I built a demographics cube based on our client data and this is the first analysis cube I have created and is in production, so forgive my ignorance about this question. We ran an analysis the other day on fiscal year end data and the numbers were not matching up. So I drilled the data down further going to the granular level of the actual client. What I found is that some clients were being counted twice because their "population identifier" changed during the fiscal year. For instance they could have been classified as a minor and then changed to an adult. Now the one client is being counted twice. I remember reading somewhere online about the "SCD" or slowly changing dimension. Is this scenario what that is referring to? I started to build a query using the ROW_NUMBER to get the latest data based on when the query is run and it works, but in the business process I am not sure where to include it. Should it be a named query and a separate dimension? I am getting confused on this issue and any help I would appreciate. If someone can point me in the right direction. Thanks
↧