Inherited distinct in join

[#1]

Hi to all!

Can some one tell me why distinct is inherited from previos load in join?

See the attachment. I am loading first table distinct and then loading second table NOT distinct and joining them. The result table turns out distinct too. Why?

[#2]

I see count=2 on test1 and count=3 on test2

[#3]

Yes. In test 1 dublicate rows are dropped and i dont understand why.

Test 1 is with distinct load in first table and the second is without.

But the data is the same in both cases.

[tresesco]

This is because your join resultant table is Groups1 which is qualified by DISTINCT before load. Therefore the table is loaded as distinct.

See the attached file, if you take a resident of the Groups1 table and then JOIN, you would get different result, because your resultant table is not qualified by DISTINCT there.

[#5 · Accepted solution]

This has been acknowledged for a while now, and considered working as designed

http://community.qlik.com/thread/5762

[#6]

I know how to work around this, the question is why it works like this. It breaks the logic.

[#7]

It costed me a half day to determine this. This should be in QlikView help file. This behavior of distinct keyword is not intuitive.