Merge branch 'hp' into 'master'

update 修改打分逻辑

See merge request root/tfse_rating!28

Rating/scripts/financial_score.py

@@ -25,29 +25,49 @@ def financial_score(param1, param2):
 
     # main
     for key, value in param2.items():
+        # 已获利息倍数为None，分数为满分
         if key == '已获利息倍数' and value is None:
             scores[key] = 5
+        elif value is None:
+            scores[key] = 0
         else:
             refer = list(map(float, refers[key]))
             weight = weights[key]
-            if value is None:
-                score = 0
-            elif value > refer[0]:
+            # 判断打分是正相关还是反相关
+            if refer[0] > refer[1]:
+                # 正相关判断
+                if value >= refer[0]:
                     score = weight
                 elif value > refer[1]:
                     standard_score = round(weight * 1, 2)
-                score = linear_correlation_type(value, standard_score, refer[0], refer[1])
+                    score = linear_correlation_type_positive(value, standard_score, refer[0])
                 elif value > refer[2]:
                     standard_score = round(weight * 0.8, 2)
-                score = linear_correlation_type(value, standard_score, refer[1], refer[2])
+                    score = linear_correlation_type_positive(value, standard_score, refer[1])
                 elif value > refer[3]:
                     standard_score = round(weight * 0.5, 2)
-                score = linear_correlation_type(value, standard_score, refer[2], refer[3])
+                    score = linear_correlation_type_positive(value, standard_score, refer[2])
                 elif value > refer[4]:
                     standard_score = round(weight * 0.25, 2)
-                score = linear_correlation_type(value, standard_score, refer[3], refer[4])
+                    score = linear_correlation_type_positive(value, standard_score, refer[3])
                 else:
                     score = 0
+            else:
+                if value <= refer[0]:
+                    score = weight
+                elif value < refer[1]:
+                    standard_score = round(weight * 1, 2)
+                    score = linear_correlation_type_anti(value, standard_score, refer[0])
+                elif value < refer[2]:
+                    standard_score = round(weight * 0.8, 2)
+                    score = linear_correlation_type_anti(value, standard_score, refer[1])
+                elif value < refer[3]:
+                    standard_score = round(weight * 0.5, 2)
+                    score = linear_correlation_type_anti(value, standard_score, refer[2])
+                elif value < refer[4]:
+                    standard_score = round(weight * 0.25, 2)
+                    score = linear_correlation_type_anti(value, standard_score, refer[3])
+
             scores[key] = round(score, 2)
 
     result = dict()
@@ -92,7 +112,7 @@ def linear_correlation_type(value, standard_score, refer1, refer2):
         elif value >= refer1:
             score = standard_score
         else:
-            score = standard_score - standard_score * (1 - (refer1 - value) / (refer1 - refer2))
+            score = standard_score * (1 - (refer1 - value) / (refer1 - refer2))
     # 反相关打分
     else:
         if value <= refer2:
@@ -100,6 +120,36 @@ def linear_correlation_type(value, standard_score, refer1, refer2):
         elif value >= refer1:
             score = 0
         else:
-            score = standard_score - standard_score * (1 - (value - refer2) / (refer1 - refer2))
+            score = standard_score * (1 - (value - refer2) / (refer1 - refer2))
+
+    return score
+
+
+def linear_correlation_type_positive(value, standard_score, refer1):
+    """
+    线性相关类型 正相关/反相关
+    Parameters
+        value float 指标值
+        standard_score float 标准分
+        refer1 阈值1
+    Returns
+        score float 得分
+    """
+    score = standard_score * (value / refer1)
+
+    return score
+
+
+def linear_correlation_type_anti(value, standard_score, refer1):
+    """
+    线性相关类型 正相关/反相关
+    Parameters
+        value float 指标值
+        standard_score float 标准分
+        refer1 阈值1
+    Returns
+        score float 得分
+    """
+    score = standard_score - (standard_score * (value / refer1) - standard_score)
 
     return score