Another possibly helpful piece of advice is to add toolboxes to the
bottom of the Matlab path. I think this will stop most toolboxes from
overloading the spm functions. I've had similar problems with Marsbar
before, so I do this out of habit now.
-Neil
Guillaume Flandin wrote:
> Dear Yannis,
>
> this could indeed explain the error you observed.
>
> Two general piece of advice:
> 1/ always use "Add Folder..." and not "Add with Subfolders..." when
> adding SPM to Matlab path.
> 2/ try to only include SPM toolboxes in Matlab path when you are
> actually using them, especially those that overload some legitimate SPM
> functions.
>
> Best regards,
> Guillaume.
>
>
> Paloyelis, Yannis wrote:
>
>> Dear Guillaume and Donald,
>>
>> Thanks very much for your responses. There were a few other instances of the Maximum recursion limit reached problem on the Matlab support site, and trying out the solution described here:
>>
>> www.mathworks.com/support/solutions/data/1-1TV6VZ.html?solution=1-1TV
>>
>> has worked for me. Solving this problem seems to have made the first problem disappear (fingers crossed!).
>>
>> In my path, I had explicitly included all the subfolders of the marsbar folder; removing them (apparently not needed to be explicitly defined even for marsbar) appears to have done it. (One of the marsbar subfolders, which was at the top of the path, apparently contained some files for earlier SPM versions, which might be the source of the problem).
>>
>> Thanks very much for our help!
>>
>> Best ishes,
>> Yannis
>>
>>
>> ________________________________________
>> From: Guillaume Flandin [[log in to unmask]]
>> Sent: 19 May 2009 19:34
>> To: Paloyelis, Yannis
>> Cc: [log in to unmask]
>> Subject: Re: [SPM] spm contrast error
>>
>> Dear Yannis,
>>
>> the first error you encountered has been reported a couple of time:
>> https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=SPM;yiAskQ;20090513160651%2B0200
>> https://www.jiscmail.ac.uk/cgi-bin/wa.exe?A2=SPM;umWLDA;20080516095323%2B0100
>> I can reproduce the error under some circumstances when the callback to
>> display the contrast on top of the design matrix is triggered and the
>> contrast as not yet been defined but it should not prevent you from
>> entering a valid contrast and submit it. Is the error display blocking?
>>
>> Concerning your second error, it is rather surprising as it occurs in
>> spm_create_image, which is not part of SPM anymore since SPM2. Would you
>> have some leftovers of SPM99 somewhere in your Matlab path?
>>
>> Best regards,
>> Guillaume.
>>
>> Yannis Paloyelis wrote:
>>
>>> Dear all,
>>>
>>> I am getting some error messages which I cannot explain - hopefully you
>>> could help. This is a new installation of SPM5 and recent update on a CentOS
>>> 64 bit cluster, using MATLAB R2008b. I run sessions connecting with ssh -X
>>> from a Linux 9.04 terminal. Preprocessing, design estimation and creation of
>>> contrast images have all run smoothly.
>>>
>>> I estimate the SPM.mat for a two-sample t-test. I open Results to define
>>> contrasts (for group comparisons). I get the following message when I try to
>>> define a legitimate T-contrast:
>>>
>>> +++++++++++++++++++++++++++++++++++++++++++++++++
>>>
>>> SPM5: spm_results_ui (v$Rev: 1776 $) 21:24:34 - 18/05/2009
>>> ========================================================================
>>> ??? Error using ==> mrdivide
>>> Matrix dimensions must agree.
>>>
>>> Error in ==> spm_conman at 1121
>>> h = image((xCon(i).c'/max(abs(xCon(i).c(:)))+1)*32);
>>>
>>> Error in ==> spm_conman at 1692
>>> spm_conman('GraphCons',DxCon,-1,F) %-Depict contrast
>>>
>>> ??? Error using ==> waitfor
>>> Error while evaluating uicontrol Callback
>>>
>>> ++++++++++++++++++++++++++++++++++++++++++++
>>>
>>> Then, when I try to run an F-contrast, I get this error:
>>>
>>> ++++++++++++++++++++++++++++++++++++++++++++++++
>>> ESS image 1 : ??? Maximum recursion limit of 500
>>> reached. Use set(0,'RecursionLimit',N)
>>> to change the limit. Be aware that exceeding your available stack space can
>>> crash MATLAB and/or your computer.
>>>
>>> Error in ==> spm_create_image
>>>
>>>
>>> ??? Error while evaluating uicontrol Callback
>>>
>>>
>>>
>>>
>>>
>>> I hope someone could help me!
>>>
>>> Kind regards, and many thanks for your time!
>>>
>>> Yannis
>>>
>
>
>
|