我用google搜索和stackoverflowed高低,无法找到任何具体的解决方案来显示每组的第二(第n)记录。
考虑下表(按created_at desc排序):
----------------------------------------------
bid_id | status | created_at
----------------------------------------------
1 | cancelled | 2015-10-03
1 | awarded | 2015-10-02
1 | pending | 2015-10-01
2 | pending | 2015-10-01
3 | denied | 2015-10-02
3 | pending | 2015-10-01
结果输出应如下所示(按bid_id分组):
bid_id | status | created_at
----------------------------------------------
1 | awarded | 2014-10-02
3 | pending | 2014-10-01
实现这一目标的有效SQL是什么?非常感谢任何提示!
答案 0 :(得分:7)
您可以使用row_number
在bi_id
列排序的每个created_at
分区中分配行号。然后,您只需选择第二行或您需要的行。这适用于postgres
,因为它支持window
个函数。
select bid_id, status, created_at
from
(
select bid_id, status, created_at,
row_number() over(partition by bid_id order by created_at desc) as rn
from tablename
) x
where x.rn = 2;