This is part of my fake/test-data I use:
What I want to do is:
1. For each product be able to tell who is top1 to topN shop based on the price.
I managed that with the FirstSortedValue (but if there are two on top1 using the rank would be better)
2. I want to see the distribution of price-positions per shop.
Thus with the given data shop A has:
Top1: 4 times
Top2: 6 times
Top 3: -
Top1: 3 times
Top2: 5 times
Top 3: 1 time
In a table it would be possible to show the AVG(position) per shop (if shop is the dimension in the table)