Modify count of customers per product to be based on dimension table attributes

[bjackson202]

Hello all,

I would like to change this calculated column definition to reference customer & product definitions found in dimProduct & dimCustomer tables:

Go from this:

// Distinct Count of Customer Key's per Product Key

CALCULATE(
DISTINCTCOUNT(factInvoices[CustomerKey]),
filter(factInvoices,factInvoices[ProductKey]=EARLIER(factInvoices[ProductKey]))
)

to this:

// Distinct Count of Customer Parents's per ProductParent

CALCULATE(
DISTINCTCOUNT(dimCustomer[CustomerParent]),
filter(factInvoices,dimProduct[ProductParent]=EARLIER(dimProduct[ProductParent]))

)

How can I force the calculated column expression to use row syntax for a related product & customer dimensions? 

Thank you!!!!

[Anonymous]

Hi @bjackson202 ,

You can try to use RELATED() function.

Please try:

CALCULATE(
    DISTINCTCOUNT(dimCustomer[CustomerParent]),
    FILTER(
        factInvoices,
        RELATED(dimProduct[ProductParent]) = EARLIER(RELATED(dimProduct[ProductParent]))
    )
)

CALCULATE(
    DISTINCTCOUNT(dimCustomer[CustomerParent]),
    FILTER(
        factInvoices,
        factInvoices[ProductKey] = EARLIER(dimProduct[ProductKey])
            && factInvoices[CustomerKey] = RELATED(dimCustomer[CustomerKey])
    )
)

If the above one can't help you get the desired result, please provide some sample data in your tables (exclude sensitive data) with Text format and your expected result with backend logic and special examples. It is better if you can share a simplified pbix file. Thank you.

Best Regards,

Neeko Tang

If this post  helps, then please consider Accept it as the solution  to help the other members find it more quickly. 

[bjackson202 · Accepted solution]

Thank you @Anonymous !

I ended up finding a post here that helped me out: https://www.daxpatterns.com/related-distinct-count/

The earlier() functionality was giving me issues with the dimension table attributes, so I changed the distinct count to a measure named [# Customers]: 

# Customers =

VAR CustsFromInvoices =

    SUMMARIZE ( FILTER(factInvoices,factInvoices[CurrentorBase]="Current"), dimCustomer[CustomerParent])

VAR Result =

    SUMX ( CustsFromInvoices, 1 )      -- optimization for COUNTROWS ( CustsFromInvoices )

RETURN

    Result

Then called that measure in a calculated column here: 

# Customer Parents per Product Parent =

VAR numCustPerProd =

    CALCULATE(

        [# Customers],                                              

        ALLEXCEPT(factInvoices,                                             

            dimProduct[ProductParent]

        )            

    )    

This got me my desired result, thank you for your time & help, and apologies for the lack of sample data & context!