AdaptiveMultiMRW

This class implements an Adaptive Multi-site Metropolis random walk
algorithm.

MCMC

node

Implements a non-parametric source attribution model to attribute
cases of disease to sources in Bayesian framework with source and type effects.
Type effects are clustered using a Dirichlet Process. Multiple times and
locations are supported.

Poppy Miller

sourceR

Fits a Non-Parametric Bayesian Source Attribution Model

Chris Jewell

Jonathan Marshall

Nigel French

AdaptiveMultiMRW function

Object of <code><a rd-options="" href="/link/R6Class?package=sourceR&version=1.0.1" data-mini-rdoc="sourceR::R6Class">R6Class</a></code> with methods for updating a <a rd-options="" href="/link/Node?package=sourceR&version=1.0.1" data-mini-rdoc="sourceR::Node">Node</a> instance.

Format

<dl class="dl-horizontal">
<dt><code>cov</code></dt><dd>the current covariance</dd></dl><dt><code>burnin</code></dt><dd>the number of updates to burn in</dd><dt><code>tune</code></dt><dd>the current tuning matrix</dd><dt><code>naccept</code></dt><dd>the number of accepted proposals</dd><dt><code>ncalls</code></dt><dd>the number of times <code>update</code> has been called</dd><dt><code>node</code></dt><dd>the node to which the updater is attached</dd>

Fields

<dl class="dl-horizontal">
 <dt><code>new(node, tune = rep(0.1, length(node$getData())), burning = 100)</code></dt><dd>constructor takes an instance of a <a rd-options="" href="/link/StochasticNode?package=sourceR&version=1.0.1" data-mini-rdoc="sourceR::StochasticNode">StochasticNode</a>
 node, initial tuning vector (diagonal of adaptive tuning matrix), and number of burnin calls.</dd>
 <dt><code>update()</code></dt><dd>when called, updates <code>node</code></dd>
 <dt><code>acceptance()</code></dt><dd>return the acceptance rate</dd>
</dl>

Methods