Practical Monte Carlo Simulation Tool

Monte Carlo Simulation for generic Matlab commands

Description

This practical function provides a tool for quickly including a random process within a Monte Carlo Simulation framework.

The main input argument, is a string containing commands in Matlab syntax. Each execution of such string is an instance of a random process. The function expressionMCS
samples an opened number of user-specified variables and expressions, by running the input string for N iterations.

The function help is documented with four meaningful examples of various uses.

The purpose of this function is to perform simple Monte Carlo testings for any kind of stochastic processes, without having to create a single .m file. Hence it is not ment to be practicle, rather than fast. I would like to acknowledge Dimitri Shvorob for his comments.

• MATLAB release: MATLAB 6.5 (R13)

[Dimitri Shvorob]

You basically wrote a restrictive, inefficient wrapper for a simple for..end loop. Restrictive, because it only calculates mean/variance. Inefficient, because it is SLOW: a loop calculation of [y,y2] = expressionMCS({'mu' mu ; 'sigma' sigma},'y = normrnd(mu,sigma)',10000,'y','y^2')
did the job 50 times faster.

[Dimitri Shvorob]

Correction: 250 times faster: that fancy progress-report stuff takes time. But, credit where it's due: if only all FEX authors (incldig myself) provided such an extensive 'help'!

[mohammad surrati]

 

[Ahmed Farag]

I got this message when i ran the file

Input argument "parameters" is undefined.

Error in ==> expressionMCS at 93
for i=1:size(parameters,1)