我想使用以下两个(简化)表格计算每个用户访问我网站的唯一天数:
Table: Users
+--------------+------------------+
| Field | Type |
+--------------+------------------+
| userid | int(10) |
| firstname | varchar(50) |
| lastname | varchar(50) |
+--------------+------------------+
Table: Visitors
+--------------+------------------+
| Field | Type |
+--------------+------------------+
| visitorid | int(10) |
| userid | int(10) |
| visitdate | timestamp |
+--------------+------------------+
首先,我按DATE(visitdate)
分组,按日期对记录进行分组。
然后,我按userid
分组,并对每个用户访问过的天数进行分组。
我对此的质询是:
SELECT count(userid) as count, userid, firstname, lastname FROM
(SELECT users.userid, firstname, lastname, DATE(`visitdate`) AS dateonly FROM
visitors JOIN users ON visitors.userid = users.userid
GROUP BY DATE(`visitdate`)) b
GROUP BY userid ORDER BY count DESC
然而,这会给出一个不正确的计数,当我只是添加一个WHERE
子句来限制用户ID时,会对此进行验证。以下查询为单个用户提供了正确的计数。
SELECT count(userid) as count, userid, firstname, lastname FROM
(SELECT users.userid, firstname, lastname, DATE(`visitdate`) AS dateonly FROM
visitors JOIN users ON visitors.userid = users.userid
WHERE users.userid = 167
GROUP BY DATE(`visitdate`)) b
GROUP BY userid ORDER BY count DESC
显然我错了 - 任何帮助都会受到赞赏。感谢。
答案 0 :(得分:0)
如果我理解你的问题,我认为这就是你想要的:
SELECT u.userid, u.firstname, u.lastname, COUNT(DISTINCT CAST(v.visitdate AS DATE)) AS visits
FROM users u
LEFT JOIN visitors v ON v.userid = u.userid
GROUP BY u.userid, u.firstname, u.lastname
这将显示每个用户的不同访问日期的用户ID,名字,姓氏和计数(如果没有访问则为0,因为它使用左连接来包含缺少访问者表中项目的用户)。