1 Reply Latest reply on Oct 21, 2016 2:56 PM by Patrick Van Der Hyde

    Plot Distribution from mean and variance only

    Brian Moore

      I have a client that is processing web log data in realtime to a certain aggregated interval (e.g. hour or minute) such that their database table appears similar to the below:


      date_time     metric_name     metric_sum     count     min     max     mean     variance

      1/1/1970 00:00     revenue     145000.00     2343     6.39     370.94     61.89     43.22222

      1/1/1970 01:00     revenue     116000.00     1843     8.99     540.60     62.94     54.11117


      My understanding is that Tableau does not innately plot histograms or distributions without the raw underlying data. Feasibly, that is not an option for this client as they are producing tens of millions of rows of data per day and need to take advantage of the pre-processing to output statistical summary data.


      Can Tableau use this type of data? What modifications could I make to the real-time processor to reduce the rows of the original data set while making the output summary data accessible to Tableau for plotting distribution?

        • 1. Re: Plot Distribution from mean and variance only
          Patrick Van Der Hyde

          Hello Brian,

          Thank you for submitting your question to the forums. 


          As I understand here, the question relates to an underlying data that is being created in log files and then aggregated into a move compact data set from this data at a periodic frequency? 

          What is the data destination for the aggregate data?  Tableau connects to most data sources even statistical packages such as R. 


          What have you tried so far and are you running into specific issues?