分组统计并且累加的问题

分组统计并且累加的问题

　　问：说有张表 Table (Date, Name, Income). 我想按Name和Date进行月份分组合计Income(不同的姓名和月份要区分出来)，并且增加一个字段Cum，显示各月的Income的累加(当然月份是按升序排列)。Date字段是日期型，格式:mm/dd/yyyy,这个查询怎么写?谢谢!

　　答：如果没猜错的话你的原意，Income 和Cum是两个字段，结果集应该是:

　　Month Name Income/Month Cum

Cum应该显示逐月累加的值，Income/Month显示的是各月的小计值。

Declare @t Table(Date datetime,name varchar(50),Income int)
Insert @t Select '2006-5-9','A',200
Union all Select '2006-5-19','B',300
Union all Select '2006-6-19','A',300
Union all Select '2006-6-29','B',300

--group by convert(varchar(7) , date , 121)

/*
Select sum(Income) ,convert(varchar(7) , Date , 121) , Name
from @t  group by name, convert(varchar(7) , Date , 121)   with rollup
order by  convert(varchar(7) , Date , 121)
*/

SELECT   convert(varchar(7) , Date , 121) as date1,
      --   CASE WHEN (GROUPING( convert(varchar(7) ,Date,121) ) = 1) THEN 'ALL'
        --   ELSE   convert(varchar(7) , Date , 121) --ISNULL(convert(varchar(7) , Date , 121), 'UNKNOWN')
--WHEN (GROUPING( convert(varchar(7) ,Date,121) ) <> 1) THEN convert(varchar(7) ,Date,121)
--         END AS Date2,
       CASE WHEN (GROUPING(name) = 1) THEN 'ALL'
            ELSE ISNULL(name, 'UNKNOWN')
       END AS Name,
       SUM(Income) AS QtySum
FROM @t
--where 
--date1 is not null
GROUP BY convert(varchar(7) , Date , 121), name WITH ROLLUP
order by convert(varchar(7) , Date , 121) asc

-------
/*

Month  Name                                               Income/Month CUM        
------ -------------------------------------------------- ------------ -----------
200605 A                                                  200          200
200606 A                                                  300          500
200605 B                                                  300          300
200606 B                                                  300          600
*/