topic Re: Avoid reloading table using 'resident' function in QlikView

Avoid reloading table using 'resident' function

Fri, 07 Oct 2011 08:05:55 GMT

Hi,

I want to perform an age calcuation at a point in time in my cloud.

I have two tables: Action (at a certain date) and Person (which contains the birthdate of every person performing an action).

The example and screenshot gives some extra info.

Regards,

Maarten

Avoid reloading table using 'resident' function

its_anandrjs — Fri, 07 Oct 2011 08:14:04 GMT

Hi,

If you make a QVD of the tables and then calculated table like Action and use this QVD it will much faster when you acces it it will faster like 1/10 times in comparision to other table loading.

Rgds

Anand

Avoid reloading table using 'resident' function

Fri, 07 Oct 2011 08:29:27 GMT

Hi Anand,

You mean something like this?

I see it does no longer use Synthetic keys (so that's definately a + :-)), but in that case I still need to do some kind of reload of all of the data (BUT according to you, it should be definately a lot faster), right?

Kind regards,

Maarten

Avoid reloading table using 'resident' function

its_anandrjs — Fri, 07 Oct 2011 08:36:42 GMT

Hi,

Yes thats correct also you try to include in QVD the condition that you make like for [Age at Action Date] and use that QVD.

Rgds

Anand

Avoid reloading table using 'resident' function

SunilChauhan — Fri, 07 Oct 2011 08:39:00 GMT

Load two table i.e Action and person. using left join as u did already

and use age funtionwhile representing

in expression age(.actn_date,prsn,prsn_birthdate) and give the expression name [Age at Action date];

if you are using in dimession the use calculated dimension

write age(.actn_date,prsn,prsn_birthdate) and give the label name [Age at Action date];

hope this helps

Re: Avoid reloading table using 'resident' function

Fri, 07 Oct 2011 09:05:29 GMT

@Anand,

I gave that serveral tries, but QlikView can't recognize the variable coming from the other table...

@Sunil,

If I look at the age function you are mentioning, it takes more than 2 parameters... Is that correct? Further I don't see the exact difference between the two examples you are showing... I suppose one is done in the script and the other one is done in the cloud itself? (this last one is not an option for 36M rows :-)).

Ps: I just created a QVD with the source data, to stop fooling arround with inline tables (as the actual data will also be fed from a QVD).

Thanks for your support!

Regards,

Maarten