Question

我有一个包含名为MySubId的列的行的表。此列中的值可以重复。我想找到MySubId值以及出现次数最多的MySubId值的行数。

我有以下查询：

SELECT MySubId, COUNT(MySubId) AS MySubIdCount
FROM MyTable
GROUP BY MySubId
HAVING COUNT(MySubId)=MAX(COUNT(MySubId))

但我收到错误：

消息130，级别15，状态1，行4无法执行聚合对包含聚合或子查询的表达式起作用。

使用简单查询无法做到这一点吗？在HAVING子句中使用它之前，我必须嵌入一个子查询来计算MAX吗？

更新

我看到很多anwers使用TOP 1过滤结果集，所以我假设没有办法使用MAX函数将此查询过滤到最大值{{ {1}}值？

Answer 1

select top 1 with ties MySubId, Count(MySubId) as MySubIdCount
from MyTable
group by MuSubId
order by 2 desc

Answer 2

select top 1 MySubId, Count(MySubId)
from MyTable
group by MySubId
order by count(MySubId) DESC

Answer 3

这只是一个FYI，因为我相信这将是海报的下一个合乎逻辑的步骤

如果要删除所有以最新...

开头的重复项

DECLARE @ROWCOUNT INT
SET @ROWCOUNT = 1

WHILE @ROWCOUNT > 0
BEGIN
     DELETE 
     FROM MyTable
     WHERE ID IN
     (
        SELECT MAX(ID)
        FROM MyTable
        GROUP By MySubID
        HAVING COUNT(1) > 1
     )
     SET @ROWCOUNT = @@ROWCOUNT
END

Answer 4

DECLARE @MyTable TABLE (
    MySubId INT
)

INSERT INTO @MyTable (MySubId) VALUES (1)
INSERT INTO @MyTable (MySubId) VALUES (1)
INSERT INTO @MyTable (MySubId) VALUES (1)
INSERT INTO @MyTable (MySubId) VALUES (2)
INSERT INTO @MyTable (MySubId) VALUES (2)
INSERT INTO @MyTable (MySubId) VALUES (2)
INSERT INTO @MyTable (MySubId) VALUES (2)
INSERT INTO @MyTable (MySubId) VALUES (3)
INSERT INTO @MyTable (MySubId) VALUES (3)
INSERT INTO @MyTable (MySubId) VALUES (3)
INSERT INTO @MyTable (MySubId) VALUES (3)

;WITH Counts_CTE (MySubId, MySubIdCount, RowNumber)
AS
(
    SELECT 
        MySubId, 
        COUNT(MySubId) AS MySubIdCount,
        DENSE_RANK() OVER (ORDER BY COUNT(MySubId) DESC) AS RowNumber
    FROM @MyTable
    GROUP BY MySubId
)
SELECT *
FROM Counts_CTE
WHERE RowNumber = 1

SQL Server GROUP BY进退两难

4 个答案: