Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hello Qlikperts
I need a small help regarding set analysis ....
I have 15 colomns from various dimensions in my app & out of those 15 I want manipulate my selections from 5 colomns & i want to ignore other selections from the remaining 10 colomns .
In a breif let us assume that i have 3 dimensions like
Dim 1 : Dim2: Dim3:
colomn1 colomn6 colomn9
colomn2 colomn7 colomn10
colomn3 colomn8 colomn11
colomn4 colomn12
colomn5 colomn13
colomn14
colomn15
and i want to make selections modify my chart values from colomn1,colomn3,colomn7,colomn11,colomn15 and i want to neglect / ignore the impact of other colomns selection on my chart so
In my set analysis i tried to write including these colomns like
sum({1<colomn1=,....,colomn15=>}(sales/gross_sales))
but it's not working & i tried in other way like hard coded one
sum({<colomns i want to ignore>sales/gross)
but this one is also not giving me my required valuations , Is there any thing i'm going wrong on it or else can anyone help me with ignoring those in a set analysis
Thanking u all
Meher
I would expect your second approach to work:
sum({<colomn2=,colomn4=,...colomn14=>} sales/gross_sales)
What problems are you seeing with it?
To avoid manually listing ten (or hundreds) of fields, try Smakovskiy's clever approach suggested at the end of his document on this very subject:
Thanks John ,
I tried in Smakovskiy's clever approach but i'm failing to declaremy set in a variable as he said , & i'm trying it in my 2nd way of approach but my manual calculation values are matching withsome & not matching with some . I don't know where i'm going wrong .
thank you
Meher
Well, he's not declaring the set in a variable. It's just a complicated set analysis expression. But if you can't get it to work by manually listing the fields, his approach won't work either, as it's just a way of automatically listing the same fields.
Is it possible to post a small sample application demonstrating the problem with the manual approach? You might be able to do that by reducing the data in your real application if you can't come up with a simple, fake example. You can also scramble any sensitive information as required. Hopefully it'll be easy to see what's going wrong if we have something to look at.
Thanks John ,
I'll try to give you a copy/sample ASAP
thank you
Meher
John I fixed it , it is not a error in set analysis , it's a error in data
thank you
Meher