Count using Set Analysis

[PPAT1981]

Col A ID	Source	Target	ColB ID
148	Sys A	Sys K	C1
141	Sys B	Sys K	C1
130	Sys C	Sys K	C1
121		Sys K	C1

I'll use the above table to illustrate an issue I'm having. 

I want to create a set analysis to find 'Distinct Col A ID' if 'Source' is null and 'Target' is not. ( for example , COL A ID 121 in this case)

This is a measure field that is used at the row level in the pivot table.

I created set analysis using the below options and it gives incorrect result.

1)count({<[Target]-={""} distinct Col A ID) - count({<[source]-={""},[Target]-={""}>} distinct Col A ID) 

2) count({<[source]={""},[Target]-={""}>} distinct Col A ID)

what am I doing wrong here?

[GaryGiles]

The second expression works, with a slight change to the field name and using []:

count({<[Source]={""},[Target]-={""}>} distinct [Col A ID])

Not sure if that is the issue you were having.  I the expression above does not work, you may want to check to see that Source is actually getting populated with Null values.

[OmarBenSalem · Accepted solution]

null values are not always defined by "field={""}", for me the best way to find null values in set analysis would be like this :

 count({<[source]={"=len(trim(source))=0"},[Target]={"=len(trim(Target))<>0"}>} distinct Col A ID)

[PPAT1981]

Thanks but not sure why this doesn't work,  maybe because this is one of the many measures in the pivot table.

[OmarBenSalem]

can you attach a picture of what you have to see the input and its result?

[PPAT1981 · Accepted solution]

 Thanks, It's an interesting way to count null values. In my case, only the below expression calculates the value. However, I will examine the underlying data to determine a problem.

count({<[Target]-={""} distinct Col A ID) - count({<[source]-={""},[Target]-={""}>} distinct Col A ID) 

Thank you !!