Period on Period Growth (without above)

[nsm1234567]

Hi There,

I'm hoping to get some input on something I'm working on.  I have a request to do a "Season on Season" growth calculation where the % growth for each season can be seen.  So Summer of 2016 will be compared to Summer of 2015.  Ordinarily, I'd build an object for this using an "above" to fetch the previous record, however, Above is sensitive to the sort order of the object and the user is going to have the option to sort the object which complicates the above, and I also battled to get the Totals working in a pivot table using "Above."

I came up with an alternate solution where I paired each season with its corresponding season in the previous year and then fiddled with aggr to get the result in the attached.  This does what I want and is not sensitive to sorting, but I'm concerned that if I use it with a large dataset I'll experience performance issues.  Is there a way I can improve the efficiency of the attached, or is my only alternative to go the "Above" route if I run into performance issues?

[sunny_talwar · Accepted solution]

Your best bet is to use The As-Of Table

[nsm1234567]

Hey there,

Thanks very much for the suggestion.  I hadn't thought of using As-Of tables for this.  I've implemented using an As-Of table which seems to work quite well.  The only downside is having the second dimension in the data model (I'm working actually on implementing in Qliksense, which makes having a second dimension trickier), but this is much more practical than the aggr expression I did.