Question

我正在针对 MS SQL Server 2008 运行查询，并且正在选择一个帐号以及按帐号分组的列mydate的最大值：

select AccountNumber,
       max(mydate),
from #SampleData
group by AccountNumber

我想在结果中添加一列，其中包含与AccountNumber组关联的第二高的mydate。我知道它必须是这样的：

select max(mydate)
from #SampleData
where mydate < (select max(mydate) from #SampleData)

但是如何在一个选择查询中获得最大和第二个最大值？

Answer 1

您没有指定DBMS，因此这是ANSI SQL：

select accountnumber,
       rn,
       mydate
from (
   select accountnumber, 
          mydate, 
          row_number() over (partition by accountnumber order by mydate desc) as rn
   from #SampleData
) t
where rn <= 2;

Answer 2

试试这个

Select AccountNumber,
MAX(Case when Rnum = 1 Then mydate END) mydate_1,
MAX(Case when Rnum = 2 Then mydate END) mydate_2
From
(
select 
AccountNumber, mydate,
ROW_NUMBER() OVER (PARTITION By AccountNumber ORDER BY mydate DESC) as Rnum
from #SampleData
) V
Group By AccountNumber

Answer 3

这样的东西应该选择第二高：

select AccountNumber,
   max(mydate),
   (select max(SD2.mydate) from #SampleData SD2 where SD2.AccountNumber=#SampleData.AccountNumber AND SD2.mydate<max(#SampleData.mydate))
from #SampleData
group by AccountNumber

Answer 4

您还可以使用TOP N子句结合订单：

select    
          TOP 2
          accountnumber, 
          mydate, 
          row_number() over (partition by accountnumber order by mydate desc) as rn
from #SampleData
ORDER BY
          row_number() over (partition by accountnumber order by mydate desc)

SQL Select MAX和2nd MAX

4 个答案: