Hello, everybody.
We have a cube with lot's of measures that need to be aggregated normally
along all dimensions except time. Along the time dimension every new member
is displaying a new state of the measure at that moment so we'd like to
calculate the difference between every current member and the previous
member. This can be done quite easily by creating a calculated member for
every measure as in:

c1: ([measures].m1 - [measures].m1,[time].currentmember.prevmember)

c2: ([measures].m2 - [measures].m2,[time].currentmember.prevmember)

....

Can this be done some other way or do we really have to create 50+ calculate
members?

Thanks in advance,
Koni.

If you place calc(s) in dimension other than measures (i.e. a special calc
dimension) then they operate agains all measures -- one calc can do the work
of N calcs.

fwiw, I'm not at all partial to cubes that have lots of measures. Whenever
possible I look for opportunities to reduce the number of measures by adding
new dimensions. Ideally I like just one measure. This is because the
measures dim is so inflexible.

public @ the domain below
www.tomchester.net

"Koni Kogan" <k_kogan (AT) hotmail (DOT) com> wrote