topic Re: Reducing the size of a qvw in QlikView

Reducing the size of a qvw

Anonymous — Wed, 03 Sep 2014 18:02:48 GMT

How does "group by" function help reducing the size of a qvw?

Thanks!

Re: Reducing the size of a qvw

Anonymous — Wed, 03 Sep 2014 18:07:29 GMT

If your raw data is say at City Level and you aggregate it at Country level exploiting Group By then you have less data and so smaller qvw.

You do need to look at the wider picture though and decide whether or not disposing of the City level data is a good or bad thing to do.

Re: Reducing the size of a qvw

JonnyPoole — Wed, 03 Sep 2014 18:09:18 GMT

Hi , A group by in your SQL will usually reduce the number of rows you are loading into QLIK. Sheer data volume does have an impact on your QVW size but also number of columns as well as the uniqueness of data.

If you do a group by there will typically be less data , potentially a lot of less data , but your data will also be more unique (summarized) and won't compress as much.

So yes by and large it will reduce the QVW but the exact amount will depend. Hope it helps.

Re: Reducing the size of a qvw

Anonymous — Wed, 03 Sep 2014 18:11:48 GMT

Great, thanks!

Re: Reducing the size of a qvw

JonnyPoole — Wed, 03 Sep 2014 18:24:48 GMT

However, if you are just testing limited data sets, there are a couple options:

Add the following just before your load statement where 10 is the number of records that you want the load to load

First 10

Load *

from source;

You can alternatively hit the 'debug' button on the left side select 'limited load' to load only a few records. The first option allows you to control the load per table so a bit more detailed.

Re: Reducing the size of a qvw

Anonymous — Wed, 03 Sep 2014 18:29:08 GMT

Thanks... what are the cons of using group by? Thanks again!

Re: Reducing the size of a qvw

JonnyPoole — Wed, 03 Sep 2014 20:32:17 GMT

I don't really see one except that you won't have drill down access to the underlying details.

Re: Reducing the size of a qvw

Anonymous — Thu, 04 Sep 2014 14:26:59 GMT

Thanks!