Dynamic segemenation. How to then split by a column?

[gazpage]

I've used the Clustered Sales example from sqlbi (Dynamic Segmentation – DAX Patterns) to good effect. In my case, the properties is effectively the customer and UnitTape contains the sales.

OrigNumberOfPropertiesClustered = 
CALCULATE (
 [NumProperties],
    FILTER (
        ADDCOLUMNS (
            'Properties',
            "RankValue", CALCULATE (
                SUM ( UnitTape[Last Bank Appraisal Value (€)] ),
                CALCULATETABLE ( 'Properties' ),
                ALLSELECTED ()
            )
        ),
        COUNTROWS (
            FILTER (
                ValueSegment,
                [RankValue] > ValueSegment[Min]
                    && [RankValue] <= ValueSegment[Max]
            )
        )
            > 0
    )
)

This works perfectly. What I now want to do is take the resulting table and split the sales by type (a column in the UnitTape table). However, if I put a filter on the table then it dynamically recalculates the constituents. I suppose I really want to freeze the results of the second clause which determines which properties are included in which bucket, and then be able to segment.

I tried wrapping the [NumProperties] measure in another CALCULATE to apply a filter, but it didn't work.

Thanks