YouвЂ™ll notice that I kept it as an inner join, matched on State (to ensure the ranking is done within the state) and then also included a condition that the Sales of the State & Customer must be less than or equal to the Sales of the Aggregate Rows step. For all intents and purposes, IвЂ™m joining the aggregation to itself: Join the Aggregate Rows to the previous aggregate step.Add a Step after the aggregation that contains the aggregate rows:.Create an aggregate of Sales by State and Customer:.If I want to rank by customer within state, here are the steps I might take: LetвЂ™s take the same data set and see how we might achieve ranking in Tableau Prep. What if you could rank the source? Ranking in Tableau Prep The difficulty with table calculations is 1) you have to set the right addressing and partitioning (not too bad in this case) and 2) you have already gone down several layers of complexity, so deeper analysis is going to be even more complex. Using Tableau, I have the option to use the Rank() table calculation, and it isnвЂ™t too difficult to come up with something like this: LetвЂ™s say I want to rank the customers for each state based on the amount of sales. LetвЂ™s use the familiar Superstore Sample data set which has Sales for Orders made by Customers in various States. Here a possibility for sorting and ranking in Tableau Prep: But to compute a rank or sort specifically, there is a work-around that will work in certain cases. This would solve a lot of use cases, including adding row numbers, rank, sort, handling cumulative values, and more. One thing IвЂ™ve submitted an idea for Tableau Prep (and would love if you voted for it!) is the ability to use Table Calculations, especially Index() and PreviousValue().

Tableau prep