2

我有这个 sql 代码：

SELECT   NoteOID = HCN.ObjectID,
         PatientOID = HCN.PatientID,
         PatientVisitOID = 0,
         CollectedDT = HCN.CollectedDT
FROM     HClinicalNote HCN WITH(NOLOCK)
where    HCN.enddt is NULL
         and HCN.visitid in (select distinct visitOID from @tblCensus)
order by HCN.PatientID,HCN.CollectedDT desc

给出这些结果：

 NoteOID    PatientOID     CollectedDT
 181382     890855         2011-09-14 21:31:00
 169115     890855         2011-09-12 18:38:00
 177466     890855         2011-09-09 19:49:00
 175150     890855         2011-09-07 19:34:00
 174057     890855         2011-09-06 19:25:00
 172429     890855         2011-09-04 09:00:00
 181387     13462666       2011-09-14 21:37:00
 182224     13462666       2011-09-14 13:24:00
 179269     13462666       2011-09-12 18:12:00

我想从每组 PatientOID 中获得前 2 个 CollectedDT。

4

2 回答 2

4

如果您至少使用 SQL-Server 2005，则可以使用CTEwithROW_NUMBER函数：

WITH CTE AS(
   SELECT  NoteOID = HCN.ObjectID,
           PatientOID = HCN.PatientID,
           PatientVisitOID = 0,
           CollectedDT = HCN.CollectedDT,
           RN = ROW_NUMBER()OVER(PARTITION BY PatientOID ORDER BY CollectedDT ASC)
   FROM HClinicalNote HCN
   WHERE HCN.enddt is NULL
   AND HCN.visitid in (select distinct visitOID from @tblCensus)
)
SELECT * FROM CTE
WHERE RN <= 2
于 2012-07-19T15:04:06.967 回答
0 
select T2.* from (SELECT  NoteOID = HCN.ObjectID, 
       PatientOID = HCN.PatientID,   
       PatientVisitOID = 0,   
       CollectedDT = HCN.CollectedDT,  
       RN = ROW_NUMBER()OVER(PARTITION BY PatientOID ORDER BY CollectedDT ASC)  
FROM HClinicalNote HCN   
WHERE HCN.enddt is NULL  
AND HCN.visitid in (select distinct visitOID from @tblCensus) ) as T2   
where T2.RN >=2
于 2012-07-20T08:14:15.567 回答