Question

我正在尝试编写一个查询来将用户表连接到活动记录表，并为每个用户返回以下内容：

A）他们上次登录的时间。

B）过去3个月的登录次数。

这是我到目前为止所提出的：

SELECT A.UserID, COUNT( Activity ) AS Logins, MAX( TIME ) AS LastLogin  
FROM UserMaster A  
LEFT JOIN UserWebActivity B ON A.UserID = B.UserID  
AND Activity =  'Login'  
AND TIME BETWEEN DATE_SUB( NOW( ) , INTERVAL 3 MONTH ) AND NOW( )  
GROUP BY A.UserID

这几乎可以使用，但它不会返回最近3个月内未登录的用户的最新登录信息。如何让count（）和max（）正常协同工作？

Answer 1

首先分别解决每个问题：

SELECT A.UserID, MAX(TIME) AS LastLogin  
FROM UserMaster A  
LEFT JOIN UserWebActivity B
ON A.UserID = B.UserID  
AND Activity = 'Login'  
GROUP BY A.UserID

SELECT A.UserID, COUNT(Activity) AS Logins
FROM UserMaster A  
LEFT JOIN UserWebActivity B ON A.UserID = B.UserID  
AND Activity = 'Login'  
AND TIME BETWEEN (NOW() - INTERVAL 3 MONTH) AND NOW( )  
GROUP BY A.UserID

单独测试它们以确保每个查询都能按您的需要工作，并在必要时进行调整。

然后当你对他们都工作感到高兴时，将结果加在一起：

SELECT T1.UserID, T1.LastLogin, T2.Logins
FROM
(
    SELECT A.UserID, MAX(TIME) AS LastLogin  
    FROM UserMaster A  
    LEFT JOIN UserWebActivity B
    ON A.UserID = B.UserID  
    AND Activity = 'Login'  
    GROUP BY A.UserID
) AS T1
JOIN
(
    SELECT A.UserID, COUNT(Activity) AS Logins
    FROM UserMaster A  
    LEFT JOIN UserWebActivity B
    ON A.UserID = B.UserID  
    AND Activity = 'Login'
    AND TIME BETWEEN (NOW() - INTERVAL 3 MONTH) AND NOW()
    GROUP BY A.UserID
) AS T2
ON T1.UserID = T2.UserID

这将允许MySQL充分利用不同查询的索引。

Answer 2

您可以在CASE中使用COUNT声明：

SELECT A.UserID,
  COUNT(
    CASE WHEN TIME BETWEEN DATE_SUB( NOW( ) , INTERVAL 3 MONTH ) AND NOW( )
    THEN Activity
  END ) AS Logins,
  MAX( TIME ) AS LastLogin  
FROM UserMaster A  
LEFT JOIN UserWebActivity B ON A.UserID = B.UserID  
AND Activity =  'Login'  
GROUP BY A.UserID

如果您拥有大量数据，此解决方案的效率将低于@MarkByers。

Answer 3

就像一个更疯狂的想法 - 而不是MAX（TIME）

coalesce(max(time), (select max(time) from UserWebActivity where ...) )

MYSQL上次登录和过去3个月的登录次数

3 个答案: