dim_reduce

Performs dimension reduction and prewhitening based on probabilistic PCA using
SVD. If dimensionality is not specified, it is estimated using the method
described in Minka (2008).

internal

Implements Bayesian brain mapping with population-derived priors,
including the original model described in Mejia et al. (2020)
<doi:10.1080/01621459.2019.1679638>, the model with spatial priors described
in Mejia et al. (2022) <doi:10.1080/10618600.2022.2104289>, and the model
with population-derived priors on functional connectivity described in Mejia
et al. (2025) <doi:10.1093/biostatistics/kxaf022>. Population-derived priors
are based on templates representing established brain network maps, for
example derived from independent component analysis (ICA), parcellations,
or other methods.  Model estimation is based on expectation-maximization or
variational Bayes algorithms. Includes direct support for 'CIFTI', 'GIFTI',
and 'NIFTI' neuroimaging file formats.

Amanda Mejia

BayesBrainMap

Estimate Brain Networks and Connectivity with Population-Derived
Priors

Damon Pham

Nohelia Da Silva

dim_reduce function

<dl><dt>X</dt>
<dd>A numeric matrix, with each column being a centered timeseries.
For fMRI data, <code>X</code> should be <code>T</code> timepoints by <code>V</code> brain
locations.</dd>
<dt>Q</dt>
<dd>Number of latent dimensions to estimate. If <code>NULL</code> (default),
estimated using PESEL (Sobczyka et al. 2020).</dd>
<dt>Q_max</dt>
<dd>Maximal number of principal components for automatic
dimensionality selection with PESEL. Default: <code>100</code>.</dd></dl>

Arguments

PCA-based Dimension Reduction and Prewhitening — dim_reduce

<dl>

<dt>X</dt>
<dd>A numeric matrix, with each column being a centered timeseries.
For fMRI data, <code>X</code> should be <code>T</code> timepoints by <code>V</code> brain
locations.</dd>


<dt>Q</dt>
<dd>Number of latent dimensions to estimate. If <code>NULL</code> (default),
estimated using PESEL (Sobczyka et al. 2020).</dd>


<dt>Q_max</dt>
<dd>Maximal number of principal components for automatic
dimensionality selection with PESEL. Default: <code>100</code>.</dd>

</dl>

dim_reduce: PCA-based Dimension Reduction and Prewhitening

Description

Usage

Value

Arguments