In proto4426/PissoortThesis: Thesis in Extreme Value Theory

knitr::opts_chunk$set(echo = TRUE)

• Burn-in M number of deleted values is small, meaning that starting values are not very influent.
• Total N is the advised number of iterations, which is actually quite close to our "choice" (=2000)
• Lower bound Nmin is the minimum sample size based on zero autocorrelation. Here it is relatively low, so it is a good point.
• Dependence factor I informs about the dependence into the chains, or the extent to which autocorrelation inflates the required sample size. It is common to say that values above 5for this criterion indicate a strong autocorrelation. Here we see that it is slightly the case, especially for $\xi$.

proto4426/PissoortThesis documentation built on May 26, 2019, 10:31 a.m.