Dear Jerome,

I cannot reproduce your problem. [7 7] for the frequency window works
fine in both GUI and batch. If the images get written out then the
results module runs fine. It looks like your batch has several modules
without dependencies so the inversion module fails while the results
module runs. I tried putting [7 7] both in the results module and in
the inversion module where there is a possibility to define a
frequency window. In the results it works perfectly fine. In inversion
it technically runs but returns all NaNs  but I don't get the error
you got. Note that trying to get exactly 7Hz you might not get exactly
what you are asking for. The frequency resolution of any TF
decomposition is limited by the data length. If e.g. your trial length
is 500ms then your frequency bins will go from 2 Hz in increments of 2
so you'll have 6Hz and 8 Hz but not 7Hz. In spm_eeg_invert the
frequency window is set by taking those 'natural' frequency components
and then selecting the ones that are between the bounds you specified.
So in most cases there will be no component which is exactly 7Hz and
nothing will get selected, hence the NaNs. In the results module it's
slightly different so this problem doesn't happen, but still it might
be a good idea to have a window around your frequency of interest
rather than a single frequency.

Best,

Vladimir


On Tue, Oct 2, 2012 at 1:03 AM, Jérôme Courtemanche
<[log in to unmask]> wrote:
> I tried with a higher frequency and the same problem occurs. The images get
> written and I can do stats on them, but the error message steals me
> confidence in the results.
>
> Thanks for your input Dr Moratti,
>
> Jerome
>
> 2012/10/1 [log in to unmask] <[log in to unmask]>
>
>>
>> Are you sure it is because of [x x]? I did steady state localization at
>> 10Hz [10 10] and had not such problems.
>>
>> best,
>>
>> Stephan
>>
>> ________________________________________________________
>> Stephan Moratti, PhD
>>
>> see also: http://web.me.com/smoratti/
>>
>> Universidad Complutense de Madrid
>> Facultad de Psicología
>> Departamento de Psicología Básica I
>> Campus de Somosaguas
>> 28223 Pozuelo de Alarcón (Madrid)
>> Spain
>>
>> and
>>
>> Center for Biomedical Technology
>> Laboratory for Cognitive and Computational Neuroscience
>> Parque Científico y Tecnológico de la Universidad Politecnica de Madrid
>> Campus Montegancedo
>> 28223 Pozuelo de Alarcón (Madrid)
>> Spain
>>
>>
>> email: [log in to unmask]
>> Tel.:    +34 679219982
>>
>> El 01/10/2012, a las 18:09, Jérôme Courtemanche escribió:
>>
>> Hi all,
>>
>> In the MEG source inversion, the frequency of window of interest cannot
>> be [x x] but can be [x x+1]. If it is [x x], a window of error popped out:
>>
>> Error in job execution Error using ==> <a href="matlab:
>> opentoline('cfg_util.m,835,0)">cfg_util
>> at 835</a> Job execution failed. The full log of this
>> run can be found un MATLAB command window, starting with
>> the lines (look for the line showing the
>> exact #job as displayed in this error message)
>> _______
>> Running job #1
>> _______
>> In the matlab window, there is no message error, only this:
>> Done 'M/EEG inversion results'
>> The following modules did not run:
>> Failed: M/EEG source inversion
>>
>> So how could I localize precisely 7Hz activity, for instance?
>> Any help welcome,
>>
>> Jerome
>>
>>
>