The simulation module can run the following methods for equations involving random variables with known distributions:

- Monte Carlo
- Latin Hypercube
- Orthogonal Latin Hypercube

When using variables with unknown distributions, the Rosenblueth 2^(k+1) simulation type is available.

Various data-generation algorithms are also available:

- Random events in time
- Uncorrelated time series
- Spatial events (with circular, rectangular, or arbitrary boundaries)
- Random Walk
- Random Fields

All simulation types are carried out by a thread pool, which insures that all of the available processors are used in an optimal way. Thus, it runs the simulations one after the other if the machine has a single-core processor (Pentium IV, AMD64, etc), but it handles multiple cores if present (Core2Duo, i7, AMD X2, etc) to run simulations in parallel. For example, a quad-core processor allows the module to perform 4 simulations at once, making the entire process 4 times faster than with Mathematica, Mathlab, or any other script-interpreting software package.

You can generate simulation reports just like this one

2r_sim.txt · Last modified: 2011/07/27 10:55 by diego · []