Archives of the TeradataForum

Message Posted: Wed, 03 Jun 2009 @ 09:15:27 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Chaim

This may give you an idea of how to approach your problem

regards

david pracy

     ct
     djp.test_table ( name char(20), department char(20)) insert into djp.test_table values
     ('John','IT'); insert into djp.test_table values ('Steven','IT'); insert into
     djp.test_table values ('Susan','IT'); insert into djp.test_table values ('Paul','HR');
     insert into djp.test_table values ('Pauline','HR'); sel
     * from djp.test_table;
     Sel
     * from ( Sel department
     ,
     name
     ,
     ROW_NUMBER() OVER (PARTITION BY department ORDER BY name desc) rank_asc from
     djp.test_table )DT_1 where rank_asc = 1; /
     /* Obviously I would be expecting to order by something more meaningful then a name eg a
     hire date etc */
     /*

     Name Department
     John IT
     Steven IT
     Susan IT
     Paul HR
     Pauline HR

     Would provide the following two rows

     Susan IT
     Pauline HR*/

	<Prev	Next>		<<First	<Prev	Next>	Last>>