1 Reply Latest reply on Aug 3, 2018 5:49 AM by Michael Hesser

    Non-Normal formatting question

    David Appleby

      this is with desktop 10.3.2

       

      hi all.

       

      just looking for a pointer.

       

      I have a table that show numbers by day of month.

       

      What I need to do is highlight values where they are x% greater than the average of the row

      excluding nulls.

       

      I can Heat Map it, but, that gives a very muddled display as the outliers are wildly off from

      the main data (Average of say 50 with outliers in the 500+ range, order of magnitude stuff)

       

       

      However, it is a large table, and, hard to spot, hence the highlight.

       

      Any help appreciated.

       

      Regards

       

      DaveA

        • 1. Re: Non-Normal formatting question
          Michael Hesser

          Hi David;

           

          Could you share a sample of your data? It's difficult to make a prognosis without seeing the patient

           

          If you're looking for row-level data, I'm thinking one or two LOD calculations are needed.

           

          When it comes to outliers that throw everything off, normalizing those excessive values is needed. You can do this quite a few ways:

          • You can format the color, adding an End Value. Your huge values won't stretch the continuous color range.
          • You can modify the color value by taking the log, or the square root, etc. While this may not work seamlessly with negative numbers and 0, it will keep your highlight color as a function of the difference.
          • You can manually evaluate the color value and reset any above a certain point.
          • You can manually evaluate the color value and place them into discrete categories. This method works okay if you have just a handful of "steps" you want to show.

           

          Hope there may be a glimmer of an idea in there for you! Good luck! --Michael