我在过去几天一直在研究这个问题,无论我做什么,我都会得到同样错误的记录集。
表:
帐户
id | Customer | dateOpened
--------------------------
1 | ConAgra | 11/01/2013
2 | Fedex | 06/21/2014
CaseStatus
id | caseStatus
---------------
1 | A
2 | B
3 | M
4 | C
5 | H
6 | W
m_account_caseStatus
AccountID | caseStatusID | startDate | endDate
----------------------------------------------
1 | 2 | 11/01/2013| 12/15/2013
1 | 1 | 12/15/2013| 2/03/2014
1 | 2 | 2/03/2014 | 3/17/2014
2 | 6 | 6/21/2014 | 8/25/2014
2 | 3 | 8/25/2014 | 10/21/2014
2 | 1 | 10/21/2014| NULL
我需要的是过去两年中开设的所有帐户以及最早的startDate和caseStatus,其中caseStatus是A或B(不是两者都有,不管是先分配的)。我一直得到的只是minDate是A或B的帐户,这个帐户明显更少。
select
c.ID, c.Customer, cs.caseStatus
from
m_account_caseStatus m
left outer join
caseStatus cs on m.caseStatusID = cs.ID
left outer join
Account a on m.accountID = a.ID
where
a.dateOpened >= dateAdd(yyyy, -2, getDate())
and caseStatus IN ('A','B')
and (a.startDate = (select min(startdate)
from m_account_caseStatus sub1
where sub1.accountID = c.ID))
我多次移动子查询和caseStatus重新查询了查询,但我总是最终只得到他们以A或B状态开始而不是所有情况和日期/状态的情况它是第一个A或B.
感谢您提供的任何帮助。
答案 0 :(得分:0)
declare @year2 datetime = dateAdd(yyyy, -2, getdate())
select accountID from
(
-- all A or B accounts withing 2 years
-- num = 1 (2,3..) is the 1st (2nd, 3rd...) date when account got A or B status
select
m.accountID,
row_number() over (partition by m.accountID order by startDate) as num
from
m_account_caseStatus m
join
Account a on m.accountID = a.ID
where
a.dateOpened >= year2
and m.caseStatusId in (1,2)
) T
where T.num = 1
答案 1 :(得分:0)
;with c as
(
select accountID,caseStatusID,startDate,
ROW_NUMBER() OVER (PARTITION BY accountID ORDER BY startDate DESC) rnk
from m_account_caseStatus
)
select c.accountID,a.Customer,a.dateOpened,c.caseStatusID,cs.caseStatus,c.startDate
from c
join Account a on a.ID=c.accountID
join caseStatus cs on cs.ID=c.caseStatusID
where rnk=1 and a.dateOpened>=dateAdd(yyyy, -2, getDate())