SQL Server更新组

Question

我试图在MS-SQL上执行此操作,但只是在逐行返回错误

update #temp
Set Dos=Count(1)
From Temp_Table2010 s
where Id=s.Total and s.total in (Select Id from #temp)
group by s.Total

有谁知道如何解决这个性能良好的问题.

Answer 1

尝试

;with counts 
AS 
( 
    SELECT total, COUNT(*) as dos
    FROM temp_table2010 
    WHERE total in (select id from #temp)
)
UPDATE T 
SET dos=counts.dos
FROM #temp T 
INNER JOIN counts 
    ON t.id = counts.total 

Answer 2

在SQL Server中,您可以在更新查询中进行聚合,只需在子查询中执行聚合,然后将其加入要更新的表中.

UPDATE  #temp
SET     Dos = Cnt
FROM    #temp 
    INNER JOIN (SELECT Total, COUNT(*) AS Cnt FROM Temp_Table2010 GROUP BY Total) AS s
        ON Id = s.Total 

这样做:

WHERE total in (select id from #temp)

然后:

 INNER JOIN counts 
    ON t.id = counts.total 

多余.

联接解决了"总在(...)"的要求.组上键然后加入.

Answer 3

您不能在UPDATE查询中使用聚合,对于初学者 - 虽然您没有在原始问题中包含错误消息,但我怀疑这是它告诉您的内容.

您需要在更新之前计算聚合并将结果存储在临时表中,然后加入该表以进行更新.