Guilherme Wood wrote:
> Hi, SPM'ers!
>
> What is expected to occur when we estimate a PPM? Is something like a
> PPM.mat file created or is SPM.mat just actualized with the new
> estimations?
> I tried to estimate one in the last days but I didn't realize changes in my
> SPM... :)
>
A PPM field will be appended to the SPM data structure ie.
SPM.PPM
This contains information that is used by spm_contrasts
to make inferences based on the posterior densities.
In fact the information in SPM.PPM is quite minimal
as a Taylor series expansion trick is
used so that the posterior covariance matrix at each voxel
(a p x p matrix when we have p regressors) can be reconstructed from the
info in SPM.PPM and a voxel-wise noise variance parameter.
For further details see the
section on `A computational saving' in
Friston & Penny (2003) Neuroimage 19, 1240-1249
best wishes,
Will.
> All the best,
> Guilherme
> ==========================================================
> Dipl.psych. Guilherme Maia de Oliveira Wood
> Neurologic Clinic-Neuropsychology, RWTH-Aachen
> Pauwelstrasse, 30, 52074-Aachen, Germany
> XX49(0241)8089-909
> XX49(0241)886-0752
> [log in to unmask]
> ==========================================================
>
> _________________________________________________________________
> The new MSN 8: advanced junk mail protection and 2 months FREE*
> http://join.msn.com/?page=features/junkmail
>
>
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7478
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
--
William D. Penny
Wellcome Department of Imaging Neuroscience
University College London
12 Queen Square
London WC1N 3BG
Tel: 020 7833 7478
FAX: 020 7813 1420
Email: [log in to unmask]
URL: http://www.fil.ion.ucl.ac.uk/~wpenny/
|