plot-est.Regression-method: Plot method for the Bayesian estimation results

In BaPreStoPro: Bayesian Prediction of Stochastic Processes

Description

Plot method for the estimation results of the regression model.

Usage

## S4 method for signature 'est.Regression'
plot(x, par.options, style = c("chains", "acf",
  "density"), par2plot, reduced = FALSE, thinning, burnIn,
  priorMeans = TRUE, col.priorMean = 2, lty.priorMean = 1, ...)

Arguments

x	est.Regression class, created with method estimate,Regression-method
par.options	list of options for function par()
style	one out of "chains", "acf", "density"
par2plot	logical vector, which parameters to be plotted, order: (φ, γ^2)
reduced	logical (1), if TRUE, the chains are thinned and burn-in phase is dropped
thinning	thinning rate, if missing, the proposed one by the estimation procedure is taken
burnIn	burn-in phase, if missing, the proposed one by the estimation procedure is taken
priorMeans	logical(1), if TRUE (default), prior means are marked with a line
col.priorMean	color of the prior mean line, default 2
lty.priorMean	linetype of the prior mean line, default 1
...	optional plot parameters

Examples

model <- set.to.class("Regression", fun = function(phi, t) phi[1]*t + phi[2],
    parameter = list(phi = c(1, 2), gamma2 = 0.1))
data <- simulate(model, t = seq(0, 1, by = 0.01), plot.series = TRUE)
est <- estimate(model, t = seq(0, 1, by = 0.01), data, 1000)  # nMCMC small for example
plot(est)
plot(est, burnIn = 100, thinning = 2, reduced = TRUE)
plot(est, par.options = list(mar = c(5, 4.5, 4, 2) + 0.1, mfrow = c(3,1)), xlab = "iteration")
plot(est, style = "acf", main = "", par2plot = c(TRUE, TRUE, FALSE))
plot(est, style = "density", lwd = 2, priorMean = FALSE)
plot(est, style = "density", col.priorMean = 1, lty.priorMean = 2, main = "posterior")
plot(est, style = "acf", par.options = list(), main = "", par2plot = c(FALSE, FALSE, TRUE))

BaPreStoPro documentation built on May 2, 2019, 3:34 p.m.