ranking customers in a matrix using a dynamic filter

[jums]

I have a superstore sales dataset 2011-2015. I want to to rank Customers using a dynamic filter Top,5,20,30,100. by total volume bought.

this is the pbix

https://ln5.sync.com/dl/3c5d67f70/jbxxwhgw-5rd3agfb-p9jte22c-tewk8c7u

The matrix show customers in rows, sub categories in columns and the value is the quantity(named as volume) of each sub category divided by the total quantity the customer bought. The various measures used are in the file .

my tables 

Measures used in the dashboard

1)Ranking Select = IF(HASONEVALUE('Ranking Selections'[Ranking]), VALUES('Ranking Selections'[Rank Number]),20000)

2)Total Volume (LBS) =

SUMX(

    FactTrans,

    FactTrans[Quantity]

)

I tried to create this measure for the matrix  

3)Top Customers by  Volume subcategoryincluded =

VAR RankingDimension = VALUES(DimCustomers[Customer Name])

VAR RankingSelect = [Ranking Select]

RETURN

CALCULATE(

 CALCULATE(

    [Total Volume (LBS)],

    ALLEXCEPT(

        'DimItems',

        DimItems[Sub-Category])

        ) / CALCULATE(

            [Total Volume (LBS)],

            ALL('DimItems')

            ),

    FILTER(RankingDimension,

        RANKX(ALL(DimCustomers[Customer Name]),[Total Volume (LBS)], ,DESC) <= RankingSelect)

        )

However I don't get the image below which I want as the expected result.

If Top 5 is selected this is the expexted result .

[lbendlin]

I want to to rank Customers using a dynamic filter Top,5,20,30,100. by total volume bought.

You don't need ranking for that. A simple sort descending by volume bought already gives you the desired result.