Why do I have a null displayed dimension in chart with SA?

[#1]

I've created a simple stright table with 2 dimensions (my actual one had more, but result was the same for other than one dim) and one expression with set analysis. The qvw is attached. I want to have CID, NUM and minimum pdate displayed for the records with NULL fdate. But for some reason my chart shows me NUM field with NULL (displayed as -) for CID = 1. Why? How can I (if any way) avoid it? Thanks!

[kaushiknsolanki]

Hi,

What is the expected O/P?

Regards,

Kaushik Solanki

[#3]

Don't quite catch, what stands for 'O/P', but I expected to see NUM = 111 in the row. Globally I wanted to filter out the rows with fdate not null.

[kaushiknsolanki]

O/P is output.

From the data set which you gave what is the output u want.

Regards,

Kaushik Solanki

[#5]

Hmmm, as I've already written I'd expected to have NUM field with '111' in the straight table for the first row.

[kaushiknsolanki]

Hi,

There is no value as Num 111 in your data set.

Regards,

Kaushik Solanki

[Anonymous · Accepted solution]

Is this what you are searching? to show 1-123-20.02.2017; 3-333-10.08.2017?

[#8]

Sorry, my bad, I really meant 123 for num.

[#9]

Yes, that's the point. But why the... ))). Why is it so? Why do the Qlik behave such unpredictable (for me) way? Where can I read about it to know where to use aggr (or something like that) instead of simple functions? Thanks anyway )

[Anonymous]

Hi,

i think this post on AGGR is nice

Aggr() function‌‌

PS: You can use the formula WITHOUT "aggr" function. In this case there is no need for aggr, because you have already in the dimension NUM and CID.

Best regards,

Cosmina