Exclude certain rows in pivot table from sum

[blipblopdk]

In the attached example, I cannot find a way to calculate the column to the right (column 3) correctly.

Column 3 is generally calculated as the difference between column 1 and 2 (see measure below).

In rows where column 1 is 0, no value should be shown in column 3. This I have achieved as shown in the example.

The problem is that the row marked with yellow does not sum to what I want. As it is now, it sums to 238, but I want it to sum to 308. Right now, the sum incorporates the difference between column 1 and 2 for those rows where column 0 is 0, and I would like it to NOT include those differences, so it should only include differences, where column 1 is NOT 0.

It would be easily solved, if I just added "{<[Udbudte pladser] -= {0}>}" to the second part of the measure. Unfortunately, this is not possible with the way my data are structured right now since "[Udbudte pladser]" lies in one set of rows and "[CPR-nummer]" lies in a different set of rows. I could probably do some "tricks" in the data load to make up for this. But do you see other ways to achieve what I want?

BR Kim