IT分销/经销数据库
能用sql实现吗？请求db2高人得帮助
表1
acctno bal trandate
--------------------------------------------
102012011 81.23 2010-01-03
102012011 100.21 2010-01-05
102012011 481.23 2010-02-04
102012011 1300.21 2010-04-05
102012011 281.23 2010-04-07
102012011 1020.21 2010-05-01
102012011 821.23 2010-05-21
102012011 100.21 2010-05-26
表二
acctno bal trandate trandate2
------------------------------------------------------
102012011 81.23 2010-01-03 2010-01-05
102012011 100.21 2010-01-05 2010-02-04
102012011 481.23 2010-02-04 2010-04-05
102012011 1300.21 2010-04-05 2010-04-07
102012011 281.23 2010-04-07 2010-05-01
102012011 1020.21 2010-05-01 2010-05-21
102012011 821.23 2010-05-21 2010-05-26
102012011 100.21 2010-05-26 2012-12-31
表一到表二的变换能用sql实现吗，如何解决这个问题请求db2高人的指导
acctno bal trandate
--------------------------------------------
102012011 81.23 2010-01-03
102012011 100.21 2010-01-05
102012011 481.23 2010-02-04
102012011 1300.21 2010-04-05
102012011 281.23 2010-04-07
102012011 1020.21 2010-05-01
102012011 821.23 2010-05-21
102012011 100.21 2010-05-26
表二
acctno bal trandate trandate2
------------------------------------------------------
102012011 81.23 2010-01-03 2010-01-05
102012011 100.21 2010-01-05 2010-02-04
102012011 481.23 2010-02-04 2010-04-05
102012011 1300.21 2010-04-05 2010-04-07
102012011 281.23 2010-04-07 2010-05-01
102012011 1020.21 2010-05-01 2010-05-21
102012011 821.23 2010-05-21 2010-05-26
102012011 100.21 2010-05-26 2012-12-31
表一到表二的变换能用sql实现吗，如何解决这个问题请求db2高人的指导
21同行回答
给你个olap 函数参考:
select accto, bal, trandate,
coalesce(
min(trandate) over
(partition by accto order by to_date(trandate, 'YYYY-MM-DD HH24:MI:SS') nulls last rows between 1 following and 1 following),
'2010-12-31 00:00:00'
) next_trandate
from table(values
(102012011, 81.23, '2010-01-03 00:00:00'),
(102012011, 100.21, '2010-01-05 00:00:00'),
(102012011, 481.23, '2010-02-04 00:00:00')
)
as t(accto, bal, trandate)收起
select accto, bal, trandate,
coalesce(
min(trandate) over
(partition by accto order by to_date(trandate, 'YYYY-MM-DD HH24:MI:SS') nulls last rows between 1 following and 1 following),
'2010-12-31 00:00:00'
) next_trandate
from table(values
(102012011, 81.23, '2010-01-03 00:00:00'),
(102012011, 100.21, '2010-01-05 00:00:00'),
(102012011, 481.23, '2010-02-04 00:00:00')
)
as t(accto, bal, trandate)收起
浏览724