merge training system

core/enhanced_reward_calculator.py

@@ -46,6 +46,8 @@ class PredictionRecord:
     confidence: float
     current_price: float
     model_name: str
+    # Optional state vector used for prediction/training (standardized feature/state)
+    state_vector: Optional[list] = None
     
     # Outcome fields (set when outcome is determined)
     actual_price: Optional[float] = None
@@ -144,7 +146,8 @@ class EnhancedRewardCalculator:
                       predicted_direction: int,
                       confidence: float,
                       current_price: float,
-                      model_name: str) -> str:
+                      model_name: str,
+                      state_vector: Optional[list] = None) -> str:
         """
         Add a new prediction to track
         
@@ -169,7 +172,8 @@ class EnhancedRewardCalculator:
                 predicted_direction=predicted_direction,
                 confidence=confidence,
                 current_price=current_price,
-                model_name=model_name
+                model_name=model_name,
+                state_vector=state_vector
             )
             
             # If predicted_return provided, prefer computing implied predicted_price