update

task/search_strategy_d.py

@@ -71,10 +71,10 @@ for t in range(0, task_days):
     today_str = now.strftime("%Y%m%d")
     yesterday_str = (now + datetime.timedelta(days=-1)).strftime("%Y%m%d")
     one_week_age_str = (now + datetime.timedelta(days=-7)).strftime("%Y%m%d")
-    sql_distinct_device_id = """SELECT partition_date,device_id FROM 
+    sql_distinct_device_id = r"""SELECT partition_date,device_id FROM 
         (
             SELECT user_id,partition_date,
-                    if(size(device_list) > 0, device_list [ 0 ], '') AS device_id
+                    if(size(device_list) > 0, device_list [0], '') AS device_id
               FROM online.ml_user_updates
               WHERE partition_date>='{yesterday_str}' AND partition_date<'{today_str}'
         )t1
@@ -113,10 +113,10 @@ for t in range(0, task_days):
                 ON t1.device_id = t2.device_id
         )t3
         on t1.user_id=t3.user_id group by partition_date,device_id
-""".format(today_str=today_str, yesterday_str=yesterday_str )
+""".format(today_str=today_str, yesterday_str=yesterday_str)
     print(sql_distinct_device_id)
     distinct_device_id_df = spark.sql(sql_distinct_device_id)
-    distinct_device_id_df.show(1)
+    # distinct_device_id_df.show(20)
     sql_res = distinct_device_id_df.collect()
     for res in sql_res:
         print(res)