Does wblrnd use the inverse tranform method to generate random numbers?

Hi everyone,

im working with the Statistics Toolbox, and i want to perform a Monte Carlo Simulation for a Weibull pdf. After reading a lot of papers i found the inverse transform method. My question is, if i use the command wblrnd it generates random values of the pdf defined by shape and scale parameter. If i use the random number generator between 0 and 1 and use the inverse cdf of this Weibull pdf, do i get the same results as wblrnd? In the description of wblrnd there was no explanation how the numbers are generated.

WeibullPDFGB=fitdist(MTTF,'Weibull');

i can extract the shapefactor=2 and scalefactor=4800 for example

is:

z=rand(200,1); y=icdf('Weibull',z,shapefactor,scalefactor);

the same as:

y=wblrnd(shapefactor,scalefactor,200,1); ?

Thanks a lot for the support

Jeremy

Answer 1

[John D'Errico]

The answer to many questions about MATLAB, is to just read the code. So, I did this...

type wblrnd

Then look at the code. the last two lines are...

% Generate uniform random values, and apply the Weibull inverse CDF.
r = A .* (-log(rand(sizeOut))) .^ (1./B); % == wblinv(u, A, B)

If that does not seem sufficiently clear, I see this comment in the code:

%   WBLRND uses the inversion method.

[Jeremy Bieler]

Thanks a lot for the quick answer.