I teach Tableau to both coworkers and college students and recently came across a change to the Returns data in the Sample Superstore Data file where there are multiple rows per order ID in the returns data associated with 2018.x versions.
I generally use this data when covering joins and blending but now when processing the join it is inaccurate due to the many to many relationship of order IDs in this data.
Has anyone else used this for similar and come up with a work-around?
Attached is a file illustrating the issue - I've created datasources joining the Orders Table to Returns table for both the 10.3 and 2018.1 versions of the superstore data (as well as drill down on one specific returned order when looking at just the orders table from each version.)
The only workaround I've currently come up with is providing the older version of the file.