Reasons why a slicer may not eliminate rows

[DIFF1]

 Hello everyone, 

 I have such a situation, namely:

 I have a slicer with names of different stores.

And two visuals where I track forecast accuracy. 

I want to move ahead and just delete that Location column from the visual, but before I do that I need to understand how I can eliminate Rite Aid when I select "Walmart" from a slicer. 

I know many people would say that the issue is the relationship.  I have 2 relationships between the tables that I use 

table1 from where comes the slicer and Actuals/Forecast - data about actuals/forecast. 

Relationships are many to many because one unit is forecast and sold in multiple locations will be appearing in different locations. 

Does someone may have an idea what could be the reason why the slicer keeps not eliminating rows ?

Any information will be greatly appreciated. 

[v-stephen-msft · Accepted solution]

Hi @DIFF1 ,

I created some dummy data to restore your problem.

The year and year columns create an active relationship, and the LOC and LOC columns create an inactive relationship.

The LOC column in the slicer is from the 'Table' table. The LOC column, YEAR column and Sales column in the table visual are all from 'Table(2)' table.

The problem is that your LOC column relationship is not active, please refer to Create and manage relationships in Power BI Desktop - Power BI | Microsoft Docs. An inactive relationship means that the relationship is not enabled, and the filter will not have the effect of the relationship.

If the LOC column in the Table table is put into the table visual, the RITE AD can be filtered out.

Best Regards,

Stephen Tao

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[bcdobbs]

The problem is that the relationship between location is not active. You can only have one active relationship between two tables.

I would create a locations dimension table with a distinct list of locations (use either power query or dax) which can then be related to both tables via a 1 to many relationship.

However I think that will bring ambiguity into your model because of the other bidirectional relationship. Can you give more description about what the other relationship does?