2 Replies Latest reply on Sep 19, 2016 7:53 AM by Rick Savoia

    Machine states over time....must be missing something really basic

    Rick Savoia



      I must be doing something really basically wrong here but I was trying to take a spreadsheet that had machine states tracked on a day by day basis over time.  I have it in excel and it looks something like this:


      Sep 1
      Sep 2
      Sep 3
      Sep 4

      Sep 5

      Machine 1GGYYG
      Machine 2RRRGG

      Machine 3

      Machine 4RRRRR



      This goes across 9 months from January to September.


      What I wanted to do was use Tableau's functionality to show breakdowns of states by Quarter, Month, Week, etc. 


      I imported the spreadsheet but all the dates got listed as dimensions unlike other spreadsheets where I've been working with multiple dates.


      Is this because the dates on my source data are individual columns instead of a single column of dates?


      In Excel, I could do a 'countif(range, "G")' and it would tell me how many days the systems were green (ready to use), etc.


      I'd like that to be more dynamic so I can compare against other projects.


      Any ideas?



        • 1. Re: Machine states over time....must be missing something really basic
          Sagar Agarwal

          Hi Rick,


          This is expected behavior. For a datasource such as Excel, Tableau treats every column in it either as a Dimension or Measure. You will have to probably prepare / massage the data at the source before you consume in Tableau. How big / small the data file is really?

          • 2. Re: Machine states over time....must be missing something really basic
            Rick Savoia

            Hi Sagar,


            The data file isn't all that big.   It's just long. It's tracking machine states from Jan 2016 to today so about 9 months of daily data.


            This is only across about 15 machines though.


            Hmm...I'm going to have to think about this.  I really want to be able to show:

            • Daily status of the machines (Functioning/Degraded/Blocked) on a daily basis
            • Status of the machines on a sprint boundary
            • Status of the machines on a quarterly boundary

            Essentially, the argument I would be making is:  "If machine states are degraded, then progress is impacted.  Our highest priority is to make sure machine states are not impacted by the introduction of new code".


            Right now, I get the argument that the team owning the machines needs to do more to work around problems but their primary purpose is not to workaround problems but validate scenario functionality. 


            Any other suggestions would be helpful.  Thanks for responding with the information you have though. It does clear up the behavior question on Tableau/Excel