For Reference:
Fact Table: Encounters, primary key: EncID
Cross Reference Table : Secondary Diagnosis, primary Key: EncId, DxId
Dimension: Diagnosis (DxId)
An encounter can have many Secondary Diagnosis, and a Secondary Diagnosis can belong to many different encounters. Not all encounters will have a secondary diagnosis, and not all secondary diagnosis will have an encounter.
So basically datamodel in the mart looks like this
Encounter (EncId) ------> Secondary Diagnosis (EncID, DxID) <-------- Diagnosis (DxId)
In the cube to model this using a Many to Many relationship I created the Encounters Dim modeled below
Encounter Fact -----------> Encounters Dim <-------- Secondary Diagnosis (Fact) ----------> Diagnosis
When I process the cube, I get server processing errors stating the EncID XXXX is not found in Secondary Diagnosis fact, which is a true statement since not all encounters will have a secondary diagnosis. Is there any way that I can model this so I do not get the processing errors. The "Encounters Dim" I limited to only encounter ids that do have a secondary diagnonis, but that did not work. Not sure what my best option is, or even if I have any.
THanks for your help