去重复的问题，急求。

[TLJewel]

我建了一张表，里边包括 ID，时间和 TAG,表里很多TAG和ID是重复的，现在我想去除TAG的重复，但要获取最新时间的ID，语句该怎么写？

[新鲜鱼排]

create table #t(tag int, _date datetime)
insert #t select 1, '2006-06-07'
union all select 1,'2007-02-03'
union all select 1,'2007-02-04'
union all select 2,'2007-02-03'

delete from #t
where convert(varchar,tag) + '_' + convert(varchar,_date ,120)
not in (select convert(varchar,tag) + '_' + convert(varchar,max(_date),120) from #t group by tag)

select * from #t

/*
tag _date
----------- ------------------------------------------------------
1 2007-02-04 00:00:00.000
2 2007-02-03 00:00:00.000
*/

drop table #t

[新鲜鱼排]

http://community.csdn.net/Expert/topic/5073/5073353.xml?temp=.7302973

[XJY123]

select identity(int,1,1) as autoID, * into #Tmp from tableName
select min(autoID) as autoID into #Tmp2 from #Tmp group by ID,autoID
select * from #Tmp where autoID in(select autoID from #tmp2)
最后一个select即得到了ID，Tag不重复的结果集（但多了一个autoID字段，实际写时可以写在select子句中省去此列）

[TLJewel]

谢谢大家，正在试

[TLJewel]

已经搞定了，2楼的朋友帮了很大的忙，谢谢啊。