我无法提出查询,找到所有已购买PROD1和PROD2的客户。
这是一个伪查询,看起来像我想要做的事情:(显然这不起作用)
SELECT COUNT(DISTINCT userid)
FROM TRANSACTIONS
WHERE product_id = 'prod1'
AND product_id = 'prod2'
所以基本上我正在尝试计算product_id“transactions
”和“prod1
”prod2
表中有交易的不同用户ID的数量。每个事务都存储在transactions
表中的一行中。
答案 0 :(得分:5)
SELECT userid
FROM TRANSACTIONS
WHERE product_id in ('prod1', 'prod2')
GROUP BY userid
HAVING COUNT(DISTINCT product_id) = 2
答案 1 :(得分:5)
我通过以下方式进行此类查询:
SELECT COUNT(DISTINCT t1.userid) AS user_count
FROM TRANSACTIONS t1
JOIN TRANSACTIONS t2 USING (userid)
WHERE t1.product_id = 'prod1'
AND t2.product_id = 'prod2';
@najmeddine的GROUP BY
解决方案shown也会产生你想要的答案,但它在MySQL上表现不佳。 MySQL很难优化GROUP BY
个查询。
您应该尝试两个查询,使用EXPLAIN
分析优化,并运行一些测试并根据数据库中的数据量计算结果。
答案 2 :(得分:2)
(使用用户提供的附加信息在下面添加了新选项)
尝试
SELECT * FROM Customers WHERE
EXISTS (SELECT * FROM Purchases WHERE ProductID = 'PROD1' AND CustID = Customers.CustID)
AND
EXISTS (SELECT * FROM Purchases WHERE ProductID = 'PROD2' AND CustID = Customers.CustID)
或者
SELECT * FROM Customers WHERE
CustID IN (SELECT CustID FROM Purchases WHERE ProductID = 'PROD1')
AND
CustID IN (SELECT CustID FROM Purchases WHERE ProductID = 'PROD2')
或者
SELECT UserID FROM Transactions WHERE ProductID = 'PROD1'
AND EXISTS (SELECT * FROM Transactions WHERE UserID = T1.UserID
AND ProductID = 'PROD2')
或者
SELECT UserID FROM Transactions WHERE ProductID = 'PROD1'
AND UserID IN (SELECT UserID FROM Transactions WHERE ProductID = 'PROD2')
答案 3 :(得分:0)
这是一个基于臭名昭着的Northwind示例数据库的Access答案。你应该很容易在mySql中翻译它。
SELECT o.CustomerID, Sum([ProductID]='Prod1') AS Expr1, Sum([productid]='Prod1') AS Expr2
FROM Orders AS o INNER JOIN [Order Details] AS d ON o.OrderID = d.OrderID
GROUP BY o.CustomerID
HAVING (((Sum([ProductID]='Prod1'))<>0) AND ((Sum([productid]='Prod1'))<>0));
答案 4 :(得分:0)
SELECT COUNT(DISTINCT userId)
FROM(
SELECT userId
FROM transactions
WHERE product = 'PROD1'
INTERSECT
SELECT userId
FROM transactions
WHERE product = 'PROD2');
查询创建两个中间表,一个包含购买PROD1的客户的userId,另一个包含购买PROD2的用户的相同表。 交集运算符返回一个表,该表仅包含在前两个表中找到的行,即那些购买两个产品的行。
答案 5 :(得分:0)
sakila db的例子:
SELECT R.customer_id, GROUP_CONCAT(I.film_id)
FROM sakila.rental R
RIGHT OUTER JOIN sakila.inventory I ON R.inventory_id = I.inventory_id
WHERE I.film_id IN (22,44) GROUP BY R.customer_id HAVING COUNT(*) = 2