sql window function

The SQL window function allows user to define a set of rows related to current row which is called a frame. Then user can apply aggregate function to the frame. It is standardized in SQL 2003 and supported by major RDBMS like:

•   SQL Server 2005+
•   Oracle
•   DB2
•   Postgre SQL

The window function can solve problem like ranking very easily without resorting to complex sub queries.

Here is an example. First comes the schema and sample data:

--dbext:profile=uca_dw_sh
create table empsalary (
    empno       bigint                not null,
    depname     varchar(30)           not null,
    salary      decimal(20, 2)        not null
    constraint PK_empsalary primary key (empno)
)
go

insert into empsalary(depname, empno, salary)
       values
       ('develop',8,6000),
       ('develop',10,5200),
       ('develop',11,5200),
       ('develop',9,4500),
       ('develop',7,4200),
       ('personnel',2,3900),
       ('personnel',5,3500),
       ('sales',1,5000),
       ('sales',4,4800),
       ('sales',3,4800)
go

 Then here are some SQLs to demonstrate the use of window function

-- list employee who is paid above department average
select t1.*
  from ( select empno,
                depname,
                salary,
                avg(salary) over (partition by depname) as dep_avg
           from empsalary
       ) t1
 where t1.salary > t1.dep_avg
go

-- rank employee accoring to salary within his department
select empno,
       depname,
       salary,
       rank() over (partition by depname order by salary desc)
  from empsalary
go

 

A good explanation of the use of window function can be found:
http://www.postgresql.org/docs/8.4/interactive/tutorial-window.html