分组后取每组内排名的Top N的SQL语句

给个MySQL例子参考 

-----查询每门课程的前2名成绩   

CREATE   TABLE   StudentGrade( 
stuId   CHAR(4),         --学号 
subId   INT,                 --课程号 
grade   INT,                 --成绩 
PRIMARY   KEY   (stuId,subId) 
) 
GO 
--表中数据如下 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '001 ',1,97); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '001 ',2,50); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '001 ',3,70); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '002 ',1,92); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '002 ',2,80); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '002 ',3,30); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '003 ',1,93); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '003 ',2,95); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '003 ',3,85); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '004 ',1,73); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '004 ',2,78); 
INSERT   INTO   StudentGrade(stuId,subId,grade)   VALUES( '004 ',3,87); 
GO 
 
--查看数据 
select   *   from   StudentGrade 


--假如出现并列时，也只取两个同学的话。 
--方法一： 
select   distinct   * 
from   studentgrade   as   t1 
where   stuid   in 
(select   top   2   stuid 
  from   studentgrade   as   t2 
  where   t1.subid=t2.subid 
                  order   by   t2.grade   desc) 
order   by   subid,   grade   desc 

--方法二： 
select   *   from   StudentGrade   a   where   (select   count(1)   from   studentGrade   where   subId=a.subId   and   grade> =a.grade) <=2 

--方法三： 
select   *   from   StudentGrade   t 
where   (select   count(1)   from   StudentGrade   where   subid=t.subid   and   grade> t.grade) <=1 
order   by   subId,grade   desc 

--结果 
 

drop   table   StudentGrade