语法：row_number() over(partition by …)
分组取最大的objid那一条,根据objid倒叙排序，取rn=1
select * from (select a.* ,row_number()over(partiotion by col1,col2 order by col1,col2) r from tablename ) where r=1;
分析函数和聚合函数的不同之处是什么?
普通的聚合函数用group by分组，每个分组返回一个统计值，而分析函数采用partition by分组，并且每组每行都可以返回一个统计值。
分析函数的形式
分析函数带有一个开窗函数over()，包含三个分析子句:分组(partition by), 排序(order by), 窗口(rows) ，他们的使用形式如下：over(partition by xxx order by yyy rows between zzz)
–利用开窗函数取每组数据的最后一条数据
sql>select * from (select row_number() over (partition by accountid order by updatetimestamp desc) rn, plm_openaccountrecon.* from plm_openaccountrecon) where rn = 1 and t.accountid = :accountid
–unbounded preceding and unbouned following针对当前所有记录的前一条、后一条记录，也就是表中的所有记录
–unbounded：不受控制的，无限的
–preceding：在…之前
–following：在…之后
–第一行到当前行
sql> select t.objid, t.applyempid, t.applydate, t.status, t.leavetype, t.leavedays, sum(leavedays) over(partition by applyempid order by objid rows between unbounded preceding and current row) lds from rt_gtcxleave t;
–当前行至最后一行
sql> select t.objid, t.applyempid, t.applydate, t.status, t.leavetype, t.leavedays, sum(leavedays) over(partition by applyempid order by objid rows between current row and unbounded following) lds from rt_gtcxleave t;
–当前行的上一行 (rownum – 1) 到当前行
–between 1 preceding and current row
sql> select t.objid, t.applyempid, t.applydate, t.status, t.leavetype, t.leavedays, sum(leavedays) over(partition by applyempid order by objid rows between 1 preceding and current row) lds from rt_gtcxleave t;
–当前行的上一行 (rownum – 1) 到当前行的下两行 (rownum + 2)
–rows between 1 preceding and 2 following