Subtraction of subsets

[george55]

Hi all,

I am using this expression which works fine:

Count(Distinct {<[Calld]={'Inbound'}>} [%CallId]) - Count(Distinct {<[Calld]={'Inbound'}, [ET]={'TALKING'}>} [%CallId])

The result is correct when an other dimension  is filted(sliced) with only one value (the mentioned dimension is not part of any dimension in this expression).

The result becomes wrong when the filted dimension has more than one values. Can anybody help?

Thanks.

George

[dwforest]

Sounds like you might need an Aggr(), where Dimension(s) are the dimension you need to slice by

Sum(<[Calld]={'Inbound'}>} [%CallId])>} Aggr(Count(Distinct {<[Calld]={'Inbound'}>} [%CallId]),Dimension(s)))

-Sum(<[Calld]={'Inbound'}, [ET]={'TALKING'}>} Aggr( Count(Distinct {<[Calld]={'Inbound'}, [ET]={'TALKING'}>} [%CallId]),Dimension(s)))

[george55]

In your solution there are some syntax errors with some curly brackets. Couldn't figure it out to fix it by myself. Thanks for your help.

Info: "[Calld]" is a tablename

[dwforest]

Syntax is available in online help

https://help.qlik.com/en-US/sense/September2019/Subsystems/Hub/Content/Sense_Hub/ChartFunctions/aggr...

[george55]

Thank you for your reply. I checked the online help already. My Problem is, how to connect the bridges between these Aggregations. E.g.:

"Sum(<["  
Must be this Sum of everything or only for the first part, before the subtraction?
Same question with the first "<["

Sum(<[Calld]={'Inbound'}>} [%CallId])>}
In your solution there are two ">}" . For me not logical. It is missing one "<}"

Thanks.

[george55 · Accepted solution]

This is the correct answer:

Sum(Aggr(Count(Distinct {<[Calld]={'Inbound'}>} %CallId), Dimension(s)))
-
Sum(Aggr(Count(Distinct {<[Calld]={'Inbound'}, [ET]={'TALKING'}>} %CallId), Dimension(s)))