Filtering out rows based on field value

Question

Hello,
I have something like this:
create table stuff
avalue varchar2,
anothervalue varchar2,
thirdvalue varchar2,
anumber number
INSERT INTO stuff Values('abc', 'cde', efg', 2);
INSERT INTO stuff Values('abc', 'cde', efgggg', 4);
INSERT INTO stuff Values('abc', 'xxx', efg', 1);
INSERT INTO stuff Values('abc', 'xxx', efgxxx', 2);
INSERT INTO stuff Values('abc', 'yyy', efg', 2);
I would like to recieve the rows with the higest 'anumber'-value, grouped by 'avalue', and 'anothervalue' (but not 'thirdvalue'). So in this example I want row 2, 4 and 5.
Anyone have an idea? I'm thinking a 'order by desc' is not the best way here?

Answer 1

select  avalue,
        anothervalue,
        thirdvalue,
        anumber
  from  (
         select  s.*,
                 rank() over(partition by avalue,anothervalue order by anumber desc) rnk
           from  stuff s
  where rnk = 1
AVALUE     ANOTHERVAL THIRDVALUE    ANUMBER
abc        cde        efgggg              4
abc        xxx        efgxxx              2
abc        yyy        efg                 2
SQL> SY.