Question

我有一个像这样的表REPORTS：

DBSCODE   RPTDATE     DBSSEQ
AAA       2012-01-31  1
AAA       2012-02-29  1
AAA       2012-02-29  2
AAA       2012-03-31  1
BBB       2012-01-31  1
BBB       2012-01-31  2

我想为每个dbscode返回每个rptdate的最大序列的详细信息。但是，我现在正在运行的查询仅在最大dbsseq＆gt;时返回给我。 1.

我正在执行：

select a.dbscode, a.rptdate, a.dbseq
from REPORTS as a 
inner join 
    (select rptdate, max(dbsseq) as 'Last'
    from REPORTS
    group by rptdate) as b
  on a.rprtdate = b.last

我的结果是：

DBSCODE, RPTDATE, DBSSEQ
AAA, 2012-02-29, 2
BBB, 2012-01-31, 2

Answer 1

<强>更新

我重新读了你的问题，我相信你要找的是这个简单的查询..

SELECT 
  dbscode, rptdate, max(dbsseq)
FROM 
  REPORTS
GROUP BY
  dbscode, 
  rptdate

演示 http://sqlfiddle.com/#!3/b16ba/13

初步回答

~~在子查询中，您还需要按DBSCODE进行分组。~~

SELECT a.dbscode, a.rptdate, a.dbsseq FROM REPORTS as a INNER JOIN (SELECT dbscode, rptdate, MAX(dbsseq) AS 'Last' FROM REPORTS GROUP BY dbscode, rptdate) as b ON a.dbsseq = b.last AND a.dbscode = b.dbscode AND a.rptdate = b.rptdate

<击>

Answer 2

select dbscode, 
       rptdate, 
       dbseq
from (
   select dbscode, rptdate, dbseq,
          row_number() over (partition by dbscode order by rptdate desc) as rn
    from reports
) t
where rn = 1;

如果有多个行具有相同的rptdate，则只返回一个。如果您不想这样做，请使用dense_rank()代替row_number()

SQL查询 - 当它= 1时返回Max

2 个答案: