param_i_qda_linearized

Calculates the function W described in Metodiev et al. (2025).

internal

Implements the truncated harmonic mean estimator (THAMES)
of the reciprocal marginal likelihood for uni- and multivariate mixture
models using posterior samples and unnormalized log posterior values via
reciprocal importance sampling.
Metodiev, Irons, Perrot-Dockès, Latouche & Raftery (2025)
<doi:10.48550/arXiv.2504.21812>.

Martin Metodiev

thamesmix

Truncated Harmonic Mean Estimator of the Marginal Likelihood for
Mixtures

Nicholas J. Irons

Marie Perrot-Dockès

param_i_qda_linearized function

<dl><dt>g</dt>
<dd>number indicating the component g</dd>
<dt>sims</dt>
<dd>n_simul x G x (u+1) array of parameters sampled from
the posterior, where
 n_simul is the number of simulations from the posterior,
 G is the number of components,
 u is the number of mixture component parameters
 (parameter u+1 is the mixture weight)</dd>
<dt>meanhat</dt>
<dd>list of means of component parameters</dd>
<dt>sigmahat</dt>
<dd>list of covariance matrices of component parameters</dd>
<dt>non_I_set</dt>
<dd>complement of an independent set of a DAG</dd></dl>

Arguments

Calculates the function W described in Metodiev et al. (2025). — param_i_qda_linearized

<dl>

<dt>g</dt>
<dd>number indicating the component g</dd>


<dt>sims</dt>
<dd>n_simul x G x (u+1) array of parameters sampled from
the posterior, where
 n_simul is the number of simulations from the posterior,
 G is the number of components,
 u is the number of mixture component parameters
 (parameter u+1 is the mixture weight)</dd>


<dt>meanhat</dt>
<dd>list of means of component parameters</dd>


<dt>sigmahat</dt>
<dd>list of covariance matrices of component parameters</dd>


<dt>non_I_set</dt>
<dd>complement of an independent set of a DAG</dd>

</dl>

param_i_qda_linearized: Calculates the function W described in Metodiev et al. (2025).

Description

Usage

Value

Arguments

References