聚合数据

[sg2333333234]

连续聚合数据（提示可以用游标 循环）
create table sales
( empno int,
saledate datetime,
qty int
);

insert into sales values(1,'2007-11-1',10);
insert into sales values(1,'2007-11-1',10);
insert into sales values(1,'2007-11-2',20);
insert into sales values(1,'2007-11-2',20);
insert into sales values(1,'2007-11-3',20);
insert into sales values(1,'2007-11-4',30);
insert into sales values(1,'2007-11-4',30);

insert into sales values(2,'2007-11-1',10);
insert into sales values(2,'2007-11-1',10);
insert into sales values(2,'2007-11-2',20);
insert into sales values(2,'2007-11-2',20);
insert into sales values(2,'2007-11-3',20);
insert into sales values(2,'2007-11-4',30);
insert into sales values(2,'2007-11-4',30);

示例数据结果显示如下：
1 2007-11-1 20
1 2007-11-2 60
1 2007-11-3 80
1 2007-11-4 140
2 2007-11-1 20
2 2007-11-2 60
2 2007-11-3 80
2 2007-11-4 140

[CN_SQL]

什么意思？最后是你想要的结果？

[-狙击手-]

create table sales
( empno int,
saledate datetime,
qty int
);

－－

select empno,convert(char(10),saledate,120) as saledate,sum(qty) as qty
from sales
group by empno,convert(char(10),saledate,120)

[sg2333333234]

这个是想要得到的结果

数据结果显示如下：
1 2007-11-1 20
1 2007-11-2 60
1 2007-11-3 80
1 2007-11-4 140
2 2007-11-1 20
2 2007-11-2 60
2 2007-11-3 80
2 2007-11-4 140

[-狙击手-]

create table sales 
( empno int, 
  saledate datetime, 
  qty  int 
); 

insert into sales values(1,'2007-11-1',10); 
insert into sales values(1,'2007-11-1',10); 
insert into sales values(1,'2007-11-2',20); 
insert into sales values(1,'2007-11-2',20); 
insert into sales values(1,'2007-11-3',20); 
insert into sales values(1,'2007-11-4',30); 
insert into sales values(1,'2007-11-4',30); 

insert into sales values(2,'2007-11-1',10); 
insert into sales values(2,'2007-11-1',10); 
insert into sales values(2,'2007-11-2',20); 
insert into sales values(2,'2007-11-2',20); 
insert into sales values(2,'2007-11-3',20); 
insert into sales values(2,'2007-11-4',30); 
insert into sales values(2,'2007-11-4',30); 

select empno,saledate,(select sum(qty) from sales where empno = b.empno and saledate <= b.saledate)
from(
select distinct empno,convert(char(10),saledate,120) as saledate 
from sales a
) b
order by 1,2

drop table sales

/*
empno       saledate               
----------- ---------- ----------- 
1           2007-11-01 20
1           2007-11-02 60
1           2007-11-03 80
1           2007-11-04 140
2           2007-11-01 20
2           2007-11-02 60
2           2007-11-03 80
2           2007-11-04 140

（所影响的行数为 8 行）

*/

[CN_SQL]

啊，那就是一般的聚合嘛，参考happyflystone的了。

[百年树人]

create table sales 
( empno int, 
  saledate datetime, 
  qty  int 
); 

insert into sales values(1,'2007-11-1',10); 
insert into sales values(1,'2007-11-1',10); 
insert into sales values(1,'2007-11-2',20); 
insert into sales values(1,'2007-11-2',20); 
insert into sales values(1,'2007-11-3',20); 
insert into sales values(1,'2007-11-4',30); 
insert into sales values(1,'2007-11-4',30); 

insert into sales values(2,'2007-11-1',10); 
insert into sales values(2,'2007-11-1',10); 
insert into sales values(2,'2007-11-2',20); 
insert into sales values(2,'2007-11-2',20); 
insert into sales values(2,'2007-11-3',20); 
insert into sales values(2,'2007-11-4',30); 
insert into sales values(2,'2007-11-4',30); 

--begin
select empno,convert(varchar(10),saledate,120) as saledate,
qty=(select sum(qty) from sales  where saledate<=a.saledate and empno=a.empno) from sales a
group by empno,saledate
order by empno,saledate

--result
empno       saledate   qty         
----------- ---------- ----------- 
1           2007-11-01 20
1           2007-11-02 60
1           2007-11-03 80
1           2007-11-04 140
2           2007-11-01 20
2           2007-11-02 60
2           2007-11-03 80
2           2007-11-04 140

（所影响的行数为 8 行）