8 Replies Latest reply on May 3, 2017 3:52 AM by Tjerk Boorsma

    Double data in one period

    Tjerk Boorsma

      Hi all,

       

      Did anyone ever come across a situation where Tableau counts data double?

       

      This is data from a survey and the period January 2016 seems to have double the sample size as other months. In fact, filtered on one brand the data for January 2016 looks like a duplicate from November 2015, and the measure is the exact same. The most obvious reason would be that it must be in the data but this is not it, at least there is nothing to be found.

       

      There are 2 screenshots attached: The first screenshot is taken when filtered on one brand and two measures (Drivers - Unweighted); these two measures look identical for this brand. The second screenshot is for all brands combined and include all measures; there are no more identical measures and January 2016 no longer has exactly double sample from November 2015 but close to double.

       

      Sum(Number of Records) is the sample size and these are correct for all periods apart from January 2016. The measures for the other periods are also fine.

      Data is stacked on lowest respondent level and all other calculations and measures are working as it should.

      There is a Union in the workbook that combines all periods from different Excel worksheets. Working on an extract in Tableau Desktop 10.2.

       

      I'm very curious about anything that could have caused Tableau to count something double.

       

      Many thanks, Tjerk