Plot results of fitting a twins model using MCMC output. Plots similar to those in the Held et al. (2006) paper are generated.

## Usage

# S3 method for atwins
plot(x, which=c(1,4,6,7), ask=TRUE, ...)

## Arguments

x

An object of class "atwins" as returned by algo.twins.

which

a vector containing the different plot types to show

1

A plot of the observed time series Z is shown together with posterior means for the number of endemic cases (X) and number of epidemic cases (Y).

2

This plot shows trace plots of the gamma parameters over all MCMC samples.

3

This shows a trace plot of psi, which controls the overdispersion in the model.

4

Autocorrelation functions for K and psi are shown in order to judge whether the MCMC sampler has converged.

5

Shows a plot of the posterior mean of the seasonal model nu[t] together with 95% credibility intervals based on the quantiles of the posterior.

6

Histograms illustrating the posterior density for K and psi. The first one corresponds to Fig. 4(f) in the paper.

7

Histograms illustrating the predictive posterior density for the next observed number of cases Z[n+1]. Compare with Fig.5 in the paper.

Boolean indicating whether to ask for a newline before showing the next plot (only if multiple are shown).

...

Additional arguments for stsplot_time, used for plot type 1.

## Details

For details see the plots in the paper. Basically MCMC output is visualized. This function is experimental, as is algo.twins.

## References

Held, L., Hofmann, M., Höhle, M. and Schmid V. (2006) A two-component model for counts of infectious diseases, Biostatistics, 7, pp. 422--437.

## Author

M. Hofmann and M. Höhle

algo.twins (with an example)