Hi.
I'm new to this forum, and have limited experience with MDX.
I have a cube with customer accounts over time (monthly statements). The Monthly Statement dimension has attributes indicating the status of an account (new, still active, closed). The fact table is linked to the monthly statement dimension, as well as a "Date Loaded" dimension, and a "Statement Date" dimension, and a "Customer" dimension.
I need to produce a matrix with the statement date months on the x-axis, date loaded months on the y-matrix, and showing the number of customers who had at least one new account on "date loaded", and have at least one open account at a "statement date".
The answer basically stipulates which customers I took on (new) and compared to that number, which customers I still have with my product.
I have tried intersect, filter, and exists, but I can't get it to work - I keep getting "different dimensionality" errors. I'm not sure if this is a cube design problem or mdx problem; the granularity of the cube is account, but I have also tried a second fact table "customer status" (rolling up accounts to customer level) but can't get it to work.
Thank you for any help.