poisson_control

Set computational options for the sampling algorithms

Fit multi-level models with possibly correlated
random effects using Markov Chain Monte Carlo simulation.
Such models allow smoothing over space and time and are useful in,
for example, small area estimation.

Harm Jan Boonstra

mcmcsae

Markov Chain Monte Carlo Small Area Estimation

Grzegorz Baltissen

poisson_control function

<dl><dt>nb.shape</dt>
<dd>shape parameter of the negative binomial distribution used
internally to approximate the Poisson distribution. This should be set to a relatively
large value (default is 100), corresponding to negligible overdispersion, to obtain a
good approximation to the Poisson sampling distribution. However, note that very large
values may cause slow MCMC exploration of the posterior distribution.</dd>
<dt>PG.approx</dt>
<dd>whether Polya-Gamma draws for logistic binomial models are
approximated by a hybrid gamma convolution approach. If not, <code>BayesLogit::rpg</code>
is used, which is exact for some values of the shape parameter.</dd>
<dt>PG.approx.m</dt>
<dd>if <code>PG.approx=TRUE</code>, the number of explicit gamma draws in the
sum-of-gammas representation of the Polya-Gamma distribution. The remainder (infinite)
convolution is approximated by a single moment-matching gamma draw. Special values are:
<code>-2L</code> for a default choice depending on the value of the shape parameter
balancing performance and accuracy, <code>-1L</code> for a moment-matching normal approximation,
and <code>0L</code> for a moment-matching gamma approximation.</dd></dl>

Arguments

Set computational options for the sampling algorithms — poisson_control

<dl>

<dt>nb.shape</dt>
<dd>shape parameter of the negative binomial distribution used
internally to approximate the Poisson distribution. This should be set to a relatively
large value (default is 100), corresponding to negligible overdispersion, to obtain a
good approximation to the Poisson sampling distribution. However, note that very large
values may cause slow MCMC exploration of the posterior distribution.</dd>


<dt>PG.approx</dt>
<dd>whether Polya-Gamma draws for logistic binomial models are
approximated by a hybrid gamma convolution approach. If not, <code>BayesLogit::rpg</code>
is used, which is exact for some values of the shape parameter.</dd>


<dt>PG.approx.m</dt>
<dd>if <code>PG.approx=TRUE</code>, the number of explicit gamma draws in the
sum-of-gammas representation of the Polya-Gamma distribution. The remainder (infinite)
convolution is approximated by a single moment-matching gamma draw. Special values are:
<code>-2L</code> for a default choice depending on the value of the shape parameter
balancing performance and accuracy, <code>-1L</code> for a moment-matching normal approximation,
and <code>0L</code> for a moment-matching gamma approximation.</dd>

</dl>

poisson_control: Set computational options for the sampling algorithms

Description

Usage

Value

Arguments