Set analysis questions

[flo2]

Hello, 

i still don't understand some points about set analysis even with doc and previous questions on this forum

for example https://community.qlik.com/t5/QlikView-App-Dev/Set-analysis-ignore-the-selection/td-p/1143580

and https://help.qlik.com/en-US/qlikview/May2021/Subsystems/Client/Content/QV_QlikView/ChartFunctions/Se...

if I have well understood :

- Structure of set analysis like this one : Sum({1<Category = {'category1'}>}[Number of people])
will allow not to take into account any selections on any fields ( cause of the identifier 1 ) 

- If we want precisely not to take into account selections of some fields we use structures like this one  ( we put the fields we don't want to take into account like the ones on red )
Sum({$<Category = {'category1'}, Field1, Field2, Field3, Field4, Field5.....>}[Number of people])

- but I really don't understand how to deal with cases where we have to use operators for example * or +.
Have we to put on all parts of set analysis ? like
Sum({1<Field2 = {'category1'}>+1<Field3 = {'category3'}>}[Number of people])

have we to put the 1 on the second part ? If we don't put it, the selecions for field3 is gonna be taken into account but not the selections for Field2 ?

Thanks a lot for reading !

[marcus_sommer · Accepted solution]

It depends on your data and requirements if you need an identifier on each subset of a set analysis statement or not. I think the easiest way to see the particular impact of each measure would be to create a straight-table with Field2 and Field3 and [Number of People] as dimensions and then using each (sub) statement as a single expression.

- Marcus

[flo2]

Oh ok thanks. So it's better to put always a 1  but it's not necessary mandatory depending on the data.

I will but the 1 in every part so 🙂