从具有相同 id 的一组值中选择最大值

Question

我有这两张表：

CREATE TABLE Category (

CategoryID int NOT NULL,
CategoryName varchar(255),

PRIMARY KEY (CategoryID)
);

CREATE TABLE Team (

CategoryID int NOT NULL,
TeamName varchar(255) NOT NULL,
Points int(255),

PRIMARY KEY (TeamName),
FOREIGN KEY (CategoryID) REFERENCES Category(CategoryID)
);

我想从每个类别中找出得分最高的球队。

到目前为止，我已经尝试过这些，我知道它们是错误的，但我走上了死胡同。任何建议都会有所帮助。

SELECT category.categoryid, team.teamname, MAX(team.points)
FROM category, team
GROUP BY team.teamname
HAVING MAX(team.points);


select category.categoryid, team.teamname
from category
right join team on team.categoryid = category.categoryid
group by team.teamname
having max(team.points);


select team.categoryid, team.teamname
from team, team a
having max(team.points);


select category.categoryid, team.teamname, max(team.points), team.points
from team
inner join category on category.categoryid = team.categoryid
group by teamname
having (select distinct max(team.points) from team);


select category.categoryid, team.teamname, team.points
from team
inner join category on category.categoryid = team.categoryid;

Answer 1

我想从每个类别中找出得分最高的球队。

WITH cte AS ( SELECT *, RANK() OVER (PARTITION BY CategoryID ORDER BY Team.Points DESC) rnk
              FROM Category 
              NATURAL JOIN Team )
SELECT *
FROM cte
WHERE rnk = 1

但是如果一个类别中有超过一个团队具有相同的最高分，那么所有团队都将被退回。

如果在这种情况下您只需要其中一个，则需要额外的排序（并且它必须提供行唯一性）。例如，它可以是

WITH cte AS ( SELECT *, RANK() OVER (PARTITION BY CategoryID ORDER BY Team.Points DESC, Team.TeamName) rnk
              FROM Category 
              NATURAL JOIN Team )
SELECT *
FROM cte
WHERE rnk = 1;

https://dbfiddle.uk/?rdbms=mysql_8.0&fiddle=5c73ac6a99e270535c70b173a65c02ff

当然，您可以只选择需要的列SELECT CategoryName, TeamName。

Answer 2

8.0 版支持dense_rank()

SELECT teamname
FROM (
    SELECT categoryid, teamname,
       dense_rank() over(partition by categoryid order by points desc) rnk
    FROM team
)t
WHERE rnk = 1
GROUP BY teamname
HAVING count(*) = (SELECT count(*) FROM Category);