'Basket Analysis' pattern extension - Unique customers to a store

Thanks for going to the effort of making an example. Definitely helped me understand your working better. There's a few things that wouldn't work in your solution however and I'll try and explain them as best I can here.

• The source data actually has almost a billion rows. Query Editor will struggle to perform manipulations on that much data because (I believe) it's not part of the optimised analysis cube model until after the Query Editor step.
• In your solution, tables are only calculated when data is refreshed therefore they aren't dynamic. This affects:
  • Time Slicing - As well as slicing by a target store, I will also be slicing by time period. (E.g. find me [Sole Store Shoppers] at store 101 between 2017-08 and 2018-02)
  • Multi-selection - Selecting multiple target stores should find customers who are unique to either store. 

Appreciate the unique solution to the one specific problem, but it doesn't mesh well with the rest of my solution due to the issues mentioned above.

There's definitely going to be a way to achieve the result with a DAX expression, I've just hit the limits of my understanding on how to achieve it.