Skip to main content
Announcements
Have questions about Qlik Connect? Join us live on April 10th, at 11 AM ET: SIGN UP NOW
cancel
Showing results for 
Search instead for 
Did you mean: 
chiru_thota
Specialist
Specialist

Do we have any constraint on number dimension putting into a pivot table or Straight table ?

Hi All,

I am working on a application where I need to keep 300 dimension & 100 metrics  in pivot table and do conditional hide and show .

Is there any problem if I put those many dimensions in to Pivot table or Straight table ?

Thanks,

Chiru

1 Solution

Accepted Solutions
chiru_thota
Specialist
Specialist
Author

After Testing a lot we came to conclusion.

Pivot table :

(1) After 127 dimensions you see weird behavior in Pivot table like metrics will come to rows and dimension will jump to columns.

(2) Let say you are using conditional hide and show and have only selected 128th dimension and some metrics.

Still we will see weird behavior.

(3) up to 127 Dimensions and as many as metrics Pivot table will work fine.

Straight table :

No constraint on number of dimensions.

Thanks,

Chiru.

View solution in original post

4 Replies
edg_baltazar
Partner - Creator
Partner - Creator

Hi Thota!

In my experience I would say that maybe you could have performance issues with such a large report , but I would use a Pivot table instead of a Straight Table because the pivot table allows you to group values and maybe this would make easier to read the report, also you can collapse the user can collapse the dimensions that they don't want to use and this also avoids to have displayed a large report all the time.

Hope this helps on something!

Best Regards!

jonathandienst
Partner - Champion III
Partner - Champion III

Hi

I have never tried that many dimensions and expressions in a single chart, but I would not be surprised if the performance was terrible.

What I would do in this scenario, is group the dimensions and metrics into logical groups (such as dates, organisation structure, customers & customer hierarchy, products and product hierarchy etc), and create one pivot table for each of these groups. These pivots should only have the metrics relevant for that dimension group.

I would inline load a selector field that I would use in a list box to select which pivot is to be shown, and conditionally hide the rest. The dimension and metric selectors would have to sensitive to the group selector, so I would include the group selector in the inline load of the metrics and dimensions.

You could even split this into multiple tabs, one for each group. I do not think that the monster one table fits all is a good design for the users either, but engage with them to decide which design works best for them.

Regards

Jonathan

Logic will get you from a to b. Imagination will take you everywhere. - A Einstein
preminqlik
Specialist II
Specialist II

see "Whats new in qlikviw 11" reporting tab

chiru_thota
Specialist
Specialist
Author

After Testing a lot we came to conclusion.

Pivot table :

(1) After 127 dimensions you see weird behavior in Pivot table like metrics will come to rows and dimension will jump to columns.

(2) Let say you are using conditional hide and show and have only selected 128th dimension and some metrics.

Still we will see weird behavior.

(3) up to 127 Dimensions and as many as metrics Pivot table will work fine.

Straight table :

No constraint on number of dimensions.

Thanks,

Chiru.