--建表
Create table 表A
(
A1 datetime,
A2 varchar(10),
A3 varchar(10),
A4 int
)
Create Table 表B
(
B1 datetime,
B2 varchar(10),
B3 int
)
--插入数据
insert into 表A values('20141102','zz1','Y',800)
insert into 表A values('20141115','zz1','Y',999)
insert into 表A values('20141116','zz1','N',444)
insert into 表A values('20141202','zz1','Y',555)
insert into 表A values('20141127','as1','Y',333)
insert into 表A values('20141127','as1','Y',222)
insert into 表A values('20141129','as1','N',110)
insert into 表A values('20141210','as1','Y',800)
insert into 表A values('20141010','as1','Y',300)
insert into 表A values('20141010','zz1','Y',400)
insert into 表B values('20141105','zz1',100)
insert into 表B values('20141128','zz1',200)
insert into 表B values('20141214','zz1',300)
insert into 表B values('20141130','as1',150)
insert into 表B values('20141219','as1',300)
insert into 表B values('20141226','as1',200)
insert into 表B values('20141026','as1',200)
insert into 表B values('20141026','zz1',200)
--创建存储过程(你的结果需要动态行列转换)
Create Procedure sp_GetBaobiao(@StartDT datetime,@EndDT datetime)
As
Begin
/*
思路
1.把表A,表B某个时间段的数据保存到临时表
2.把两个临时表的日期转换位年月,并拼成字符串,如@[201411月],@[201412月]这样,这里加了@是后面需要加限定符
3.把临时表按年月汇总后,做行列转换(这里的年月是动态的),再按A2=B2做连接
4.计算未付款合计和全年已付款比率%
搞定
*/
Declare @Sql Varchar(8000)
Declare @F Varchar(1000)=''
Select * Into #表A From 表A Where A1 Between @StartDT And @EndDT
Select * Into #表B From 表B Where B1 Between @StartDT And @EndDT
Select @F=@F+'@['+A1+'],' From
(
Select Convert(Varchar(6),A1,112)+'月' As A1 From #表A
union
Select Convert(Varchar(6),B1,112)+'月' As A1 From #表B
)S
Set @F=STUFF(@F,LEN(@F),1,'')
Set @SQL='Select M.A2,'+replace(@F,'@','M.')+',应付款合计,'+replace(@F,'@','N.')+',
应付款合计-已付款合计 As 未付款合计,
Round(100.0*已付款合计/应付款合计,2) as [全年已付款比率%]
From
(
Select *,(Select Sum(Case When A3=''Y'' Then 1 Else -1 End*A4) From
#表A Where A2=pvt.A2) As 应付款合计
From
(
Select Convert(Varchar(6),A1,112)+''月'' As Y1,A2,
Sum(Case When A3=''Y'' Then 1 Else -1 End*A4) A4 From #表A
Group by Convert(Varchar(6),A1,112),A2
) A
Pivot
(
Sum(A4)
For Y1 in('+replace(@F,'@','')+')
)As pvt
) M
Left Join
(
Select *,(Select Sum(B3) From #表B Where B2=pvt.B2) As 已付款合计
From
(
Select Convert(Varchar(6),B1,112)+''月'' As Y1,B2,Sum(B3) As B3 From #表B
Group by Convert(Varchar(6),B1,112),B2
) A
Pivot
(
Sum(B3)
For Y1 in('+replace(@F,'@','')+')
)As pvt
) N on M.A2=N.B2'
Exec (@sql)
Drop Table #表A
Drop Table #表B
End
--得到结果,开始日期到结束日期(必须是同一年,否则全年已付款比率列不正确)
Exec sp_GetBaobiao '20141001','20141230'
--效果
图片附件: 游客没有浏览图片的权限,请
登录 或
注册
[
本帖最后由 mxbing1984 于 2015-3-3 11:53 编辑 ]