WARNING! DISCLAIMER! I make no claims about the software provided here. It is likely still under active development and may work poorly or not at all, may be poorly or misleadingly commented and may bork your computer! Use at your own risk! Please contact me if you would like to use any code here for your own research. It is presented here only for tutorial purposes. I also have Fortran code available - please contact me if you are interested in looking at that.
Matlab code. This code performs a Markov-chain Monte Carlo inversion on parameters of some simple model. Currently,
these models include damped harmonic oscillator, 1D nonlinear advection and the Lorenz (1963) 3-component
model. Note that mcmc_toy.m calls mcmc_fwd.m, which may call some other code, depending on what model you choose.
Please take time to read comments for a more thorough documentation.
Update -- September 16, 2011: code posted.
Update -- September 17, 2011: development halted on mcmc_toy.m
mcmc_toy.m | mcmc_fwd.m | mcmc_lorenz.m | mcmc_dampharm.m
Matlab code. This code replaces mcmc_toy.m and contains improvements including implementation of
a correlated proposal distribution, as well as the option of delayed rejection, adaptive Metropolis, or
DRAM samplers.
Update -- December 04, 2011: code posted.
This page uses the Perfect 'Left Menu' 2 Column Liquid Layout by Matthew James Taylor. View more website layouts and web design articles.
This personal webpage is not an official University of Miami webpage. See disclaimer.
© 2011 Marcus van Lier-Walqui