Hi,
I'm not sure if this is the same problem, but using SPM2 in Matlab 7.0.4
we were unable to define contrasts. The box for defining the contrast
vector would not allow any input. We solved this by turning off the
java virtual machine. Have you tried starting matab with "matlab -nojvm" ?
Hope this helps,
Mark
Kirsch, Peter wrote:
>Johannes,
>
>All your problems have to do with MATLAB 7. The warning does not influence
>your results. I also do not like it but I learned to live with it. The
>update file in the SPM2-update folder which says that it fixes the problem
>does not work on my machine.
>However, the contrast problem is more severe since it prevents that you can
>calculate t-contrast. I figured out that the problem is only present when
>you use Matlab Release 14 Service Pack 2. After updating my Matlab 7 to SP2
>the problem occurred and after downgrading to SP1 it disappeared again. As
>fas as I see, uninstalling SP2 is the easiest way to get rid of it.
>
>Best wishes,
>
>Peter
>
>-----Ursprüngliche Nachricht-----
>Von: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] Im
>Auftrag von Johannes Gerber
>Gesendet: Mittwoch, 1. Juni 2005 19:22
>An: [log in to unmask]
>Betreff: [SPM] Failure on contrast defining
>
>
>Dear all,
>
>a week ago Gang Chen wrote under the above mentioned subject:
>
>
>
>>I have trouble defining contrast tests in SPM2. For example, when I run
>>a
>>
>>
>one-sample t test, after I specify a t-contrast
>
>
>>on the SPM contrast manager by typing "1" in the window of "contrast
>>
>>
>weights vector" and click "... submit", there is no
>
>
>>response. And at the bottom of the manager window, it says "name
>>defined,
>>
>>
>contrast not defined".
>
>
>>If I switch to F-contrast, it still doesn't work if I specify a "1" in
>>
>>
>the "contrast weights matrix" window. Only if I
>
>
>>define the contrast by typing 0 in the "columns to reduced desig"
>>window
>>
>>
>can I get the contrast successfully defined.
>
>
>>The same troulbe also occured when I was running contrasts for
>>individual
>>
>>
>subject analysis. Why is this?
>
>and as far as I can see his (or her? sorry for my ignorance) questions
>haven't been answered yet. how sad! because I am also desperately waiting
>for the answer, having the same problem after an update of MATLAB and SPM!
>
>more precisely: everything worked fine when using matlab 6.5 (R13) and SPM2
>(with bugs fixed until to 12/2004) on Windows XP. after the spring cleaning
>of my computer I now have Matlab 7 (R14, without any toolboxes as before)
>and a newly updated SPM2, again on Windows XP prof.
>
>During my (once upon a time) normal routine analysis I now get an awful lot
>of ERROR MESSAGES to which I am not accustomed.
>
>when ESTIMATING, SPM warns of the following (for all planes):
>
>....
>
>Plane 26/26 , block 5/7 :
>Warning: The argument for the %s format specifier must be of type char (a
>string).
>
>
>>In spm_s ...read & mask dataWarning: The argument for the %s
>>
>>
>format specifier must be of type char (a string).
>
>
>>In spm_s filteringWarning: The argument for the %s
>>
>>
>format specifier must be of type char (a string).
>
>
>>In spm_s estimation
>>
>>
>Plane 26/26 , block 6/7 :
>Warning: The argument for the %s format specifier must be of type char (a
>string).
>
>
>>In spm_s ...read & mask data
>>
>>
>Plane 26/26 , block 7/7 :
>Warning: The argument for the %s format specifier must be of type char (a
>string).
>
>
>>In spm_s ...read & mask dataWarning: The argument for the %s
>>
>>
>format specifier must be of type char (a string).
>
>
>>In spm_s ...saving planeWarning: The argument for the %s
>>
>>
>format specifier must be of type char (a string).
>
>
>>In spm_s ...done
>>
>>
>Warning: The argument for the %s format specifier must be of type char (a
>string).
>
>
>>In spm_s ...closing files
>>
>>
>Spatial non-sphericity (over scans) : ...estimating derivatives
>Spatial non-sphericity (over scans) : ...writing resels/voxel image
>Warning: The argument for the %s format specifier must be of type char (a
>string).
>
>
>>In spm_s ...done
>>
>>
>Saving
>results : ...writingWarning:
>The argument for the %s format specifier must be of type char (a string).
>
>
>>In spm_s ...done
>>
>>
>Completed : 18:33:49 - 01/06/2005
>...use the results section for assessment
>
>
>when pressing RESULTS, the following error messages appear (again for all
>planes):
>
>...
>
>
>
>>In spm_resss at 105
>>
>>
> In spm_contrasts at 130
> In spm_getSPM at 393
> In spm_re...plane 26/26 Warning: The argument for the %s format specifier
>must be of type char (a string).
>
>
>>In spm_resss at 124
>>
>>
> In spm_contrasts at 130
> In spm_getSPM at 3 ...written ess_0001.img
> spm{F} image
>1 : ...computingWarning: The argument for
>the %s format specifier must be of type char (a string).
>
>
>>In spm_contrasts at 198
>>
>>
> In spm_getSPM at 3 ...writingWarning: The argument for
>the %s format specifier must be of type char (a string).
>
>
>>In spm_contrasts at 216
>>
>>
> In spm_getSPM at 3 ...written spmF_0001.img
> SPM computation : ...initialising
> SPM computation : ...done
>
>
>
>
>and after that, when invoking the contrast builder, the problems described
>by Gang show up: after having typed the contrast vector, SPM would ignore
>the entries and post a warning message (cf. attachment).
>
>could anyone help with this complex problem (or, in the end it is very
>simple?) which seems to be related to updating MAtlab and/or SPM2?
>
>sorry for the inconvenience, but I am in dire straits!
>
>thank you!
>
>Johannes
>
>
|