Chinaunix首页 | 论坛 | 博客
  • 博客访问: 369593
  • 博文数量: 100
  • 博客积分: 2586
  • 博客等级: 少校
  • 技术积分: 829
  • 用 户 组: 普通用户
  • 注册时间: 2008-04-09 15:20
个人简介

我是一个Java爱好者

文章分类

全部博文(100)

文章存档

2014年(2)

2013年(7)

2012年(2)

2010年(44)

2009年(28)

2008年(17)

我的朋友

分类: Oracle

2008-10-16 10:57:02

一:分析函数over
分析函数用于计算基于组的某种聚合值,它和聚合函数的不同之处是
对于每个组返回多行,而聚合函数对于每个组只返回一行。                                       
1:统计某商店的营业额。       
    date      sale
            20
            15
            14
            18
            30
   规则:按天统计:每天都统计前面几天的总额
   得到的结果:
   DATE  SALE      SUM
   ----- -------- ------
      20       20          --1天          
      15       35          --1天+2天          
      14       49          --1天+2天+3天          
      18       67                  
      30       97           .
    
2:统计各班成绩第一名的同学信息
   NAME  CLASS S                        
   ----- ----- ----------------------
   fda      80                    
   ffd      78                    
   dss      95                    
   cfe      74                    
   gds      92                    
   gf       99                    
   ddd      99                    
   adf      45                    
   asdf     55                    
   3dd      78             
  
   通过:  
   --
   select * from                                                                      
                                                                            
   select name,class,s,rank()over(partition by class order by s desc) mm from t2
                                                                            
   where mm=1
   --
   得到结果:
   NAME  CLASS S                      MM                                                                                       
   ----- ----- ---------------------- ----------------------
   dss      95                                        
   gds      92                                        
   gf       99                                        
   ddd      99                            
  
   注意:
   1.在求第一名成绩的时候,不能用row_number(),因为如果同班有两个并列第一,row_number()只返回一个结果         
   2.rank()和dense_rank()的区别是:
     --rank()是跳跃排序,有两个第二名时接下来就是第四名
     --dense_rank()l是连续排序,有两个第二名时仍然跟着第三名
    
    
3.分类统计 (并显示信息)
                      
   -- -- ----------------------
                      
                      
                      
                      
                      
                      
                      
                      
   3
  select a,c,sum(c)over(partition by a) from t2               
  得到结果:
       SUM(C)OVER(PARTITIONBYA)     
  -- -- ------- ------------------------
                            
                            
                            
                            
                            
                            
                            
                            
                            
 
  如果用sum,group by 则只能得到
  SUM(C)                           
  -- ----------------------
                     
                     
                     
                     
  无法得到B列值
二:开窗函数          
     开窗函数指定了分析函数工作的数据窗口大小,这个数据窗口大小可能会随着行的变化而变化,举例如下:
1:    
   over(order by salary)按照salary排序进行累计,order by是个默认的开窗函数
   over(partition by deptno)按照部门分区
2:
  over(order by salary range between 5 preceding and 5 following)
  每行对应的数据窗口是之前行幅度值不超过5,之后行幅度值不超过5
  例如:对于以下列
    aa
    1
    2
    2
    2
    3
    4
    5
    6
    7
    9
  
  sum(aa)over(order by aa range between 2 preceding and 2 following)
  得出的结果是
           AA                      SUM
           ---------------------- -------------------------------------------------------
                               10                                                     
                               14                                                     
                               14                                                     
                               14                                                     
                               18                                                     
                               18                                                     
                               22                                                     
                               18                                                               
                               22                                                               
                                                                                             
            
  就是说,对于aa=5的一行 ,sum为  5-1<=aa<=5+2 的和
  对于aa=2来说,sum=1+2+2+2+3+4=14   
  又如 对于aa=9 ,9-1<=aa<=9+2 只有9一个数,所以sum=9  
             
3:其它:
     over(order by salary rows between 2 preceding and 4 following)
         每行对应的数据窗口是之前2行,之后4行
4:下面三条语句等效:          
     over(order by salary rows between unbounded preceding and unbounded following)
         每行对应的数据窗口是从第一行到最后一行,等效:
     over(order by salary range between unbounded preceding and unbounded following)
          等效
     over(partition by null)

范例:

 EMP有3个字段 EMANE,DEPTNO,SAL求查询每个部门薪水前十名的纪录 用一条SQL语句

SELECT *
 
FROM (SELECT ENAME,
               DEPTNO,
               SAL,
               RANK()
OVER(PARTITION BY DEPTNO ORDER BY SAL) RK
         
FROM EMP)
WHERE RK <= 10;

阅读(645) | 评论(0) | 转发(0) |
给主人留下些什么吧!~~