topic Re: Pivot Challenge (Show all values) in QlikView

Pivot Challenge (Show all values)

pennetzdorfer — Wed, 26 Feb 2014 10:30:11 GMT

Could someone help me out on this?

Regards,

Florian

Re: Pivot Challenge (Show all values)

marcus_sommer — Wed, 26 Feb 2014 11:05:49 GMT

I'm not sure if I understand your issue right, if not you would have probably to extend your data-model to combine all possible combination of Categorie and ItemID inclusive null-values.

- Marcus

Re: Pivot Challenge (Show all values)

pennetzdorfer — Wed, 26 Feb 2014 11:55:08 GMT

Thanks for your reply, Marcus. I understand your suggestion of adding 0-values for all possible combinations of Category and ItemID.

Of course that's no problem for the little example that I posted. But in my real application I have got about 10 fields with a lot of distinct values -> the number of possible combinations is exploding (> 10,000,000 records)

Is there a way to avoid that "script workaround"?

Regards,

Florian

Re: Pivot Challenge (Show all values)

marcus_sommer — Wed, 26 Feb 2014 13:01:51 GMT

If you want show data which doesn't exist - for a null-analysis - it will be necessary to create these data. Completely meant this to create a cartesian product - most often the amount of data is then too big and it needed to be reduced on the most important data-area and/or with splitting in slices.

To do this within the script is the easiest way but in cases from reduced/splitted date not very flexible. Within the GUI you could achieve such data-structure with dynamic dimensions - it is more flexible from the included data-areas but worse from handling and performance.

Perhaps these or similar links are helpful for you:

Re: 0 Wert in Diagramm anzeigen wenn keine Daten vorhanden

Re: Nicht-Verkäufe transparent darstellen

Re: Artikel ohne Umsatz?

Re: immer alle Lagerorte anzeigen

- Marcus

Re: Pivot Challenge (Show all values)

pennetzdorfer — Thu, 27 Feb 2014 07:10:27 GMT

Your "dynamic dimensions" hint could be helpful ... I will give that a try. Thanks for your support!

Florian