correct_marginals

The default method of computation for aghq objects computes approximate marginals using an outdated algorithm
with no known theoretical properties. The updated algorithm computes pointwise approximate marginals that
satisfy the same rate of convergence as the original approximate marginal likelihood. This function takes
a fitted aghq object and recomputes its marginals using the updated algorithm

Adaptive Gauss Hermite Quadrature for Bayesian inference.
The AGHQ method for normalizing posterior distributions
and making Bayesian inferences based on them. Functions are provided for doing
quadrature and marginal Laplace approximations, and summary methods are provided
for making inferences based on the results.
See Stringer (2021). "Implementing Adaptive Quadrature for Bayesian Inference:
the aghq Package" <arXiv:2101.04468>.

Alex Stringer

aghq

Adaptive Gauss Hermite Quadrature for Bayesian Inference

correct_marginals function

<dl><dt>quad</dt>
<dd>An object of class <code>aghq</code> returned by <code>aghq::aghq()</code>.</dd>
<dt>...</dt>
<dd>Not used</dd></dl>

Arguments

Correct the posterior marginals of a fitted aghq object — correct_marginals

<dl>

<dt>quad</dt>
<dd>An object of class <code>aghq</code> returned by <code>aghq::aghq()</code>.</dd>


<dt>...</dt>
<dd>Not used</dd>

</dl>

correct_marginals: Correct the posterior marginals of a fitted aghq object

Description

Usage

Value

Arguments