4 Replies Latest reply on Aug 30, 2016 2:12 AM by Jonathan-Raphael Reichert

    How to create customized filters from (calculating) multiple measures

    Jonathan-Raphael Reichert

      I am having trouble creating customized filters, which may be used for reporting.



      My dataset consists of several rows/text documents (observations) originating from multiple sources and domains from different demographic locations (forums discussions, tweets, articles etc.)

      The texts have been coded via a text analytical software (QDA miner/wordstats) and extracted as excel doc. Each row (each text) has several columns (variables), with either 0 or 1, where 1 denotes that the category is present for that observations (see below).


      The variables are all listed as columns without being divided into hierarchies/categories. My hope is to create categories (calculations) of several variables in Tableau, and use those categories as filters.


      1. How can I create several categories of measures, which I can use as a filter afterwards? I am interested in creating several different categories that I can use as a set of categories filter.

      - Calculating measures together as a sum is no problem. However, in order to use it as a filter, I have tried to transform the new calculated category into a dimension and use the filter function to show either how many 0 observations and 1 observations there are. My main problem here is that for each category that I calculate i get "x amount of 0", "x amount of 1" and "x amount of 2". Ideally I would only have all the ones that are flagged as 1 across all the variables chosen for that category.

      - The hand drawing below depicts how I want it to be. In the example showing 3 calculated categories, that can be used to filter on and off to change and thus change the domain distribution.




      A workbook with test data has been attached. This data is a simplification of what I want to end out with, but I hope it is sufficient for the purpose.


      Please let me know if anything is unclear!