Set Analysis A -B

[#1]

I have a list of projects from the source A. And have the second list of projects from the source B. How do I list A-B or vice verse?

[Anonymous]

Make sure you have a keyfield in both sources with the same name then it will connects automaticly when you load.

If you share your loading script I glady have a look at it.

[#3]

Yea. I named the both fields same. Currently, I can distinguish the source of the projects, whether from list A or B. However, I don't know a way to have a chart (or a list etc.) listing only projects which are from source A but not in source B (A-B) or vice verse.

[#4]

In addition to naming the keyfield the same, for example project, you can added an addtional field to identify the project as A or B.

project,

project AS project A

project AS project B.

QlikView will not see these as keys so you can identifiy each project.

These can also be counted in expressions.

Hope this helps.

Don

[#5 · Accepted solution]

Is it in the script you want to take care of it? It might not be the most optimized way, but here is an example

Table1:
Load * inline [
Num, Label1
1,a
2,b
3,c];

Table2:
Load * inline [
Num, Label2
3,c
4,d
5,e];

Final:
Load Num, Label1 as Label, 'Table1' as Source
Resident Table1
Where NOT Exists(Label2,Label1);
concatenate
Load Num, Label1 as Label, 'Both' as Source
Resident Table1
Where Exists(Label2,Label1);
concatenate
Load Num, Label2 as Label, 'Table2' as Source
Resident Table2
Where Not Exists (Label1, Label2);

Drop table Table1;
Drop table Table2;

[#6]

Your solution works fine in order to only list few fields in the manner I want . However, I have more than 20 fields along with to operate on in both Table 1 and Table 2. So, in my case QV can't manage to load the data; waiting for minutes, then freezing.

[#7]

I have fixed my issue. I used your way to distingue the sources and to collect the data from the two tables into the final table with the related information and id only. Then I reloaded Table1 and Table2 with all the fields, and making a logical connection between the final table and these two full loaded tables through id fields by naming them same.