Filter options: merge tables and bi directional relationship

[Anonymous]

Hi, 

I have several filters coming from dimension tables and it should only show values in the fact table. 

To make this possible I have to set the directions of relationship  to "BOTH" or I would have to merge all dimension tables into the fact table and use the columns in the flattened fact table as filter.

Conventional wisdom of datamodelling says that you should not flattening tables (merge option) and avoid BI directional relationship. 

So which of the "bad" option is the best, or is there another way to filter fact values??

[Anonymous]

Hi,

Instead of doing any of this, Might I suggest that you do a DISTINCT(Fact[Diemension_Column]). You can use this as the Diemension Table with a Single Directional Relationship flowing from the Diemension Table to the Fact Table.

[Anonymous]

@Anonymous  thx for your reply.

I have thought about this and this off course would also do the trick. 

I wonder which method has a better performance: getting dimensiontable from database or using DAX to make dimension table.

[Anonymous · Accepted solution]

Hi,

I think you should consider using DAX and Creating a calculated Diemension Table.

Do "Accept as Solution" if this solves your Issue 🙂