Forum Discussion

sbotros's avatar
sbotros
Frequent Visitor
2 years ago
Solved

Star schema filtering in visuals

Hello all - looking for some quick troubleshooting/second pair of eyes as I've been staring at the problem for too long at this point. 

 

After previously using a single table schema for individual sets of data, I'm working on adapting to a star schema that contains all relevant tables and columns for any sets of questionnaire data (see below). The idea is that this structure and report could be used with any given set of data, as the fields are normalized. 

 

I am having trouble with visualizations behaving as expected. For example, I have chart 'C12' and 'C7' using X: Responses.OptionID and Y: Responses.RespondentID (distinct count). Each visual is filtered to use Questions.QuestionID with the respective question selected. However, this results in no interaction between the two visuals when anything is selected, the data either disappears in the other table or opens the table view.

 

I've tried countless other methods including using measures, using the dimension tables as the data sources, changing the cardinality of the model, checked the interaction settings, etc etc etc. 

 

I think I am missing something very obvious; would appreciate any extra eyes and brains on this. Let me know if anything needs clarification. 

 

 

 

 

  • Facts: something you measure

    Dimensions: something you filter by

     

    Re-examine your data and identify which column belongs where, how many tables you need etc.

3 Replies

  • That's not a star schema - yet.  Make all joins single directional.

    • sbotros's avatar
      sbotros
      Frequent Visitor

      Thanks for the response! I had initially tried single direction joins and it did not work.

      I've also tried using only the dimension tables fields as my sources for the visuals and that also doesn't work. 

      • lbendlin's avatar
        lbendlin
        Icon for Super User rankSuper User

        Facts: something you measure

        Dimensions: something you filter by

         

        Re-examine your data and identify which column belongs where, how many tables you need etc.