Dear Michiel,
we've been aware of this problem recently and it should now be solved.
To fix it, overwrite spm_getSPM.m of your SPM5 installation with the one
attached to this email - this will also be part of the next set of updates.
I hope this helps,
Guillaume.
Michiel de Ruiter wrote:
> Dear SPM experts,
>
> I encountered a problem using SPM5, update1111 (in combination with
> Matlab 7.1 and both Windows XP and Vista) when defining contrasts for my
> fMRI models using the GUI. When defining contrasts for different models
> in a row, the SPM.mat files for the models get mixed up. For instance, I
> first make a contrast for model 1. Accordingly, the SPM.mat file for
> model 1 is updated as can be verified by a change in date and time of
> the model. However, when I switch to model 2, both the SPM.mat files
> from model 1 AND 2 are updated. As a result, the SPM.mat files do not
> contain the right information anymore and I have to specify and estimate
> model 1 and 2 all over again.
>
>
> A colleague faces the same problems, on a independent system, with
> similar software specifications.
>
> This problem can be circumvented by closing SPM5 between specification
> of contrasts for different models. Also, it does not occur when using
> batches for contrast specification. Thanks in advance for your advise!
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|