Update ml_engine/patterns.py

ml_engine/patterns.py

@@ -1,5 +1,5 @@
 # ml_engine/patterns.py
-# (V12.0 - Unified Pattern Engine - XGBoost & Pipeline Merged)
+# (V30.0 - GEM-Architect: Config-Injectable Edition)
 
 import os
 import gc
@@ -25,7 +25,7 @@ except ImportError:
     logger.warning("⚠️ مكتبة 'hurst' غير موجودة. سيتم استخدام القيمة الافتراضية.")
 
 # ==============================================================================
-# 🛠️ INTERNAL HELPER FUNCTIONS (Previously in xgboost_pattern_v2)
+# 🛠️ INTERNAL HELPER FUNCTIONS (Essential for Feature Engineering)
 # ==============================================================================
 
 def _zv(x):
@@ -169,10 +169,26 @@ class ChartPatternAnalyzer:
         """
         self.models_dir = models_dir
         self.models = {} 
+        
+        # ✅ القيم الافتراضية (Placeholder)
+        # سيتم الكتابة عليها بواسطة Processor عند التشغيل
         self.timeframe_weights = {'15m': 0.40, '1h': 0.30, '5m': 0.20, '4h': 0.10, '1d': 0.00}
+        self.thresh_bullish = 0.60
+        self.thresh_bearish = 0.40
+        
         self.supported_timeframes = list(self.timeframe_weights.keys())
         self.initialized = False
 
+    def configure_thresholds(self, weights: dict, bull_thresh: float, bear_thresh: float):
+        """
+        دالة استقبال الإعدادات من المعالج المركزي (Processor Injection).
+        """
+        self.timeframe_weights = weights
+        self.thresh_bullish = bull_thresh
+        self.thresh_bearish = bear_thresh
+        self.supported_timeframes = list(weights.keys())
+        logger.info(f"🔧 [PatternEngine] Config Injected: Bull > {self.thresh_bullish}, Weights set.")
+
     async def initialize(self):
         """تحميل نماذج XGBoost"""
         if self.initialized: return True
@@ -183,6 +199,7 @@ class ChartPatternAnalyzer:
              return False
 
         loaded_count = 0
+        # تحميل النماذج بناءً على الأطر الزمنية المدعومة (التي تم تكوينها)
         for tf in self.supported_timeframes:
             model_path = os.path.join(self.models_dir, f"xgb_{tf}.json")
             if os.path.exists(model_path):
@@ -223,6 +240,7 @@ class ChartPatternAnalyzer:
                         prob_up = model.predict(dtest)[0] 
                         details[tf] = float(prob_up)
                         
+                        # ✅ استخدام الوزن الديناميكي
                         weight = self.timeframe_weights.get(tf, 0.0)
                         if weight > 0:
                             weighted_score_sum += prob_up * weight
@@ -236,9 +254,12 @@ class ChartPatternAnalyzer:
         if total_weight_used > 0:
             final_score = weighted_score_sum / total_weight_used
         
+        # ✅ استخدام العتبات الديناميكية للتصنيف
         pattern_text = "Neutral"
-        if final_score >= 0.60: pattern_text = "Bullish Signal"
-        elif final_score <= 0.40: pattern_text = "Bearish Signal"
+        if final_score >= self.thresh_bullish: 
+            pattern_text = "Bullish Signal"
+        elif final_score <= self.thresh_bearish: 
+            pattern_text = "Bearish Signal"
 
         return {
             'pattern_detected': pattern_text,