In the real world Monte-Carlo-Modell, the model is growing. Still relatively simple, but a full recalc takes about 1.5 seconds. This seems not very long, but when you want to do 10’000 recalcs, it starts to matter.
The model was much faster, before I added some strange formulas, joining data from different matrices. I’m sure there will be more efficient ways to do this. But before starting to optimise, it would be good to know where exactly time is spent in the recalculation of the model. In most programming language, there is a way to find out where the calculation spends its time.
My question: is there any technique to profile the model, e.g. to find out where my model spends time to recalculate?
I found some 4 year old discussions on that maybe “eclipses” cost time, as well as summary items. Is there any progress on this, e.g. is there a list of “expensive” techniques / errors. E.g. use multiple selects in the same formula with nested “and” clauses (as discussed in this post), so I would be interested to know if this might be an expensive technique.
Unfortunately there is not an easily accessible way to profile calculation, especially at any granular level that would be helpful. Because Quantrix is a Java application, you can profile it with a tool like JProfiler, but someone with Java experience would need to analyze the results. We would be happy to look at what you are doing in your model to see if there are any bottlenecks.
Quantrix Engineering Manager