Set analysis ?

[#1]

Hi !

I can t change the model, but i need to multiple in a pivot table a coefficient with using two different dimension for weeks

Is it possible ?

Thanks

Regards

[shree909]

HI

='=> ID_PRODUCT * COEF (with using Week_S) -- Possible ?'

I DONT UNDERSTANT WHY U WANT TO MULIPLY CHARACTER  B WITH COEFF

CAN U EXPLAIN CLEARLY .

WAHT U WANT TO CALULATE??

ID PRODUCT IS A CHARACTER AND COEFF IS A NUMBER SO U CANT  MULTIPLY???

[#3]

i want to multiply

count(ID_PRODUCT)  * coef !

[shree909]

TRY LIKE THIS

=COUNT(ID_PRODUCT) *SUM(Coeff)

BUT HTIS WORKS ONLY WHEN U SELECT THE COEF FROM THE LIST BOX.

TRY THE ABOVE EXPRESSION IN

[#5]

Yes but the coefficient is linked to a week and it must be dynamic

I need to do this :

=COUNT(ID_PRODUCT) * SUM(<Week_S=Week>  Coeff)

<Week_S=Week> <=> Where Week_S = Week

Is that clear ?

[#6]

Hi,

that's what I understood:

-Model unchangeable (why? maybe there is way)

-2 tables:

     -product with week_s and week with week_id

     -they are not connected (via a reference, so week_id is not an attribute in product)

-you need

     -count(id_product)*Coeff

          -for the selected Coeff

          -if week_s=week

     -0 in the else-case

Is a selection on week or what is the business relation between product and week?

Ciao

Klaus

[#7]

thanks for answer

I need this :

     -count(id_product)*Coeff

          -for the selected Coeff

          -if week_s=week

     -0 in the else-case

Is a selection on week or what is the business relation between product and week?

There is no relation between Week_s and week in the model. And there is no selection on week.

The coefficient is written in a file and there is a relation between week -Week_s and product.

In my model, i only have a relation with the product and not with the week.

But I need in my expression to have the relation between Week_s and Week to find the coefficient

Is it possible to construct a pivot table and in the set analysis to do the relation between the Week and Week_s ?

[#8]

ok....

but which weeks from week are the relevant ones for that condition (week_s=week)?

or is existence of the week_s in week sufficient for the condition?

Ciao

Klaus

[#9]

yes it's sufficient !

[#10]

so go for

where exists (week, week_s);