Running total or something like that !

[V i n c e n t]

Hi,
I need your help to solve my problem, can someone help me ?

I want that

FACT MONTH | CUSTOMER+PRODUCT LINE | CUMULATIVE CUSTOMER + PRODUCT LINE
-----------------------------------------------------------------------
     1     |        16 016         |              16 016
-----------------------------------------------------------------------
     2     |        15 302         |              22 432
-----------------------------------------------------------------------
     3     |        16 837         |              27 883 
-----------------------------------------------------------------------
     4     |        15 347         |              31 382
-----------------------------------------------------------------------

and i have that

FACT MONTH | CUSTOMER+PRODUCT LINE | CUMULATIVE CUSTOMER + PRODUCT LINE
-----------------------------------------------------------------------
     1     |        16 016         |              16 016
-----------------------------------------------------------------------
     2     |        15 302         |              31 318
-----------------------------------------------------------------------
     3     |        16 837         |              48 155 
-----------------------------------------------------------------------
     4     |        15 347         |              63 502
-----------------------------------------------------------------------

I'm working with a data warehouse (relational) and i want to do a cumulative total month by month of couples customers+ product line without recounting the couples already counted in previous months.
The running-total function do the cumulative total of any month with recount.
I try the distinct fonction but dont work.

Any idéa ? can someone help me ?

Thanks

[blom0344]

You could solve this by adding an additional query that determines the highest value of month for the combination  of customer/productline  from the datawarehouse.

Use this query in a join with the orginal one , on customer/productline and month to retain only the combinations with the highest month.

Use the output of this joined set as the input for a query that computes the cumulatives.

Instead of month a better solution would to use the period , like in YYYYMM  , say  201206.  Month itself is not unique , which may cause problems if you need data transcending years.