三张表连接后如何找到最常见的值

Question

我必须为每个用户编写一个 SQL 查询，该查询将返回该用户最常预订的房间的名称。

我创建了三个表之一：

SELECT   User.Name as user_name, Room.Name as room_reser  
FROM     Reservation 
INNER JOIN User ON User.Id = Reservation.UserId 
INNER JOIN Room ON Room.Id = Reservation.RoomId

3 人表：

Name          room_rese                             name     common_room
Jack           room_1
Anna           room_2                I need =>      Jack       room_1
Jack           room_1
Anna           room_1                               Anna        room_2
Jack           room_2
Anna           room_2

我尝试过这样的事情，但我不知道在这种情况下如何使用它：

SELECT DISTINCT r.user_name, (
    select b.room_reser
    from Reservation b
    where b.user_name = r.user_name
    group by b.user_name, b.roo_reser
    order by count(*) desc
    limit 1
    ) as roo_reser from Reservation r)`

Answer 1

如果您运行的数据库支持窗口函数，则可以使用聚合和窗口函数rank()：

select user_name, room_name
from (
    select   
        us.name as user_name, 
        ro.name as room_name,
        rank() over(partition by re.userid order by count(*) desc) rn
    from reservation re
    inner join user us on us.id = re.userid 
    inner join room ro on ro.id = re.roomid
    group by re.userid, re.roomid, us.name, ro.name
) t
where rn = 1

内部查询按用户名和房间聚合，并对每个用户的房间进行排名。外部查询过滤每个用户的顶部房间。如果有平局（即用户预订最多的两个房间的预订数量相同），则两者都会显示 - 如果即使有平局也想要一条记录，您可以添加另一个排序条件来打破平局.

---

如果您的数据库不支持窗口函数，您可以尝试在 having 子句中使用聚合相关子查询进行过滤：

select   
    us.name as user_name, 
    ro.name as room_name
from reservation re
inner join user us on us.id = re.userid 
inner join room ro on ro.id = re.roomid
group by us.name, ro.name
having count(*) = (
    select count(*) 
    from reservation re1 
    where re1.userid = re.userid
    group by re1.roomid
    order by count(*) desc
    limit 1
)

Answer 2

SELECT   DISTINCT User.Name as user_name, Room.Name as room_reser  
FROM     Reservation 
INNER JOIN User ON User.Id = Reservation.UserId 
INNER JOIN Room ON Room.Id = Reservation.RoomId 
GROUP BY user_name, room_reser 
ORDER BY COUNT(room_reser)