over the time my QlikView application became a bit slow. I mainly use simple expressions like "Sum(Revenue)". (In performance tests I already sorted out the more complex expressions)
The sluggish performance is only disturbing, if in ONE certain field (valuetype) no selection is made. So I thought about a simple solution to calculate the Sum only if a selection is made.
=if(GetSelectedCount(valuetype)=0,0,Sum(Revenue))
This works in terms of returning 0 in case no selection is made, but the (bad) performance is as it was before. The idea is to calculate the Sums only if a selection is made for valuetype, otherwise just return 0 ("do not even touch the Sum formula"). I thought this should be faster than any Sum(if(a=b,myfield)) or Sum({a=b}myfield) approach, where the Sum-function must at least do 'something'. But it is not ....
If it is okay to display the chart only when some selections are made, then you could try this:
In the chart properties, you can put calculation condition "getselectedcount(valuetype) > 0". This will show the chart only when some selection is made else will display a message which can be customized (Properties/General/Error Messages).
But still I'd like to understand QlikView in more detail.
In case of a formula like "if (condition, expression1, expression2)" ... when is 'expression2' processed? I would expect that this is done only in case of 'condition = false'. Does QlikView work this way?
