Question

假设我有一张桌子SALES

Date         ArticleID  CustomerID SalesID
2016-04-01      4          1000      1
2016-04-01      8          1000      2
2016-04-01      4          1000      3
2016-04-01      12         1000      4
2016-04-02      4          1000      5
2016-04-03      6          800       6
2016-04-03      4          1000      7
2016-04-05      99         900       8
2016-04-06      4          700       9

我的目标是创建一个查询，该查询仅返回同一客户在同一天购买相同文章的行。

在这种情况下，结果应为：

Date         ArticleID  CustomerID SalesID
2016-04-01      4          1000      1
2016-04-01      4          1000      3

如何以最佳方式编写此查询？谢谢你的帮助！

Answer 1

您可以使用GROUP BY和HAVING：

来实现

SELECT t.*
FROM tbl t
INNER JOIN (
    SELECT CustomerID, Date, ArticleID
    FROM tbl
    GROUP BY CustomerID, Date, ArticleID
    HAVING COUNT(SalesID) >= 2
)a
    ON a.CustomerID = t.CustomerID
    AND a.Date = t.Date
    AND a.ArticleID = t.ArticleID

另一种方法是使用COUNT OVER：

WITH Cte AS(
    SELECT *,
        cnt = COUNT(SalesID) OVER(PARTITION BY CustomerID, Date, ArticleID)
    FROM tbl
)
SELECT
    Date, ArticleID, CustomerID, SalesID
FROM Cte
WHERE cnt >= 2

ONLINE DEMO

Answer 2

一种方法是使用CTE group by和having，然后加入原始表：

;WITH CTE AS
(
SELECT Date, ArticleId, CustomerId
FROM Sales
GROUP BY Date, ArticleId, CustomerId
HAVING COUNT(SalesId) > 1
)

SELECT S.Date, S.ArticleId, S.CustomerId, S.SalesId
FROM Sales S
INNER JOIN CTE C ON(S.Date = C.Date AND S.ArticleId = C.ArticleId AND S.CustomerId = C.CustomerId)

Answer 3

请尝试以下查询

select date,articleid,customerid,salesid from sales 
inner join 
(select date,articleid,customerid 
from sales 
group by date,articleid,customerid 
having count(1)>1) b
on sales.date=b.date
and sales.articleid=b.articleid
and sales.customerid=b.customerid

Answer 4

使用子查询执行此操作。

SELECT A.*
FROM YourTable AS A INNER JOIN (
SELECT Date, ArticleID, CustomerID, COUNT(*) AS NumberOfCount
FROM YourTable
GROUP BY Date, ArticleID, CustomerID
HAVING COUNT(*) >1 ) AS B
ON A.Date = B.Date
AND A.ArticleID = B.ArticleID
AND A.CustomerID = B.CustomerID

Answer 5

您可以在COUNT()中使用OVER：

select Date_, ArticleID, CustomerId, SalesID
from (select *, 
             count(*) over(partition by Date_, ArticleID, CustomerId) cnt 
      from #t )x
where cnt > 1

查询以显示客户是否在同一天购买了相同的artictle

5 个答案: