SQL Server 2005 T-SQL 学习笔记：排名函数
2005中共有四个排名函数：Row_Number, Rank(), Dense_Rank(), NTILE()
Row_Number()：
按一个排序表达式，将查询结果编号，主要用于分页中使用.
Row_Number()简单示例：
select
SalesOrderID, OrderDate, CustomerID, SalesPersonID,row_number()
over
(
order
by
CustomerID)
as
RowNumber
from
Sales.SalesOrderHeader
where
SalesPersonID
is
not
null
Order
By
CustomerID
desc
--
查询结果的排序可以不同于row_number中的排序，这时RowNumber列将被打乱
结果:
SalesOrderID OrderDate CustomerID SalesPersonID RowNumber
------------ ----------------------- ----------- ------------- --------------------
50735 2003-06-01 00:00:00.000 701 285 3801
48352 2002-12-01 00:00:00.000 701 285 3802
49469 2003-03-01 00:00:00.000 701 285 3803
71843 2004-06-01 00:00:00.000 701 285 3804
53529 2003-09-01 00:00:00.000 701 285 3805
59043 2003-12-01 00:00:00.000 701 285 3806
63181 2004-02-01 00:00:00.000 700 279 3797
... ...
Row_Number中可以实现按一列的值不同而重新编号，这时在order by 前面加入Partition by [列名]
Row_Number中的分区示例:
select
SalesOrderID, OrderDate, CustomerID, SalesPersonID,row_number()
over
(partition
by
SalesPersonID
order
by
CustomerID)
as
RowNumber
from
Sales.SalesOrderHeader
where
SalesPersonID
is
not
null
and
SalesOrderID
>
65000
结果：
SalesOrderID OrderDate CustomerID SalesPersonID RowNumber
------------ ----------------------- ----------- ------------- --------------------
71919 2004-06-01 00:00:00.000 38 268 1
65298 2004-03-01 00:00:00.000 116 268 2
67286 2004-04-01 00:00:00.000 120 268 3
71779 2004-06-01 00:00:00.000 149 268 4
69528 2004-05-01 00:00:00.000 309 268 5
69545 2004-05-01 00:00:00.000 436 268 6
65294 2004-03-01 00:00:00.000 489 268 7
67277 2004-04-01 00:00:00.000 530 268 8
65310 2004-03-01 00:00:00.000 3 275 1
71889 2004-06-01 00:00:00.000 3 275 2
67292 2004-04-01 00:00:00.000 4 275 3
65315 2004-03-01 00:00:00.000 21 275 4
... ...
Rank():
用于对一列（多列也行)的值进行排名，有并列名次的，后面的名次会加上并列的次数．
Rank简单示例：
select
SalesOrderID, OrderDate, CustomerID, SalesPersonID,Rank()
over
(
order
by
CustomerID, SalesPersonID )
as
Rank
from
Sales.SalesOrderHeader
where
SalesPersonID
is
not
null
and
SalesOrderID
>
10000
结果
SalesOrderID OrderDate CustomerID SalesPersonID Rank
------------ ----------------------- ----------- ------------- --------------------
43860 2001-08-01 00:00:00.000 1 280 1
44501 2001-11-01 00:00:00.000 1 280 1
45283 2002-02-01 00:00:00.000 1 280 1
46042 2002-05-01 00:00:00.000 1 280 1
46976 2002-08-01 00:00:00.000 2 283 5
47997 2002-11-01 00:00:00.000 2 283 5
49054 2003-02-01 00:00:00.000 2 283 5
50216 2003-05-01 00:00:00.000 2 283 5
51728 2003-08-01 00:00:00.000 2 283 5
57044 2003-11-01 00:00:00.000 2 283 5
63198 2004-02-01 00:00:00.000 2 283 5
69488 2004-05-01 00:00:00.000 2 283 5
71889 2004-06-01 00:00:00.000 3 275 13
... ...
和Row_Number一样，Rank也可以使用Partition By
Dense_Rank():
和Rank很像，只不过并列名次的，后面的名次连续显示．
Dense_Rank简单示例:
select
SalesOrderID, OrderDate, CustomerID, SalesPersonID,Dense_Rank()
over
(
order
by
CustomerID, SalesPersonID )
as
Dense_Rank
from
Sales.SalesOrderHeader
where
SalesPersonID
is
not
null
and
SalesOrderID
>
10000
结果
SalesOrderID OrderDate CustomerID SalesPersonID Dense_Rank
------------ ----------------------- ----------- ------------- --------------------
43860 2001-08-01 00:00:00.000 1 280 1
44501 2001-11-01 00:00:00.000 1 280 1
45283 2002-02-01 00:00:00.000 1 280 1
46042 2002-05-01 00:00:00.000 1 280 1
46976 2002-08-01 00:00:00.000 2 283 2
47997 2002-11-01 00:00:00.000 2 283 2
49054 2003-02-01 00:00:00.000 2 283 2
50216 2003-05-01 00:00:00.000 2 283 2
51728 2003-08-01 00:00:00.000 2 283 2
57044 2003-11-01 00:00:00.000 2 283 2
63198 2004-02-01 00:00:00.000 2 283 2
69488 2004-05-01 00:00:00.000 2 283 2
71889 2004-06-01 00:00:00.000 3 275 3
... ...
NTILE(N):
给定一个总名次N，将查询结果按指定的条件排名次，并列时，名次连续.
NTILE的分区示例：
select
SalesOrderID, OrderDate, CustomerID, SalesPersonID,NTILE(
10
)
over
( partition
by
SalesPersonID
order
by
CustomerID)
as
NTILE
from
Sales.SalesOrderHeader
where
SalesPersonID
is
not
null
and
SalesOrderID
>
60000
结果
SalesOrderID OrderDate CustomerID SalesPersonID NTILE
------------ ----------------------- ----------- ------------- --------------------
71919 2004-06-01 00:00:00.000 38 268 1
65298 2004-03-01 00:00:00.000 116 268 2
67286 2004-04-01 00:00:00.000 120 268 3
71779 2004-06-01 00:00:00.000 149 268 4
61200 2004-01-01 00:00:00.000 205 268 5
69528 2004-05-01 00:00:00.000 309 268 6
61203 2004-01-01 00:00:00.000 421 268 7
69545 2004-05-01 00:00:00.000 436 268 8
65294 2004-03-01 00:00:00.000 489 268 9
67277 2004-04-01 00:00:00.000 530 268 10
65310 2004-03-01 00:00:00.000 3 275 1
71889 2004-06-01 00:00:00.000 3 275 1
67292 2004-04-01 00:00:00.000 4 275 1
61193 2004-01-01 00:00:00.000 4 275 1
65315 2004-03-01 00:00:00.000 21 275 1
71881 2004-06-01 00:00:00.000 21 275 1
69495 2004-05-01 00:00:00.000 22 275 1
63162 2004-02-01 00:00:00.000 22 275 1
65301 2004-03-01 00:00:00.000 39 275 1
71912 2004-06-01 00:00:00.000 39 275 2
... ...