Dear Tobias

There are two work arounds:

1. Make a copy of your favourite design matrix before you specify the data, this design matrix can then be copied when you need it for another study.

2. This solution is a bit more dirty: go to the directory where your analysis is located and load the SPM.mat file into the matlab workspace:

load SPM.mat

and clear the fields where the filenames and their info is stored:

SPM.xY=rmfield(SPM.xY,'P')
SPM.xY=rmfield(SPM.xY,'VY')

and save the new SPM structure:

save SPM.mat SPM

Hope this helps

Torben Lund, Danish Research Centre for MR

P.S. the second solution will keep the same HP and correlation parameters as specified in the original data file, but will hopefully recalculate the hyper parameters (please check this).

Egner, Tobias wrote:

[log in to unmask]">

 Thanks Darren,

just to clarify, I didn't try to use an SPM99 matrix file in SPM2, but was
looking for (and hoping in vain) the possibility to apply an SPM2 SPM.mat
specified design to different subjects' data, as one would have (could have)
done with the old design matrix in SPM99.
That this isn't possible in SPM2 seems a serious draw-back in terms of
analysis time. I am aware that one can possibly automate that step by using
batch routines but I wondered whether there couldn't be an easy way to
integrate this into the GUI? Given that the parameter specification in SPM2
consists of a "design" step, followed by a second step selecting the data to
be fitted to the specified design, it would be great (and somehow intuitive)
if one could choose to select different sets of data...?
Would that be fiendishly complicated?

best wishes,

Tobias Egner.

-----Original Message-----
From: [log in to unmask]
To: [log in to unmask]
Cc: [log in to unmask]
Sent: 11/24/2003 11:41 PM
Subject: Re: spm.mat in spm2

Dear Tobias:

The SPM.mat files from SPM99 and SPM2 share only their name in common.
The internals of the file are entirely different. SPM2 also has no
facility for using the old SPM_fMRIDesMtx.mat files (which I also
thought
were convenient for repeating analyses).

The bottom line is you have to setup the analyses fresh for spm2, and
you
will have to setup the analyses new each time. There is no template-type
function in spm2.  There have been a number of messages on the list
regarding batching and automation of these issues.

Regards,
Darren

On Mon, 24 Nov 2003 20:21:17 -0000 "Egner, Tobias" wrote:

Dear all,

I am used to SPM99 and have only started recently to play around with
SPM2.
In the model specification and estimation process I am confused as to
how I
can use the same design matrix (formerly known as SPM_fMRIDesMtx.mat,

now

called SPM.mat) for estimating data for different subjects. When I
choose to
"specify data" with an SPM.mat that I have used before to estimate

another

subject's data, I do not seem to have the option to select new image

files

for the same matrix... Has the capability to estimate different data

sets

with the same specified model really disappeared with SPM2?

Many thanks,

Tobias Egner.




--------------------------------
Darren R. Gitelman, M.D.
Cognitive Neurology and the Alzheimer's Disease Center
Northwestern Univ., 320 E. Superior St., Searle 11-470, Chicago, IL
60611
Voice:   (312) 908-9023     Fax:  (312) 908-8789
--------------------------------