hive支持的分析函数:
总的概括:http://www.07net01.com/linux/HIVE_chuangkoujifenxihanshu_yingyongchangjing_532180_1373989446.html
********************************************************************************************************
Rank over的用法:http://www.cnblogs.com/mycoding/archive/2010/05/29/1747065.html
原始数据:
a           b           c
----------- ----------- ----
1           3           E
2           4           A
3           2           D
3           5           B
4           2           C
2           4           B
需求:以a,b进行分组,在每个组内以b进行排名。
select *,rank() over( partition by a,b order by b) from xxxx_tab ;
数据为:
a           b           c    rank
----------- ----------- ---- --------------------
1           3           E    1
2           4           A    1
2           4           B    1
3           2           D    1
3           5           B    2
4           2           C    1
分了5个组,第2行跟第3行是一个组,其他的每行是一个组。在第2行与第3行的组内以b排名,并列为1
***************************************************************************************************************************************
Row_Number() over的用法:http://www.cnblogs.com/fxgachiever/archive/2010/09/15/1826792.html
原始数据:
empid deptid salary
1           10          5500.00
2           10          4500.00
3           20          1900.00
4           20          4800.00
5           40          6500.00
6           40          14500.00
7           40          44500.00
8           50          6500.00
9           50          7500.00
需求:根据部门分组,显示每个部门的工资等级
SQL脚本:
SELECT *, Row_Number() OVER (partition by deptid ORDER BY salary desc) rank FROM employee
预期结果:
empid       deptid      salary                                  rank
----------- ----------- --------------------------------------- --------------------
1           10          5500.00                                 1
2           10          4500.00                                 2
4           20          4800.00                                 1
3           20          1900.00                                 2
7           40          44500.00                               1
6           40          14500.00                               2
5           40          6500.00                                 3
9           50          7500.00                                 1
8           50          6500.00                                 2
********************************************************************************************
窗口函数的用法:http://blog.csdn.net/cnham/article/details/6101199
select month,sum(tot_sales) month_sales, sum(sum(tot_sales)) over(order by month rows between unbounded preceding and unbounded following) total_sales from orders group by month.
原文:http://blog.csdn.net/purisuit_knowledge/article/details/44836257