GLMEM_objfn

This returns the negative of the expected log-likelihood function.

internal

Performs a spatial Bayesian general linear model (GLM) for task
functional magnetic resonance imaging (fMRI) data on the cortical surface.
Additional models include group analysis and inference to detect thresholded
areas of activation. Includes direct support for the 'CIFTI' neuroimaging
file format. For more information see A. F. Mejia, Y. R. Yue, D. Bolin, F.
Lindgren, M. A. Lindquist (2020) <doi:10.1080/01621459.2019.1611582> and D.
Spencer, Y. R. Yue, D. Bolin, S. Ryan, A. F. Mejia (2022)
<doi:10.1016/j.neuroimage.2022.118908>.

Amanda Mejia

BayesfMRI

Spatial Bayesian Methods for Task Functional MRI Studies

Daniel Spencer

Damon Pham

David Bolin

Sarah Ryan

Yu (Ryan) Yue

GLMEM_objfn function

<dl><dt>theta</dt>
<dd>a vector containing kappa2, phi, and sigma2, in that order</dd>
<dt>spde</dt>
<dd>the spde object</dd>
<dt>model_data</dt>
<dd>the model_data object containing <code>y</code> and <code>X</code></dd>
<dt>Psi</dt>
<dd>a conversion matrix (N by V) (or N by n)</dd>
<dt>K</dt>
<dd>number of covariates</dd>
<dt>A</dt>
<dd>The value for Matrix::crossprod(X%*%Psi) (saves time on computation)</dd>
<dt>num.threads</dt>
<dd>Needed for SQUAREM (it is an argument to the fixed-point functions)</dd>
<dt>Ns</dt>
<dd>The number of samples used to approximate traces using the Hutchinson
estimator. If set to 0, the exact trace is found.</dd></dl>

Arguments

Objective function for the BayesGLM EM algorithm — GLMEM_objfn

<dl>

<dt>theta</dt>
<dd>a vector containing kappa2, phi, and sigma2, in that order</dd>


<dt>spde</dt>
<dd>the spde object</dd>


<dt>model_data</dt>
<dd>the model_data object containing <code>y</code> and <code>X</code></dd>


<dt>Psi</dt>
<dd>a conversion matrix (N by V) (or N by n)</dd>


<dt>K</dt>
<dd>number of covariates</dd>


<dt>A</dt>
<dd>The value for Matrix::crossprod(X%*%Psi) (saves time on computation)</dd>


<dt>num.threads</dt>
<dd>Needed for SQUAREM (it is an argument to the fixed-point functions)</dd>


<dt>Ns</dt>
<dd>The number of samples used to approximate traces using the Hutchinson
estimator. If set to 0, the exact trace is found.</dd>

</dl>

GLMEM_objfn: Objective function for the BayesGLM EM algorithm

Description

Usage

Value

Arguments