Question

SELECT
  ( Select 
COUNT(t2.ID) as Total_HH_Memebers_Primary,t1.SETTLEMENT_ID 
from 
PScData t1 
INNER JOIN 
PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
WHERE  EXISTS (SELECT 1
               FROM   PScData t1
               WHERE  t1.SETTLEMENT_ID = SETTLEMENT_ID 
                AND t2.EDUCATION_LEVEL = 5 
               HAVING Count(*) > 1)
               Group By t1.SETTLEMENT_ID 
) col1



  , (Select 
COUNT(t2.ID) as Total_HH_Memebers_Secondary,t1.SETTLEMENT_ID 
from 
PScData t1 
INNER JOIN 
PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
 WHERE  EXISTS (SELECT 1
               FROM   PScData t1
               WHERE  t1.SETTLEMENT_ID = SETTLEMENT_ID 
                AND t2.EDUCATION_LEVEL = 10 
              HAVING Count(*) > 1) 
               Group By t1.SETTLEMENT_ID 

) col2

这是我的查询，我想根据＆＃34; WHERE＆＃34;得到两个不同的结果。条件来自同一个表格的列，当我这样做它给出这个错误。有人如何正确吗？

Answer 1

使用条件聚合：

SELECT
    SUM(CASE WHEN t2.EDUCATION_LEVEL = 5
             THEN 1 ELSE 0 END) AS Total_HH_Members_Primary,
    SUM(CASE WHEN t2.EDUCATION_LEVEL = 10
             THEN 1 ELSE 0 END) AS Total_HH_Members_Secondary,
    t1.SETTLEMENT_ID 
FROM PScData t1 
INNER JOIN PscMemberData t2
    ON t2._PARENT_AURI = t1.URI
GROUP BY
    t1.SETTLEMENT_ID
HAVING COUNT(*) > 1;

select子句中的两个子查询对我来说完全没用。而且我认为我们可以删除EXISTS子句，并将HAVING子句移动到外部查询。

Answer 2

尝试使用cte

;with col1 as 
(  Select 
COUNT(t2.ID) as Total_HH_Memebers_Primary,t1.SETTLEMENT_ID 
from 
PScData t1 
INNER JOIN 
PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
WHERE  EXISTS (SELECT 1
               FROM   PScData t1
               WHERE  t1.SETTLEMENT_ID = SETTLEMENT_ID 
                AND t2.EDUCATION_LEVEL = 5 
               HAVING Count(*) > 1)
               Group By t1.SETTLEMENT_ID 
),col2 as 
(Select 
COUNT(t2.ID) as Total_HH_Memebers_Secondary,t1.SETTLEMENT_ID 
from 
PScData t1 
INNER JOIN 
PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
 WHERE  EXISTS (SELECT 1
               FROM   PScData t1
               WHERE  t1.SETTLEMENT_ID = SETTLEMENT_ID 
                AND t2.EDUCATION_LEVEL = 10 
              HAVING Count(*) > 1) 
               Group By t1.SETTLEMENT_ID 

)  

select col1.SETTLEMENT_ID,col2.Total_HH_Memebers_Secondary, col1.Total_HH_Memebers_Primary from col1 
join col2 on col1.SETTLEMENT_ID=col2.SETTLEMENT_ID

Answer 3

错误消息告诉您：选择列表中的子查询不允许返回多于一个值，对于每个子查询都是如此。

您有2列和此查询返回的多行：

Select COUNT(t2.ID) as Total_HH_Memebers_Primary, t1.SETTLEMENT_ID 
from PScData t1 
INNER JOIN PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
WHERE  EXISTS (SELECT 1
               FROM   PScData t1
               WHERE  t1.SETTLEMENT_ID = SETTLEMENT_ID 
                AND t2.EDUCATION_LEVEL = 5 
               HAVING Count(*) > 1)
Group By t1.SETTLEMENT_ID

所以不允许作为＆＃34;子查询＆＃34;在一个选择条款中。

您需要使用＆＃34;条件聚合＆＃34;相反，在别处解释。

Answer 4

我已接受上述答案，我也找到了另一种方法：

这两种方法都是可行的，我发现的方法如下：

WITH EDU_LEVEL AS 
(SELECT t1.SETTLEMENT_ID,t2.EDUCATION_LEVEL from 
PScData t1 
INNER JOIN 
PscMemberData t2 ON t2._PARENT_AURI  = t1.URI
)


SELECT  SETTLEMENT_ID,
(Select COUNT(*) FROM EDU_LEVEL    WHERE  SETTLEMENT_ID = T.SETTLEMENT_ID AND EDUCATION_LEVEL = 5 ) Total_HH_Memebers_Primary,
(Select COUNT(*) FROM EDU_LEVEL    WHERE  SETTLEMENT_ID = T.SETTLEMENT_ID AND EDUCATION_LEVEL = 10 ) Total_HH_Memebers_Secondary



FROM EDU_LEVEL AS T
Group By SETTLEMENT_ID

当EXISTS未引入子查询时，只能在选择列表中指定一个表达式

4 个答案: