Dear Virginia,
thanks for spotting this - there is something going wrong with this
concatenation line. Try replacing
con = [c1 con(c).c zeros(1,covs(s))];
with
c1 = [c1 con(c).c zeros(1,covs(s))];
Now, c1 will accumulate contrast weights over all sessions and con(c).c
will be changed outside this loop. This should produce a proper
concatenation. I will fix this for the next set of updates.
Volkmar
On Thu, 23 Aug 2007, <Virginia> <Flanagin> wrote:
> Hi,
>
> I recently tried to use the Factorial Design option with fMRI Model
> specification to automatically create the f-contrasts necessary for the
> appropriate first and second level analysis. The specification and
> estimation of the model worked fine but when SPM5 tried to automatically
> create the contrasts I received the following message: "Improper matrix
> reference" which was an error in spm_design_contrasts line 83
>
> I looked around in spm_design_contrasts and found the for loop:
> for s=2:nsess
> con = [c1 con(c).c zeros(1,covs(s))];
> end
> It looks like a normal array is created with the same name as a structural
> array that is then called again in the for loop, and I think that is where
> the problem is coming from.
> From the rest of the code it seems like one just needs to change the
> parentheses but I do not really want to mess around with the spm programs
> without posing the problem here first.
> Has anyone else had this problem or is it possible to fix it?
> Thanks,
> Virginia
>
--
Volkmar Glauche
-
Department of Neurology [log in to unmask]
Universitaetsklinikum Freiburg Phone 49(0)761-270-5331
Breisacher Str. 64 Fax 49(0)761-270-5416
79106 Freiburg http://fbi.uniklinik-freiburg.de/
|