Question

I am trying to create a query that will generate the following table.
I know the column names (that is start and end date of each column. These dates will vary based on the month/year. Also the number of columns will vary since some month may have 6 weeks).

The number of pools may also vary. There can be ‘n’ number of pools. So I have create a query for each pool(row) and for each week(column).
Assume there are 3 pools(3 rows) and 5 weeks (5 columns), I may need to run 3*5=15 queries.
Is there any way to create a single query to fix this?
Can anyone help me in fixing this?
I am stiil working on this..

--------------------------------------------------
row 0 | week1 | week2 | week3  | week4  | week5  |
st-end|jan1 -2|jan 3-9|jan10-17|jan18-24|jan25-31|
--------------------------------------------------
pool a| 100   | 105   | 309    | 345    | 234    |
pool b| 106   | 108   | 399    | 245    | 214    |
pool c| 120   | 145   | 359    | 365    | 274    |
--------------------------------------------------  

Answer 1

Have you tried to do it using CASE expressions? [See “Control Flow Functions”]
Something along these lines:

select POOL
      ,sum(case when dt between 20120101 and 20120107 then VAL else 0) as Wk1
      ,sum(case when dt between 20120108 and 20120114 then VAL else 0) as Wk2
    ...  ...
from MY_TABLE
group by POOL