若有多个数量列的判断

predict/dict_predict.py

@@ -130,6 +130,20 @@ class DicPredict(BasePredictor):
                 for col_diffrate in sort_li_diffrate[1:]:
                     temp_pre_model_res.pop(col_diffrate[0])
 
+        # 若有多个数量列,则进行空置率的比较, 空置率最低的选为目标列, #TODO 后续可能需要改成数量元素的占比率
+        prob_num_cols = [i for i in temp_pre_model_res if temp_pre_model_res[i] == '数量']
+        if len(prob_num_cols) >= 2:
+            li_nullrate = []
+            for prob_num_col in prob_num_cols:
+                counter = 0
+                for item in dic_data[prob_num_col]:
+                    if str(item).strip() == self.PLACEHOLDER:
+                        counter += 1
+                li_nullrate.append((prob_num_col, counter / len(dic_data[prob_num_col])))
+            sort_li_nullrate = sorted(li_nullrate, key=lambda x: x[1])
+            for col_nullrate in sort_li_nullrate[1:]:
+                temp_pre_model_res.pop(col_nullrate[0])
+
         model_id_res = {
             'std_result': temp_pre_model_res,
             'ab_result': ab_result,