auto_corr

auto_corr_plot

This function obtains the autocorrelation of the MCMC samples in an JointAI
object via <code>coda::autocorr.diag()</code>. <code>autocorr_plot()</code> visualizes the results
using ggplot2.

Joint analysis and imputation of incomplete data in the Bayesian
framework, using (generalized) linear (mixed) models and extensions there of,
survival models, or joint models for longitudinal and survival data, as
described in Erler, Rizopoulos and Lesaffre (2021) <doi:10.18637/jss.v100.i20>.
Incomplete covariates, if present, are automatically imputed.
The package performs some preprocessing of the data and creates a 'JAGS'
model, which will then automatically be passed to 'JAGS'
<https://mcmc-jags.sourceforge.io/> with the help of
the package 'rjags'.

Nicole S. Erler

JointAI

Joint Analysis and Imputation of Incomplete Data

auto_corr function

<dl><dt>object</dt>
<dd>an object of class JointAI</dd>
<dt>lags</dt>
<dd>a numeric vector indicating the lags to consider</dd>
<dt>by_chain</dt>
<dd>logical; should the autocorrelation be computed for each
chain separately?</dd>
<dt>outcome</dt>
<dd>integer; index of the outcome model for which the
autocorrelation should be plotted</dd>
<dt>start</dt>
<dd>the first iteration of interest
(see <code>window.mcmc</code>)</dd>
<dt>end</dt>
<dd>the last iteration of interest
(see <code>window.mcmc</code>)</dd>
<dt>thin</dt>
<dd>thinning interval (integer; see <code>window.mcmc</code>).
For example, <code>thin = 1</code> (default) will keep the MCMC samples
from all iterations; <code>thin = 5</code> would only keep every 5th
iteration.</dd></dl>

Arguments

Autocorrelation of MCMC samples — auto_corr

<dl>

<dt>object</dt>
<dd>an object of class JointAI</dd>


<dt>lags</dt>
<dd>a numeric vector indicating the lags to consider</dd>


<dt>by_chain</dt>
<dd>logical; should the autocorrelation be computed for each
chain separately?</dd>


<dt>outcome</dt>
<dd>integer; index of the outcome model for which the
autocorrelation should be plotted</dd>


<dt>start</dt>
<dd>the first iteration of interest
(see <code>window.mcmc</code>)</dd>


<dt>end</dt>
<dd>the last iteration of interest
(see <code>window.mcmc</code>)</dd>


<dt>thin</dt>
<dd>thinning interval (integer; see <code>window.mcmc</code>).
For example, <code>thin = 1</code> (default) will keep the MCMC samples
from all iterations; <code>thin = 5</code> would only keep every 5th
iteration.</dd>

</dl>

auto_corr: Autocorrelation of MCMC samples

Description

Usage

Value

Arguments

Examples