Hi all,

I have a fact table with columns:

Order Number OrderLine ProductCode PrgDisc Discount Quantity
001 1 ABC 1
3% 100
001 1 ABC 2
5% 100
001 2 XYZ 1
3% 200
001 2 XYZ 2
8% 200
002 1 ABC 1
3% 300
002 1 ABC 2
5% 300
002 1 ABC 3
2% 300
002 2 HHH 1
5% 150

For each order line I have many records, one for each distinct discount
applied.

I need dimensions ProductCode and Discount. The measure is obviously
quantity.

I can't sum quantity with simple aggregation by product code.

Can anyone help me?

Thanks,

Enrica

Please clarify why you need to sum the %. Normally that doesn't make sense.
Also, you can pick max or min as aggregation function. But I will still doubt
the semantics.

James

"Enrica" wrote:

Discount is a dimension, I don't need to sum %, I need to sum quantity sold
with same discount ('3%', '5%' etc are used as discount codes).
The standard aggregation with dimension discount works fine. My problem is
the aggregation with dimension Product, or other dimensions I'll add to the
cube (customer, employee who process order....)


"James Ma" <JamesMa (AT) discussions (DOT) microsoft.com> ha scritto nel messaggio
news:C76D86E0-47E4-4B62-BC80-04E07652B03B (AT) microsoft (DOT) com...

It sounds like the discount is a dimension. If you have discrete discount
values(ie a relatively small number of actual discounts - rather than a
continiously variable values) then you can model discount as a dimension
quite simply.

Alternatively if discount is continiously variable, you might want to model
ranges (0-3,3-5,5-10...)

hth

DG

"Enrica" wrote: