topic Re: Problem with full outer join in App Development

Problem with full outer join

olivier_vincon — Sat, 16 Nov 2024 05:42:55 GMT

Hello,

I have a problem in my script between table ORDRE and VOYAGE with the key 'VOYCLE', the only key between these 2 tables.

Normally, the join by default is a full OUTER JOIN in QlikSense but I don't see all the value of the key VOYCLE where VOYETAT = 'ES'

VOYAGE:
LOAD VOYCLE, *
FROM [$(PathQVD)voyage.qvd](qvd)
where
match(VOYETAT,'E','ES','I','P')

I have tried

Right Join(ORDRE)

LOAD VOYCLE, *
FROM [$(PathQVD)voyage.qvd](qvd)
where
match(VOYETAT,'E','ES','I','P')

But totally the same problem.

If I try to concatenate to table VOYAGE only the datas with the VOYETAT = 'ES', finally I have the same problem, because Qlik apply a INNER JOIN and I don't know why.

Could you tell me please ?

Best regards,

Michel

Re: Problem with full outer join

martinpohl — Wed, 29 May 2019 12:25:42 GMT

Hi!

A join in Qlik is done by all fields with the same name.

To give one field (like VOYCLE in your example) is not enough.

So rename the field you want to load but have the same name or select only fields that are necessary.

You can see which fields are match if you load the datas without join. Then you can see in data model which fields are linked.

Regards

Re: Problem with full outer join

olivier_vincon — Wed, 29 May 2019 12:44:45 GMT

Hi,

I have VOYCLE in the 2 tables ORDRE and VOYAGE. It is not the problem but the right join doesn't work, also OUTER JOIN that is by default with Qlik

Re: Problem with full outer join

Anil_Babu_Samineni — Wed, 29 May 2019 12:50:30 GMT

How is the table looks like for "ORDRE" table. Can you please share full script of yours?

Re: Problem with full outer join

olivier_vincon — Fri, 31 May 2019 07:36:45 GMT

Hi, I have isolated the problem. And it comes from the reduction of datas on the table ORDRE in the script because we have null values of VOYCLE when VOYETAT = ES in the table VOYAGE. So I have to concatenate the table ORDRE with these values.

Thank you