Need Desired Output

Question

HELLO EVERYONE,
PLZ HELP ME WITH THE BELOW QUERY.
THE BELOW CONDTION SHOULD BE APPLIED FOR THE DATA
CASE COUNT (MAX(ADATE) )
               WHEN 1 THEN '15'
               WHEN 2 THEN '30'
               WHEN 3 THEN '60'
               WHEN 4 THEN '120'
END
_SAMPLE OUTPUT_
V104       30
V105       15
V107       60
V249       120
V272      15
TABLE T1 AS (AID,ANO,ADATE)
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664604, 'V104',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664603, 'V104',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664602, 'V104',  TO_Date( '03/13/2012 07:34:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664601, 'V104',  TO_Date( '03/13/2012 07:34:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664604, 'V105',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664603, 'V105',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664602, 'V105',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664601, 'V105',  TO_Date( '03/25/2012 01:19:00 PM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664604, 'V107',  TO_Date( '12/31/2011 10:20:45 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664603, 'V107',  TO_Date( '03/24/2012 01:38:00 PM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664602, 'V107',  TO_Date( '03/24/2012 01:38:00 PM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664601, 'V107',  TO_Date( '03/24/2012 01:38:00 PM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664604, 'V249',  TO_Date( '03/03/2012 07:09:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664603, 'V249',  TO_Date( '03/03/2012 07:09:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664602, 'V249',  TO_Date( '03/03/2012 07:09:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664601, 'V249',  TO_Date( '03/03/2012 07:09:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664598, 'V272',  TO_Date( '12/31/2011 10:20:44 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664597, 'V272',  TO_Date( '02/19/2012 08:06:00 AM', 'MM/DD/YYYY HH:MI:SS AM'));
INSERT INTO T1 ( AID, ANO, ADATE ) VALUES (
664596, 'V272',  TO_Date( '03/25/2012 02:04:00 PM', 'MM/DD/YYYY HH:MI:SS AM'));
COMMIT;

Answer 1

Hi,
H.O wrote:
... Well, thanks. That would work. however, my table t1 is not a simple one. can this be achieved by analytical fn's in a simple way, to avoid using the table t1 twice?Instead of hiding some of your requirements, then waiting for a unsuitable answer and then complaining about it, consider giving all of your requirements in your first message.
Here's a way to get the same results only referencing the t1 table once:
WITH     got_r_num     AS
     SELECT       ano
     ,       adate
     ,       CASE  COUNT (*)
                      WHEN  1  THEN   15
                      WHEN  2  THEN   30
                      WHEN  3  THEN   60
                      WHEN  4  THEN  120
            END               AS cnt
     ,       RANK () OVER ( PARTITION BY  ano
                             ORDER BY      adate  DESC  NULLS LAST
                      )      AS r_num
     FROM      t1
     GROUP BY  ano
     ,            adate
SELECT    ano
,       adate          -- if wanted
,       cnt
FROM       got_r_num
WHERE       r_num     = 1
ORDER BY  ano
;It's unclear if this will be faster or slower than what I originally posted. Even though only one scan of the t1 table is required, the analytic function means that the result set of the GROUP BY has to be scanned, which could be even slower.