>

>select * from table1 join (select id, max(qty) as qty from table1 group by id) table2 on table1.qty = table2.qty
>

>
>you need to be sure that qty is unique per names; otherwise you'll need another sub-query on max(date) or whatever.
>
>
>>Sir I have these codes
>>
>>

>>	CREATE CURSOR table1 (id int,names c(10),date d,qty int)
>>	INSERT INTO table1 VALUES (1,"ADMIN",{^2017-02-19},25)
>>	INSERT INTO table1 VALUES (1,"ADMIN",{^2017-02-20},21)
>>	
>>	INSERT INTO table1 VALUES (2,"GENERAL",{^2017-02-17},29)
>>	INSERT INTO table1 VALUES (2,"GENERAL",{^2017-02-21},14)
>>	
>>	SELECT id,MAX(names),MAX(date),MAX(qty)from table1 GROUP BY id
>>
>>	brow
>>

>>
>>it show these wrong result
>>
>>1 ADMIN 2/20/2017 25
>>2 GENERAL 2/21/2017 29
>>
>>But I need this
>>
>>1 ADMIN 2/19/2017 25
>>2 GENERAL 2/17/2017 29
>>
>>The date that has greater qty
>>
>>Later on I want to use these codes in SQLSERVER 2008
>>
>>Please help


Sir, your query is displaying strange result

id_a names date qty_a id_b qty_b
1 ADMIN 2/19/2017 25 1 25
2 GENERAL 2/17/2017 29 2 29