overfittingMFA_missing_values

Gibbs sampling for fitting a mixture model of factor analyzers.

Model-based clustering of multivariate continuous data using Bayesian mixtures of factor analyzers (Papastamoulis (2019) <DOI:10.1007/s11222-019-09891-z> (2018) <DOI:10.1016/j.csda.2018.03.007>). The number of clusters is estimated using overfitting mixture models (Rousseau and Mengersen (2011) <DOI:10.1111/j.1467-9868.2011.00781.x>): suitable prior assumptions ensure that asymptotically the extra components will have zero posterior weight, therefore, the inference is based on the ``alive'' components. A Gibbs sampler is implemented in order to (approximately) sample from the posterior distribution of the overfitting mixture. A prior parallel tempering scheme is also available, which allows to run multiple parallel chains with different prior distributions on the mixture weights. These chains run in parallel and can swap states using a Metropolis-Hastings move. Eight different parameterizations give rise to parsimonious representations of the covariance per cluster (following Mc Nicholas and Murphy (2008) <DOI:10.1007/s11222-008-9056-0>). The model parameterization and number of factors is selected according to the Bayesian Information Criterion. Identifiability issues related to label switching are dealt by post-processing the simulated output with the Equivalence Classes Representatives algorithm (Papastamoulis and Iliopoulos (2010) <DOI:10.1198/jcgs.2010.09008>, Papastamoulis (2016) <DOI:10.18637/jss.v069.c01>).

Panagiotis Papastamoulis

fabMix

Overfitting Bayesian Mixtures of Factor Analyzers with
Parsimonious Covariance and Unknown Number of Components

overfittingMFA_missing_values function

<dl> <dt>missing_entries</dt>
<dd>list which contains the row number (1st entry) and column indexes (subsequent entries) for every row containing missing values.</dd> <dt>x_data</dt>
<dd>normalized data</dd> <dt>originalX</dt>
<dd>observed raw data (only for plotting purpose)</dd> <dt>outputDirectory</dt>
<dd>Name of the output folder</dd> <dt>Kmax</dt>
<dd>Number of mixture components</dd> <dt>m</dt>
<dd>Number of iterations</dd> <dt>thinning</dt>
<dd>Thinning of chain</dd> <dt>burn</dt>
<dd>Burn-in period</dd> <dt>g</dt>
<dd>Prior parameter \(g\). Default value: \(g = 2\).</dd> <dt>h</dt>
<dd>Prior parameter \(h\). Default value: \(h = 1\).</dd> <dt>alpha_prior</dt>
<dd>Parameters of the Dirichlet prior distribution of mixture weights.</dd> <dt>alpha_sigma</dt>
<dd>Prior parameter \(\alpha\). Default value: \(\alpha = 2\).</dd> <dt>beta_sigma</dt>
<dd>Prior parameter \(\beta\). Default value: \(\beta = 1\).</dd> <dt>start_values</dt>
<dd>Optional (not used)</dd> <dt>q</dt>
<dd>Number of factors.</dd> <dt>zStart</dt>
<dd>Optional (not used)</dd> <dt>gibbs_z</dt>
<dd>Optional</dd> <dt>lowerTriangular</dt>
<dd>logical value indicating whether a lower triangular parameterization should be imposed on the matrix of factor loadings (if TRUE) or not. Default: TRUE.</dd></dl>

Arguments

Author

Basic MCMC sampler for the case of missing data — overfittingMFA_missing_values

<dl>

 <dt>missing_entries</dt>
<dd>list which contains the row number (1st entry) and column indexes (subsequent entries) for every row containing missing values.</dd>

 <dt>x_data</dt>
<dd>normalized data</dd>

 <dt>originalX</dt>
<dd>observed raw data (only for plotting purpose)</dd>

 <dt>outputDirectory</dt>
<dd>Name of the output folder</dd>

 <dt>Kmax</dt>
<dd>Number of mixture components</dd>

 <dt>m</dt>
<dd>Number of iterations</dd>

 <dt>thinning</dt>
<dd>Thinning of chain</dd>

 <dt>burn</dt>
<dd>Burn-in period</dd>

 <dt>g</dt>
<dd>Prior parameter \(g\). Default value: \(g = 2\).</dd>

 <dt>h</dt>
<dd>Prior parameter \(h\). Default value: \(h = 1\).</dd>

 <dt>alpha_prior</dt>
<dd>Parameters of the Dirichlet prior distribution of mixture weights.</dd>

 <dt>alpha_sigma</dt>
<dd>Prior parameter \(\alpha\). Default value: \(\alpha = 2\).</dd>

 <dt>beta_sigma</dt>
<dd>Prior parameter \(\beta\). Default value: \(\beta = 1\).</dd>

 <dt>start_values</dt>
<dd>Optional (not used)</dd>

 <dt>q</dt>
<dd>Number of factors.</dd>

 <dt>zStart</dt>
<dd>Optional (not used)</dd>

 <dt>gibbs_z</dt>
<dd>Optional</dd>

 <dt>lowerTriangular</dt>
<dd>logical value indicating whether a lower triangular parameterization should be imposed on the matrix of factor loadings (if TRUE) or not. Default: TRUE.</dd>

</dl>

overfittingMFA_missing_values: Basic MCMC sampler for the case of missing data

Description

Usage

Value

Arguments

Author