Posterior distributions of parameters, including the tree, were estimated using Markov chain Monte Carlo sampling.
Simulations use Bayesian inference to provide probability distributions of parameters, which enable analysis of multiple minima and parameter correlation.
Disadvantages of the method pointed out by later statisticians include the different ways of assigning prior distributions of parameters and the possible sensitivity of conclusions to the choice of distributions.
It was found that estimated distributions of parameters were mildly non-Gaussian and could instead be described rather well with heavy-tailed α-stable distributions.
Means, standard errors and distributions of parameters examined in the probabilistic sensitivity analysis.
Anyway, we have checked that different distributions of parameters yielded similar results.
The empiric distributions of parameters in the LHS are represented in figure 3.
Standard deviations and ranges for the distributions were based on the literature and on the sampling distributions of parameters fit to data during model construction [28].
Posterior distributions of parameters were approximated using two independent MCMC analyses of 107 steps each, following a discarded burn-in of 106 steps.
We discarded the first 100,000 samples as a burn-in period, and used the remaining 200,000 to construct the posterior distributions of parameters and states.
Posterior distributions of parameters were investigated using Markov chain Monte Carlo (MCMC) analysis, with samples from the posterior drawn every 10,000 steps over a total of 50,000,000 steps, with the first 10% discarded as burn-in.
