Beneficiary Model - Fact Table Does Not Properly Filter Beneficiary Table

Hey WinterMist ,

I assume you are looking for this:

To achieve the above, I "simply" changed the cross-filter direction of the relationship between the "fact policy" and the "beneficiary" table from "Single" to "Both."

The powers of a measure are manifold; everything we love about measures boils down to reaping insights from our data. But there is something to measures that I consider to be on the darker side of the super powers. This is the creation of relationships, even if they do not exist within our models, e.g., the relationship between: "Insured 1" and beneficiary "Sid." The "bond" that ties these objects together is "Corn" (here, it's the fact policy key 17).

First, you find the two beneficiaries, "Sam" and "Samantha" (tied to "Insurer 1"), and then you ask what the crop is. It becomes necessary to transform the table into a "kind of" dimension table to keep the beneficiaries. This can be achieved by changing the cross-filter direction.

Keep in mind that there are a lot of people out there who consider having a cross-filter of value both:

• bad practice
• a bad data model
• evidence of poor understanding

I consider having relationships with cross-filter direction "Both" a subtle dark art, like the 5-fingers death punch.

Hopefully, this will help you find what you are looking for.

Regards,

Tom

P.S.: there is no name for a table filtered by a fact table, but I consider the beneficiary table not a fact table. I am not sure if you know the book "(The Complete Reference) Star Schema" by Christopher Adamson. I consider this book essential.