coda_diag: Create multiple diagnostic graphics for a single parameter

In MichaelMalick/r-codatools: codatools

Description

coda_diag creates a graphic that includes three diagnostic plots to help monitor convergence of an MCMC chain for a parameter. By default the function will loop through all parameters in an mcmc.list object and create a single graphic of three diagnostic plots for each parameter.

Usage

coda_diag(coda.object, parameters = NULL)

Arguments

coda.object	an mcmc.list object
parameters	character vector of parameter names to create diagnostic plots for. If none are supplied all monitored parameters are included.

Value

A graphic with three diagnostic plots:

• traceplot of all chains

• density plot of marginal distribution for each chain

• autocorrelation for each chain

Author(s)

Michael Malick

See Also

traceplot density

Examples

library(coda)
data(line)

coda_diag(line)

coda_diag(line, parameters = "alpha")

coda_diag(line, parameters = "beta")

coda_diag(line, parameters = c("alpha", "beta"))

coda_diag(line, parameters = grep("sig", varnames(line), value = TRUE))

coda_diag(line, parameters = grep("a", varnames(line), value = TRUE))

coda_diag(line, parameters = c("alpha", grep("sig", varnames(line),
          value = TRUE)))

MichaelMalick/r-codatools documentation built on May 8, 2019, 9:56 a.m.