Dear Xin,
if you have a toolbox in your path that takes precedence of the SPM
function spm_getSPM.m then the bug can still appear. Type:
>> which spm_getSPM -all
to check if that is the case. If so, remove that toolbox from your
Matlab path or patch it like this:
https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=SPM;vDpQZw;20080618180823%2B0100
Best regards,
Guillaume.
Xin Zheng wrote:
> Hi Guillaume,
>
> I forgot to mention that we already have the latest updates (3381 I
> believe). Are there specific files we can look at to help troubleshoot
> this problem?
>
> Thanks,
> Xin
>
> On Wed, Oct 7, 2009 at 5:56 AM, Guillaume Flandin
> <[log in to unmask]> wrote:
>> Dear Xin,
>>
>> what you describe seems to be symptomatic of a bug that was present in
>> updates 1111 of SPM5 so I would suggest that you install the most recent
>> updates from:
>> ftp://ftp.fil.ion.ucl.ac.uk/spm/spm5_updates/
>> so that newer files overwrite existing files.
>> Note that if it is indeed the cause of the problem you describe, you
>> will have to specify and estimate your models again as the SPM.mat has
>> been somehow corrupted.
>>
>> Best regards,
>> Guillaume.
>>
>>
>> Xin Zheng wrote:
>>> Dear SPM users.
>>>
>>> We have encountered a persistent problem with SPM5 contrast manager.
>>> We first ran the model and set up the contrast for analysis 1. We then
>>> opened another analysis (analysis 2) that was set up by another
>>> computer in the network. When we opened the contrast manager for
>>> analysis 1 again, only the contrasts pertaining to analysis 2 (not 1)
>>> popped up in the GUI. However, when we checked the directory where
>>> analysis 1 was stored, it seemed to contain all the constrasts we
>>> originally set up.
>>>
>>> We also noticed that in the SPM.mat file in analysis 1, the 'swd'
>>> variable pointed to the directory for analysis 2. Is this a path issue
>>> in SPM?
>>>
>>> We have tried restarting SPM - but it's still not showing the data for
>>> analysis 1. All the analyses were done in their own directories.
>>>
>>> Has anyone else encountered this problem? Any help would be much appreciated.
>>>
>>> Thanks,
>>> Xin & Catherine
>> --
>> Guillaume Flandin, PhD
>> Wellcome Trust Centre for Neuroimaging
>> University College London
>> 12 Queen Square
>> London WC1N 3BG
>>
>>
>
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|