Markov Chain Monte Carlo sampler: using the stretch move (implementation of the Goodman and Ware emcee)
s.m.mcmc(f, lower.inits, upper.inits, max.iter, n.walkers, ...)
function that returns a single scalar value proportional to the log probability density to sample from.
vector specifying for each parameter the lower value the initial distribution.
vector specifying for each parameter the upper value the initial distribution.
maximum number of function evaluations
number of walkers (ensemble size)
further arguments passed to f
List containing:
samples[n.walkers,chain.length,n.dim]
log.p[n.walkers,chain.length]
Goodman, J. and Weare, J. (2010) Ensemble samplers with affine invariance. Communications in Applied Mathematics and Computational Science, 5(1), 65<U+2013>80, 10.2140/camcos.2010.5.65