that would work too but 1 order could have 3 records.
OrderDate, ShipoutDate/PickupDate or CancelledDate.

On the other side, the fact table needs to be updated for different date if
it is not at the Order Type level.

What would be a better way?

"Tomasz Borawski" wrote:

> Hi,
>
> Why are not going to create dimension called Order Type with to members
> Ordered, Shipout, Pickout and Canceled? You could use COUNT to measure all
> orders and filter them if you would like to show only canceled and you do not
> need to use -1 as a value.
>
> "Ed" wrote:
>
> > Hi,
> > I am a newbie and I have been studying DW for a while and I've
> > successfully built some cubes.
> > There is still one thing I am not sure how to design it in the fact table
> > which is the COUNT. Let's say I have an Order Fact to record the order from
> > customers. Then I have OrderDate, ShipDate, CancelDate, PickupDate, and some
> > sales related measurments. How do I count how many orders were cancelled (I
> > stored -1 as a value) for CancelDate? Should I add an measurment column that
> > put a value of 1 if the order was cancelled and put 0 if the order was not
> > cancelled so that it can be aggregated in the cube? I need the same thing to
> > count the pickup since some customers pick up in the store not ship out and I
> > need to be able to generate the comparison between ShipOut, Pickup, and
> > Cancel.
> >
> > Thanks so much
> >
> > Ed
> >