Continuation: need data grouped for every week

Question

SELECT * FROM aetnah_file_emp_cust_hist
WHERE pctl_employee_seqnum= 133774This query returns too many rows.
Now my requirement is that I need to get the most recent contrib_amt from this table for every week based on date column CTL_INS_DTTM.
This column CTL_INS_DTTM stores data about when a row wass inserted into this table.
pctl_employee_seqnum CTL_INS_DTTM(MM/DD/YYYY) contrib_amt
133774 01/01/2009 100
133774 01/02/2009 200
133774 01/03/2009 300
133774 01/04/2009 400
133774 01/05/2009 500
133774 01/06/2009
133774 01/07/2009 700
133774 01/08/2009 800
133774 01/10/2009 900
133774 01/12/2009 1000
133774 01/13/2009 1100
133774 01/14/2009 1200
I will need 52 columns (1 year = 52 weeks) totally and the most recent data for each week.
Ex:
Desired output:
01/07/2009 01/14/2009 01/21/2009 01/28/2009
700 1200 NULL 200
SELECT pctl_employee_seqnum, CTL_INS_DTTM, contrib_amt
  FROM (
          SELECT pctl_employee_seqnum, CTL_INS_DTTM, contrib_amt, ROW_NUMBER() OVER(PARTITION BY TO_CHAR(CTL_INS_DTTM, 'WMONYYYY')ORDER BY CTL_INS_DTTM DESC) RNO
            FROM aetnah_file_emp_cust_hist
           WHERE pctl_employee_seqnum= 133774
WHERE rno = 1This code above is doing that but I need to transpose this row wise data into column wise as mentioned on Desired output.
Total # of columns:52
Apologize for opening a new thread.....please help on this transpose issue
Thank You All

Answer 1

Hi,
TO_CHAR (ctl_ins_dttm, 'WMONYYYY') will result in 59 or 60 groups per year, since all months (except February in common years) have (incomplete) 5th weeks.
If you want 52 equal-sized groups, then TO_CHAR (ctl_ins_dttm, 'WWYYYY') will get you closer. (You'll still have an incomplete week 53.)
To pivot those rows into one column, you can do something like:
SELECT  MAX (CASE WHEN TO_CHAR (ctl_ins_dttm, 'WW') = '01' THEN contrib_amt END)   AS week_01
,       MAX (CASE WHEN TO_CHAR (ctl_ins_dttm, 'WW') = '02' THEN contrib_amt END)   AS week_02
,       MAX (CASE WHEN TO_CHAR (ctl_ins_dttm, 'WW') = '03' THEN contrib_amt END)   AS week_03
,       ...If you want data (like "01/07/2009") as the columns headers, then you'll have to use dynamic SQL.