1 Reply Latest reply on Jul 18, 2016 7:50 AM by Russ Lyman

    Pros and Cons of Data Structure

    Ashutosh Garg

      I'm getting data ready for a dashboard and I am confused about how to best structure the data.  The dashboard would be updated monthly.


      Method 1) The first method would be to have all the months data in one row. I lean towards this structure because it's direct; everything I need is in 1 row; it doesn't take up too many rows.


      Month     Measure1     Measure2 ...      Measure30

      July          20                    30                         40


      Method 2) This would be the second method, which someone else showed me and from the looks of it, it may be more "Tableau friendly"; however, it's more row.


      Month     Measure#     Value

      July         Measure1      20

      July         Measure2     30


      July         Measure30     40


      I think preparing the data is easier for Method 1, but I'm a bit restricted in Tableu (Calculations, types of graphs). Where as Method 2 requires a bit more effort, takes up plenty of rows.


      What is best practice?

        • 1. Re: Pros and Cons of Data Structure
          Russ Lyman

          Generally speaking, Tableau prefers the data to be skinny, as opposed to wide. Option #2 would typically work better in Tableau. In your case, as long as the measure# represents the same metric (i.e., M1, M2, M3 all = Sales), option #2 is better. If each measure # should represent a different field (i.e., Profit, Sales, Discount, etc.), then shouldn't pivot the data and you would need to go with option #1.