Can you merge filters without actually combining the data sets and duplicating rows?

[Anonymous]

Hi,

I have 2 different data sources with different data, but they both contains Month Year (ex. January 2024). My questions is is there a way to data model where I can join both data sources on 1 "month" table and use that middle-man table to filter on both data sources with 1 month filter?

Thanks!

[Anonymous]

Hi @lbendlin,

I think I was able to do it. However, I'm getting yelled at when trying to cross join to multiple tables based off that one calendar table. I use the both cross-filter direction with many to 1 for each. When I try to do the same thing and make all relationships active and also add another dummy join table just with my page names, I get an argument error. I don't understand why I have to make them all single cross filter direction for it to work.

[lbendlin]

There is no need for cross joins. Control both fact tables from the one calendar dimension table. No need to join the fact tables with each other.

[Anonymous]

Hi @lbendlin 

You have to pick a cross-filte direction.

[lbendlin · Accepted solution]

the default is 1:* Single from Dimension to Fact.  All other options are sub-optimal.

[lbendlin]

You mean, create a calendar dimension table that then controls both these fact tables?

Yes, you can. And you should.