Elena
SPM should have saved the files in Matlab 6 format, but perhaps spm2 did not
do this. Are these spm99 or spm2 analyses by the way? In any case I would
guess that the mat files got saved as the Matlab version 7. I guess the
first question is why not use R2006a, but I guess you would have done that
if you could have.
However, to fix this you will need a copy of Matlab version 7. After
starting it up with the desktop visible go to File, Preferences, Mat files
and change the preference to saving as Matlab version 5 or greater. Make a
backup of all the SPM.mat files. Go to where an SPM.mat file is. Load it
(load SPM.mat). Then resave it: save('SPM.mat','SPM','-v6').
I'm not sure why you cannot run any group analyses. Perhaps the permissions
on the files got changed by your colleagues? Hopefully you did not
permanently delete the various con and spmT files. These should work fine.
I'm not sure about the xCon.mat file. I thought that file was for spm99 and
not spm2. (if you are in fact trying to add contrasts using spm2 to data
that was analyzed in spm99 that won't work at all, and it has nothing to do
with matlab. The data structures changed between those versions of spm.)
hth,
darren
> -----Original Message-----
> From: SPM (Statistical Parametric Mapping)
> [mailto:[log in to unmask]] On Behalf Of Antonova, Elena
> Sent: Friday, November 16, 2007 11:27 AM
> To: [log in to unmask]
> Subject: [SPM] SPM2, Matlab7.2 and SPM.mat
>
> Dear SPMers,
>
> We have added a new t-contrast using Matlab 7.2 (R2006a) and
> SPM2 contrast manager. All the model building and previous
> contrasts were made using Matlab 6.1. This was done
> unintentionally (through an oversight my colleagues used the
> wrong version of matlab)
>
> Now we cannot open SPM.mat files using Matlab6.1. The error
> message comes up saying that the file might be corrupted. We
> can open them using SPM2 GUI in Matlab 7.2., but cannot run
> any group analysis using the t-contrasts (old or new).
>
> I have tried deleting all the con* and spmT* files from the
> directories, but still cannot open SPM.mat file in Matlab6
> with the same error (By the way, I do not seem to have
> xCon.mat files in any of the directories.)
>
> Is there a way to 'restore' the SPM.mat files? The design is
> individual for each subject, with a large number of
> contrasts, so it will be extremely unfortunate if we have to
> start building the models from the beginning. (we have usable
> SPM.mat back-up files for some, but not all subjects)
>
> Many thanks in advance for any help.
>
> best wishes,
>
> Lena
>
> P Before printing, think about the environment
> *********************************************
> Elena Antonova, PhD
> Post-Doctoral Research Worker
> PO78 Department of Psychology
> Institute of Psychiatry
> King's College London
> De Crespigny Park
> London SE5 8AF UK
> tel: +44 (0)20 7848 0978
> fax: +44 (0)20 7848 0860
> email: [log in to unmask] <mailto:[log in to unmask]>
>
>
>
|