6 Replies Latest reply on Jan 14, 2017 1:50 AM by Yuriy Fal

    Distinct Head Counts Between 2 Dates without Running Totals (V.10)

    David Rudden

      Hello everyone,


      I apologize if this is a question that has already been posed and answered, but I haven't been able to find a solution that will work quite like I need it to for the data that I'm working with.  I'm looking to display a unique/distinct count of students enrolled on each given day, displayed over a period of time.  There have already been a number of discussion examples that would have helped  (Show Records That Fall Within a Period of Time | Tableau Software and Re: How to get counts of active enrollments (active between two dates) have been especially helpful), but I am running into issues because each variation of these solutions seems to include a RUNNING_TOTAL calculation that ends up displaying inflated counts with the data that I'm working with. 


      Each row in my data represents an individual enrollment transaction, with a date that a student enrolled in a particular class (add date) and a corresponding date that the class was dropped or they completed it (in which case the drop date is null), but there are multiple records for each student as students add or drop courses during the semester.  Additionally, the active enrollment for an individual student is not always linear over a given term, as seems to be necessary for many of the solutions that I've seen on the forum so far.  That is, an individual student can be active for a part of the semester, drop out and then re-enroll in the same semester later on and I'm trying to figure out how to be able to account for that fluctuating active/not active status at the unduplicated student level. 


      I've attached a copy of the workbook that I've tried to replicate the solutions from the knowledge base article above, as well as an Excel template that illustrates an example of what I'd like the visualization to ultimately look like. as you can see, my visualization using the blending method produces head counts that are much higher than they should be and other methods that I've tried have yielded similarly inflated counts.


      Blending Example.JPGEnrollment Dashboard Example.JPG


      Any suggestions or references to helpful discussion posts that I might have missed would be greatly appreciated!


      Thanks so much,