Point Process MCMC JavaScript program, by Jeffrey Rosenthal. Description at the bottom.

This program runs a random-scan component-wise Metropolis MCMC algorithm on a spatial point process hardcore model for a fixed number of particles, with initial distribution uniform over the pink region, and with target density (with respect to uniform) proportional to exp(-H), where

H(z_1, z_2, ..., z_N) = A * sum_{i<j} |z_i − z_j| + B * sum_{i<j} (1 / |z_i − z_j|) + C * sum_i (z_i1)
where |...| is Euclidean distance, z_i is the position of the i'th particle, and z_i1 is the first coordinate of z_i.

The program accepts the following keyboard inputs (or you can mouse click):


This program is written by Jeffrey Rosenthal. See also my other JavaScript, my Stochastic Processes book, and my Java Applets.