SMath Studio Forum
»
SMath Studio
»
Questions
»
Speeding up calculation time for Monte Carlo simulations
Rank: Member Groups: Registered
Joined: 21/11/2023(UTC) Posts: 24 Was thanked: 2 time(s) in 2 post(s)
|
Hi all, I don't even know if this is possible, but I figured I would ask anyways to see if someone might know any tricks. I wrote a worksheet to show my students an example of a method for making probabilistic calculations of a parameter instead of using deterministic methods. The probabilistic calculation using Monte Carlo simulation lends itself to calculating probability of failure, which is a nice way of explaining system reliability. Ideally, one would run a large number of simulations so that the result converges. However, it seems like the matrix calculations (or at least the way I am doing them) are quite slow, and when I assign a meaningful number of trials (say, 1000 or more), it takes the worksheet forever to calculate. I can do the same in Excel and it is almost instantaneous to do even 10000 or 20000 trials, but I hate doing these in Excel since I can show the work and the equations much more easily with SMath. Any suggestions on what might speed up the run-time? The worksheet is attached for reference. Thanks! FT Correlated Normal Variables - Bearing Capacity Example.sm (418kb) downloaded 14 time(s).
|
|
|
|
Rank: Advanced Member Groups: Registered
Joined: 15/04/2012(UTC) Posts: 1,990 Was thanked: 1128 time(s) in 724 post(s)
|
You just have to switch all cells to numeric optimization (Ctrl A to select everything, right mouse button on any cell, select Optimization> Numeric Then, with 10000 tries it took a couple of seconds on my 6 years old notebook. Correlated Normal Variables - Bearing Capacity Example_Kr.sm (418kb) downloaded 16 time(s). |
|
1 user thanked mkraska for this useful post.
|
|
|
Rank: Member Groups: Registered
Joined: 21/11/2023(UTC) Posts: 24 Was thanked: 2 time(s) in 2 post(s)
|
And there it is, easy as that Thank you so much! Originally Posted by: mkraska You just have to switch all cells to numeric optimization (Ctrl A to select everything, right mouse button on any cell, select Optimization> Numeric Then, with 10000 tries it took a couple of seconds on my 6 years old notebook. Correlated Normal Variables - Bearing Capacity Example_Kr.sm (418kb) downloaded 16 time(s).
|
1 user thanked gth802s for this useful post.
|
|
|
Rank: Guest
Groups: Registered
Joined: 04/07/2015(UTC) Posts: 6,866 Was thanked: 981 time(s) in 809 post(s)
|
Originally Posted by: gth802s Any suggestions on what might speed up the run-time? The worksheet is attached for reference. The attached document runs Monte-Carlo in two styles. 1. Logistic style 4 s 2. Ordinary style 0.7 s Observe eval(,)Jean Stat Monte Carlo [rlogis] Copy 2.sm (103kb) downloaded 8 time(s).
|
|
|
|
Rank: Member Groups: Registered
Joined: 21/11/2023(UTC) Posts: 24 Was thanked: 2 time(s) in 2 post(s)
|
Thanks for sharing, looks like it will be useful! Originally Posted by: Jean Giraud Originally Posted by: gth802s Any suggestions on what might speed up the run-time? The worksheet is attached for reference. The attached document runs Monte-Carlo in two styles. 1. Logistic style 4 s 2. Ordinary style 0.7 s Observe eval(,)Jean Stat Monte Carlo [rlogis] Copy 2.sm (103kb) downloaded 8 time(s).
|
|
|
|
SMath Studio Forum
»
SMath Studio
»
Questions
»
Speeding up calculation time for Monte Carlo simulations
Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.