SQL Server Performance Forum – Threads Archive

Year over Year analysis

Hi,
I am new to analysis services, and I have to say I am in love so far. I bought a book and have walked through most of the examples using excel as the front end. My quesiton is this – we do alot of year over year analysis. Basically, I would like to take my sales cube and link it to itself on each dimension, except have the time dimension staggered one year so I can create a year over year change measure. Off the top of my head I can think of a few ways to do this, but nothing very efficient. Is there a way to do this gracefully? Thanks,
Eric

---

Let me get this strait, you want to create cube that that will give you history trends and comparisons from last year? Why do you need to link it to itself can you explain this a little better please?? You can create a Calculated Member using MDX expressions to minipulate your data for these sorts of trends, one that just came to mind is the ParallelPeriod function. Raulie
Hewlett-Packard

---

Thanks Raulie, I will clarify. The sales cube I have has time, store, and item dimensions, and measures for total price, total cost, and total quantity. I would like to easily have a measure as well of the year over year % change and actual change built right into the cube this way I dont have to do it in excel. I think using the parallelperiod funciton is a good idea, I was also thinking about some kind of lookup, hgowever that seems like something that would be very processor intensive… Thanks for the help,
Eric

---

]]>