Is it known behaviour?

[#1]

I;am trying to understand my app

On the picture you have few fields. All fields only in one table of data model.

In source data it looks like it is no duplicate. I also checked data, which i join to this table, and it should no multiply.

When i put into table only ID field and count (ID), my table (or barchart) shows, that i have 3 rows.

If i put to this chart any addisional field (from the same table of data model), the same measure shows 1....

I know of course, that i may use distinc But what if i would like really count rows? I dont like to put "1" to each row of each table

Every suggestion is welcome.

Best regards

Darek

P.S - i'am on desktop with about 100k rows, about 10 tables in data model.

[swuehl]

So you are counting a field that is a key linking two or more tables?

You shouldn't count a key field (at least without a DISTINCT qualifier), the result is not deterministic as Qlik does not know which table records need to be scanned.

[#3]

https://community.qlik.com/mobile/mobile-access.jspa#jive-content?content=%2Fapi%2Fcore%2Fv3%2Fconte...

[swuehl]

It's also described here by Henric:

Count or Count distinct?

[#5]

Without distinct. I looked for duplicates and my id is also business number. I prepared technical key, hided it and now it is better 😉

[tomasz_tru]

Better or as expected? So it was count on key field?