly.graph_objects as go
 from plotly.subplots import make_subplots
 import yaml
@@ -71,21 +71,13 @@ class AdvancedTradingEnvironment:
         self._initialize_market_data()
         
         # Define action and observation spaces
-        self.action_space = self._create_action_space()
-        self.observation_space = self._create_observation_space()
+        self.action_space = 4  # 0: Hold, 1: Buy, 2: Sell, 3: Close Position
+        self.observation_space = (15,)  # Increased features for better state representation
         
         # Portfolio tracking
         self.portfolio_history = []
         self.action_history = []
 
-    def _create_action_space(self) -> int:
-        """Define available trading actions"""
-        return 4  # 0: Hold, 1: Buy, 2: Sell, 3: Close Position
-
-    def _create_observation_space(self) -> Tuple:
-        """Define observation space dimensions"""
-        return (15,)  # Increased features for better state representation
-
     def _initialize_market_data(self):
         """Initialize synthetic market data based on asset type"""
         n_points = 200  # Longer history for better indicators
@@ -405,7 +397,7 @@ class DQNAgent:
         actions = torch.LongTensor(actions).to(self.device)
         rewards = torch.FloatTensor(rewards).to(self.device)
         next_states = torch.FloatTensor(np.array(next_states)).to(self.device)
-        dones = torch.BoolTensor(dones).to(self.device)
+        dones = torch.BoolTensor(dones).to(self.device)  # Fixed: Use BoolTensor instead of FloatTensor
         
         # Current Q values
         current_q_values = self.q_network(states).gather(1, actions.unsqueeze(1)).squeeze(1)
@@ -413,7 +405,8 @@ class DQNAgent:
         # Next Q values from target network
         with torch.no_grad():
             next_q_values = self.target_network(next_states).max(1)[0]
-            target_q_values = rewards + self.gamma * next_q_values * (~dones)
+            # Fixed: Use proper boolean masking
+            target_q_values = rewards + self.gamma * next_q_values * (~dones).float()
         
         # Compute loss and update
         loss = self.criterion(current_q_values, target_q_values)