Week and month calculation from date column

Question

I have 3 column data like
with tab as
  select 'Topshop' brand, '10-JUL-11' deliverydate, '100' qty from dual union all
  select 'Topshop' brand, '10-JUL-11' deliverydate, '400' qty from dual union all
  select 'NewSita' brand, '11-JUL-11' deliverydate, '200' qty from dual union all
  select 'LaGress' brand, '12-JUL-11' deliverydate, '300' qty from dual union all
  select 'LaGress' brand, '10-AUG-11' deliverydate, '100' qty from dual union all
  select 'LaGress' brand, '11-AUG-11' deliverydate, '200' qty from dual union all
  select 'Topshop' brand, '12-AUG-11' deliverydate, '300' qty from dual union all
  select 'NewSita' brand, '10-SEP-11' deliverydate, '100' qty from dual union all
  select 'Topshop' brand, '11-SEP-11' deliverydate, '200' qty from dual union all
  select 'NewSita' brand, '12-SEP-11' deliverydate, '300' qty from dual
) select * from tabI need to convert it into 4 columns
Brand | Month | Week(start date) | Qty (sum)
Please let me know what are the options that i have, especially the date-time calculation functions available to solve such problems.
Thanks
w\

Answer 1

with tab as
  select 'Topshop' brand, to_date('10-JUL-11', 'dd-mon-yy') deliverydate, '100' qty from dual union all
  select 'Topshop' brand, to_date('10-JUL-11', 'dd-mon-yy') deliverydate, '400' qty from dual union all
  select 'NewSita' brand, to_date('11-JUL-11', 'dd-mon-yy') deliverydate, '200' qty from dual union all
  select 'LaGress' brand, to_date('12-JUL-11', 'dd-mon-yy') deliverydate, '300' qty from dual union all
  select 'LaGress' brand, to_date('10-AUG-11', 'dd-mon-yy') deliverydate, '100' qty from dual union all
  select 'LaGress' brand, to_date('11-AUG-11', 'dd-mon-yy') deliverydate, '200' qty from dual union all
  select 'Topshop' brand, to_date('12-AUG-11', 'dd-mon-yy') deliverydate, '300' qty from dual union all
  select 'NewSita' brand, to_date('10-SEP-11', 'dd-mon-yy') deliverydate, '100' qty from dual union all
  select 'Topshop' brand, to_date('11-SEP-11', 'dd-mon-yy') deliverydate, '200' qty from dual union all
  select 'NewSita' brand, to_date('12-SEP-11', 'dd-mon-yy') deliverydate, '300' qty from dual
), week_data as (
   select brand
   ,      to_char(deliverydate, 'Month', 'nls_date_language=american') MON
   ,      trunc(deliverydate, 'day') WEEK_START
   ,      sum(TO_NUMBER(qty)) SUM_QTY
   from   tab
   group by brand
   ,      to_char(deliverydate, 'Month', 'nls_date_language=american')
   ,      trunc(deliverydate, 'day')
), weeks as (
   select
   w.week_min + (level - 1) * 7 week_start
   from (
      select min(week_start) week_min, max(week_start) week_max
      from week_data
   ) w, dual
   connect by level <= 1 + (w.week_max - w.week_min) / 7
select
week_data.brand,
weeks.week_start,
nvl(week_data.sum_qty,0) sum_qty
from week_data
partition by (brand)
right outer join weeks
on (weeks.week_start = week_data.week_start)
order by
weeks.week_start,
week_data.brand;You construct a set of records of all "week_start" dates from your minimum to your maximum (classic connect by level trick.)
You use a partitioned outer join to fill in the gaps in your sparse data (see doc example [url http://download.oracle.com/docs/cd/E11882_01/server.112/e17118/statements_10002.htm#i2177515]here.
Hope that helps you :-)
(PS. I dislike implicit conversions - that is the only reason I have added a TO_NUMBER to your code within the sum() ;-) )
(PPS. Actually you probably should have asked this in a new question - I believe it is not really good forum etiquette to continue in a thread that has already been answered...)