Question

在以下情况中，我无法确定如何平均10列：

我有两张桌子

--------------------
|Name|ClosestSiteID|
|AAA |001          |
|BBB |002          |
--------------------

第二个表

-----------------------
|SiteID|DateTime|Value|
|001   |0       |10   |
|001   |1       |20   |
|001   |2       |20   |
|001   |3       |10   |
|002   |0       |5    |
|002   |1       |15   |
-----------------------

...等每个SiteID有数百个条目

我想在SiteID上编写这两个表的连接，它们平均每个站点ID中的10个条目，以便输出看起来像这样

-------------------------------------
|Name  |ClosestSiteID|DateTime|Value|
|AAA   |001          |0       |AVG  |
|AAA   |001          |10      |AVG  |
|BBB   |002          |0       |AVG  |
|BBB   |002          |10      |AVG  |
-------------------------------------

有没有办法在sql中执行此操作？

Answer 1

这是一种方法。为每个table2添加一个序号到SiteId。然后将此数字除以10并将结果用于聚合目的：

select t1.name, t1.ClosestSiteId, min(DateTime) as DateTime,
       avg(value) as value
from table1 t1 join
     (select t2.*, row_number() over (partition by SiteId order by date) as seqnum
      from table2 t2
     )
     on t1.ClosestSiteId = t2.SiteId;
group by t1.name, t1.ClosestSiteId, (seqnum - 1) / 10
order by name, ClosestSideId, DateTime;

平均每n列内部加入Sql Server

1 个答案: