We have an issue where we have student Majors and Major Reporting categories. The problem is that these items have a 1.N <--> 1.N relationship. Currently on the back end the database has these items stored together in the same view (as our FM model source) and there is no unique key defined for each reporting category (it has to be a combination of the Major_Code + Reporting_Category). With this being the case can determinants be used to correct the granularity. I have attempted this by defining Major as the first determinant with Major_Code as the key and then Reporting Category as my second determinant with a derived key of (Major_Code + Reporting Category). I was able to compile this into my model on the Major dimension and successfully publish the package. However, the report I built returned the same exact GPA value I was getting in the report prior to incorporating determinants.
I have looked at the documentation from IBM and found it to be less than useful. Can anyone recommend better documentation source or provide personal experience using determinants.