plot_autocorr: Checking convergence: autocorrelation plot
In sppmix: Modeling Spatial Poisson and Related Point Processes

Description Usage Arguments Author(s) See Also Examples

This function can be used to assess convergence by visualizing the autocorrelations between the draws of the Markov chain chain. The lag k autocorrelation rho_k is the correlation between every draw and its kth lag. We would expect the kth lag autocorrelation to be smaller as k increases (that is, the 100th and 1000th draws should be less correlated than the 100th and 105th draws). For higher values of k we anticipate small autocorrelation values, otherwise the chain is not mixing well (in other words we do not explore the parameter space adequately).

For examples see

http://faculty.missouri.edu/~micheasa/sppmix/sppmix_all_examples.html#plot_autocorr

1	plot_autocorr(chain, open_new_window = FALSE, maxlag = 100)

`chain`	An `Lx1` vector containing the `L` posterior realizations.
`open_new_window`	Open a new window for the plot.
`maxlag`	The maximum lag value to consider. Default is 100.

Sakis Micheas

est_mix_damcmc, rmixsurf, rsppmix

truemix_surf <- rmixsurf(m = 3, lambda=100, xlim = c(-3,3), ylim = c(-3,3))
plot(truemix_surf)
genPPP=rsppmix(intsurf = truemix_surf, truncate = FALSE)
fit <- est_mix_damcmc(pp = genPPP, m = 3)
plot_autocorr(fit$genps[,1])
plot_autocorr(fit$genps[,2])
plot_autocorr(fit$genps[,3])