部分文件id格式不对 fix

tangwang
1 parent 0b73c877
Showing 4 changed files with 20 additions and 0 deletions Show diff stats
offline_tasks/scripts/generate_session.py
offline_tasks/scripts/i2i_content_similar.py
offline_tasks/scripts/interest_aggregation.py
offline_tasks/scripts/tag_category_similar.py
@@ -190,6 +190,10 @@ def main():
         df = pd.read_sql(sql_query, engine)
         logger.info(f"获取到 {len(df)} 条记录")
  
+        # 确保ID为整数类型
+        df['item_id'] = df['item_id'].astype(int)
+        df['user_id'] = df['user_id'].astype(str)
+        
         # Debug: 显示数据详情
         if args.debug:
             log_dataframe_info(logger, df, "用户行为数据", sample_size=10)
@@ -49,6 +49,8 @@ def get_active_items(engine):
     """
  
     df = pd.read_sql(sql_query, engine)
+    # 确保ID为整数类型
+    df['item_id'] = df['item_id'].astype(int)
     return df['item_id'].tolist()
  
  
@@ -303,6 +303,14 @@ def main():
     df = pd.read_sql(sql_query, engine)
     logger.info(f"获取到 {len(df)} 条记录")
  
+    # 确保ID为整数类型
+    df['item_id'] = df['item_id'].astype(int)
+    df['user_id'] = df['user_id'].astype(str)
+    if 'category_id' in df.columns:
+        df['category_id'] = df['category_id'].astype(int)
+    if 'supplier_id' in df.columns:
+        df['supplier_id'] = df['supplier_id'].astype(int)
+    
     # 记录数据信息
     log_dataframe_info(logger, df, "用户行为数据")
  
@@ -58,6 +58,12 @@ if args.debug:
 # 执行 SQL 查询并将结果加载到 pandas DataFrame
 df = pd.read_sql(sql_query, engine)
  
+# 确保ID为整数类型
+if 'category_id' in df.columns:
+    df['category_id'] = df['category_id'].astype(int)
+if 'supplier_id' in df.columns:
+    df['supplier_id'] = df['supplier_id'].astype(int)
+
 if args.debug:
     print(f"[DEBUG] 查询完成，共 {len(df)} 条订单记录")