Snappy title, huh?
This post is a quick note on a powerful but hopefully non-obvious feature of MDX against Essbase, inspired by an interesting thread on the Network54 Essbase board. That discussion is about implementing custom groupings from the reporting layer – i.e. being able to retrieve roll-ups that don’t exist in the underlying cube. For example, the ASOsamp.Sample Products dimension looks like this:
Suppose that you want to do some reporting on the total of [Personal Electronics] and [Home Entertainment], for which there is no roll-up in the cube. MDX can do the summing, of course. But won’t that mess up any ratio- or variance-type measures, by adding the values for [Personal Electronics] and [Home Entertainment] together? As it turns out, the answer is ‘not necessarily’. Continue reading