Update scoring_calculation_system.py

scoring_calculation_system.py

@@ -1469,80 +1469,77 @@ def calculate_final_weighted_score(
     adaptability_bonus: float
 ) -> float:
     """
-    優化的最終分數計算系統
+    優化的最終分數計算系統，強化條件變化的影響力
     """
-    # 1. 基礎分數計算 - 使用更極端的權重
+    # 基礎權重 - 更極端的差異
     base_weights = {
-        'space': 0.35,        # 大幅提高空間權重
-        'exercise': 0.25,
+        'space': 0.35,        # 極度重視空間匹配
+        'exercise': 0.25,     # 重視運動需求
         'grooming': 0.15,
         'experience': 0.15,
         'health': 0.07,
-        'noise': 0.03        # 降低噪音的基礎權重
+        'noise': 0.03
     }
+
+    # 條件特殊化評分
+    special_conditions = 0.0
     
-    # 2. 條件特殊化評分
-    condition_bonus = calculate_condition_bonus(breed_info, user_prefs, scores)
-    
-    # 3. 計算加權基礎分數
+    # 極端條件判定
+    if user_prefs.living_space == 'apartment':
+        if breed_info['Size'] == 'Large':
+            special_conditions -= 0.35  # 大型犬在公寓極度不適合
+        elif breed_info['Size'] == 'Small':
+            special_conditions += 0.15  # 小型犬在公寓加分
+            
+    # 運動需求極端匹配
+    exercise_needs = breed_info.get('Exercise_Needs', 'MODERATE').upper()
+    if user_prefs.exercise_time > 120:  # 高運動量
+        if exercise_needs in ['VERY HIGH', 'HIGH']:
+            special_conditions += 0.20
+        elif exercise_needs == 'LOW':
+            special_conditions -= 0.25
+    elif user_prefs.exercise_time < 45:  # 低運動量
+        if exercise_needs in ['VERY HIGH', 'HIGH']:
+            special_conditions -= 0.25
+        elif exercise_needs == 'LOW':
+            special_conditions += 0.15
+
+    # 經驗等級極端匹配
+    if user_prefs.experience_level == 'beginner':
+        if breed_info.get('Care_Level') == 'HIGH':
+            special_conditions -= 0.30
+    elif user_prefs.experience_level == 'advanced':
+        if breed_info.get('Care_Level') == 'LOW':
+            special_conditions -= 0.20
+
+    # 計算加權基礎分數
     weighted_base = sum(score * base_weights[category] for category, score in scores.items())
     
-    # 4. 品種特性加成
+    # 品種特性加成
     breed_bonus = calculate_breed_bonus(breed_info, user_prefs)
     
-    # 5. 最終分數計算 - 改變權重分配
-    raw_score = (weighted_base * 0.60) + (breed_bonus * 0.25) + (adaptability_bonus * 0.10) + (condition_bonus * 0.05)
-    
-    # 6. 分數轉換 - 使用更激進的轉換函數
-    return amplify_score_extreme(raw_score)
+    # 最終分數計算
+    raw_score = (weighted_base * 0.65) + (breed_bonus * 0.20) + (adaptability_bonus * 0.10) + (special_conditions * 0.05)
 
-def calculate_condition_bonus(breed_info: dict, user_prefs: UserPreferences, scores: dict) -> float:
-    """
-    計算條件特殊化加分，強化極端條件的影響
-    """
-    bonus = 0.0
-    
-    # 居住空間極端匹配
-    if user_prefs.living_space == 'apartment':
-        if breed_info['Size'] == 'Small' and scores['noise'] > 0.8:
-            bonus += 0.25  # 顯著獎勵適合公寓的小型安靜犬種
-        elif breed_info['Size'] in ['Large', 'Giant']:
-            bonus -= 0.35  # 嚴重懲罰大型犬
-    
-    # 美容需求匹配
-    if user_prefs.grooming_commitment == 'low':
-        if breed_info.get('Grooming Needs') == 'HIGH':
-            bonus -= 0.30  # 嚴重懲罰高美容需求
-    
-    # 經驗等級匹配
-    if user_prefs.experience_level == 'beginner':
-        if breed_info.get('Care Level') == 'HIGH':
-            bonus -= 0.25
-    elif user_prefs.experience_level == 'advanced':
-        if breed_info.get('Care Level') == 'LOW':
-            bonus -= 0.20
+    # 分數轉換 - 使用S型曲線
+    return amplify_score_extreme(raw_score)
     
-    return bonus
 
 def amplify_score_extreme(score: float) -> float:
     """
-    更激進的分數轉換函數
+    使用S型曲線進行分數轉換，加大差異
     """
-    # 基礎範圍調整
-    base_min = 0.65  # 提高最低分
-    base_max = 0.98  # 提高最高分
+    # 基礎範圍
+    base_min = 0.65
+    base_max = 0.95
     
-    # 非線性轉換
+    # 正規化
     normalized = (score - 0.5) / 0.5
-    amplified = math.pow(abs(normalized), 1.2) * math.copysign(1, normalized)
     
     # S型曲線轉換
-    sigmoid = 1 / (1 + math.exp(-amplified * 3))
+    sigmoid = 1 / (1 + math.exp(-normalized * 4))
     
     # 映射到目標範圍
     final = base_min + (base_max - base_min) * sigmoid
     
-    # 加入隨機微擾動以打破相似分數
-    noise = random.uniform(-0.002, 0.002)
-    
-    return round(min(base_max, max(base_min, final + noise)), 4)
+    return round(min(base_max, max(base_min, final)), 4)