求一条SQL,去掉重复数据

[wxylvmnn]

表T的数据如下，
A B
-------
A1 B1
A1 B2
A2 B1
A3 B2

希望得到
A B
--------
A1 B1
A2 B2
A3

[长垣小康]

select distict（*） from T

[--小F--]

select * from tb t where not exists(select 1 from tb where a=t.a and b<t.b)

[Me_zzx]

 select a,min(b)  b from tb group by a

[wxylvmnn]

都不对啊。。。
结果变成了
A1 B1
A2 B2
A3 B1
想要的是
A1 B1
A2 B2
A3

[十三门徒]

你给的数据不对吧 你看看你给的数据是怎么个重复法啊

[zhousq00]

你这个没有一个去除规则啊

[开着拖拉机泡妞]

select min(A) as A,isnull(min(B),'') AS B from tbl

[夏小笨]

select distinct a,b from tb

[--小F--]

;with f as
(
select id=row_number()over(partition by a order by getdate()),* from tb
)
select
  isnull(a.a,'') as a,isnull(b.b,'') as b
from
(select px=row_number()over(order by getdate()),* from (select distinct id,a from f where id=1)t)a
left join
 (select px=row_number()over(order by getdate()),* from (select distinct id,b from f where id=1)t)b
on
 a.px=b.px

[libystu]

有点难度啊

[libystu]

利用游标好了

[tim_spac_126]

create table T (a varchar(16), b varchar(16))
insert into T
select 'A1','B1' union
select 'A1','B2' union
select 'A2','B1' union
select 'A3','B2'

declare @tb_a table (id int identity(1,1), ch varchar(16))
declare @tb_b table (id int identity(1,1), ch varchar(16))

insert into @tb_a
select distinct a 
from T 
order by a

insert into @tb_b
select distinct b
from T 
order by b

select i.id, a.ch, b.ch
from (select id from @tb_a union select id from @tb_b) as i
left join @tb_a as a on a.id = i.id
left join @tb_b as b on b.id = i.id

drop table T


-- id	ch	ch
-- 1	A1	B1
-- 2	A2	B2
-- 3	A3	

[tim_spac_126]

create table T (a varchar(16), b varchar(16))
insert into T
select 'A1','B1' union
select 'A1','B2' union
select 'A2','B1' union
select 'A3','B2'

declare @tb_a table (id int identity(1,1), ch varchar(16))
declare @tb_b table (id int identity(1,1), ch varchar(16))

insert into @tb_a
select distinct a 
from T 
order by a

insert into @tb_b
select distinct b
from T 
order by b

select i.id, a.ch, b.ch
from (select id from @tb_a union select id from @tb_b) as i
left join @tb_a as a on a.id = i.id
left join @tb_b as b on b.id = i.id

drop table T


-- id	ch	ch
-- 1	A1	B1
-- 2	A2	B2
-- 3	A3	

[tim_spac_126]

create table T (a varchar(16), b varchar(16))
insert into T
select 'A1','B1' union
select 'A1','B2' union
select 'A2','B1' union
select 'A3','B2'

declare @tb_a table (id int identity(1,1), ch varchar(16))
declare @tb_b table (id int identity(1,1), ch varchar(16))

insert into @tb_a
select distinct a 
from T 
order by a

insert into @tb_b
select distinct b
from T 
order by b

select i.id, a.ch, b.ch
from (select id from @tb_a union select id from @tb_b) as i
left join @tb_a as a on a.id = i.id
left join @tb_b as b on b.id = i.id

drop table T


-- id	ch	ch
-- 1	A1	B1
-- 2	A2	B2
-- 3	A3	

[anzhiqiang_touzi]

[Quote=引用 15 楼 的回复:]
SQL code

create table T (a varchar(16), b varchar(16))
insert into T
select 'A1','B1' union
select 'A1','B2' union
select 'A2','B1' union
select 'A3','B2'

declare @tb_a table (id int identi……
[/Quote]