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: 
Anonymous
Not applicable

Performance - Set Analysis or separate columns

I have a scenario where we are attributing costs per unit at several different cost centers.  So I need to display the cost per unit and extended across each cost center.  If I have a dozen or more cost centers, is it better to have a data model with one fact column and do set analysis for each expression needed in my UI charts or rather have 12 columns of facts in my data (1 for every cost center) and then have each expression calculate on the specific needed column without set analysis?

I already created the model using the thinner fact table, but am noticing some pretty substantial latency on a relatively small data set.  Before I reengineer the facts into a wider format, I want to get the group's input.

Ideas?  Best practices here?

1 Reply
Not applicable
Author

My knee-jerk response would be to go for the 12 columns in the fact table, but of course it always depends on your data model... best thing is always to test!

If the cost centre is unique to each unit and that is all that is going in the set analysis, it might not really make that much difference to have just the 1 extra column in your fact table for the cost centre.