3 Replies Latest reply on Apr 18, 2018 12:38 PM by Paul Gonzalez

    Aggregation Aggravation

    Paul Gonzalez

      As I started to compose this post it occurred to me that Tableau likes to aggregate data from two different sources, and that is the reason for my frustration. But hopefully by stating my problem here one of you knows a good work around.


      I'm building dashboards to display production metrics based on time stamps. I've found that I can display pretty nice graphs from a single data source, but when I start blending and or joining data, it aggregates in unpredictable and undesirable ways.


      For instance, this is what I want:


      And this is what I don't want, and is the result of blending server data with a local table (the excel file below):



      What I'm trying to do I connect to a table with some constant values, in this case I need median values and a constant stating that there are 24 hours in a day to display my planned dashboard correctly.  This is EASY to accomplish in Excel by adding a column and putting values into it(see below), but since I'm pulling these values from our Tableau server it isn't as straight forward.  I suppose I could work out a solution with the server admin as to how to have these values available in our server connection, but it would be so much easier if I could find a way to add constants into my Tableau calculations.




      Does anyone have any idea of how to add these constants into my data source without having it force me to aggregate?

        • 1. Re: Aggregation Aggravation
          Branden Kornell

          I think it's safe to hard-code the 24 hours/day into your calculations. That's not changing anytime soon. ;-)


          Can you join the values to your main data source on Process, or do you need the multiple different Process values available to your calculations?


          if that's the case, try pivoting the values for your constants so that, instead of rows, they appear in columns (Column A, B, C, etc., just like in Excel).


          The resulting table would be 1 row long, and you would join to your main data set on a 1=1 join. Then each row will have access to all the constant values.


          Though this seems like it's redundant (putting every value on every line), Tableau is efficient and there shouldn't be a big performance or space impact.

          • 2. Re: Aggregation Aggravation
            Shinichiro Murakami

            Hi Paul,


            It's quite difficult to understand your issue without seeing the data.

            Could you attach sample sample data and expected result as illustration.




            • 3. Re: Aggregation Aggravation
              Paul Gonzalez

              Sorry to both of you who responded, I didn't get back right away because I was off for a few days.


              Thank you Branden, that is a great idea. I'm also corresponding with my Tab Server Admin about how to handle this too.  But I think that your idea is perfect for proof of concept. 


              Shin, I'll try to mock something up.