Can anyone advise whether it is more efficient to use an aggregation with set analysis (eg: Sum({$<Field={'value'}>} Count) or one with no set analysis (eg: Sum(Count) and a combination of "Suppress When Value is Null" on the Dimensions tab and "Suppress Zero-Values" on the Presentation tab?
In my use case both achieve the same result. My thinking is that set analysis avoids doing unecessary aggregations that are then suppressed and so should be more efficient, but it depends on how smart QlikView is under the hood.
I haven't tested it, but I would guess that "Presentation -> Suppress Zero-values" is more efficient since it uses existing selection, whereas Set Analysis makes a new selection that needs to be evaluated.
The setting "Dimensions -> Suppress when value is NULL" only affects the dimension, and not the "Count" that you use in your aggregation.
.png "Former Employee"){kind=icon}
