redefinetrial_offset option

Fri Apr 16 15:02:07 CEST 2010

Hi Natalia,

The 'offset'-column in a trl-matrix tells you how many points you have  
to move away from t=0, to get to the begin of the trial. In other  
words, for a given trial, if the offset value is negative, this means  
that you have to move X samples to the right to get to time point t=0.  
Consequently, the first value of the corresponding time-axis is  
negative. Important to keep in mind is that the third column of the  
trl-matrix defines the 'local time axis' of your epoch of interest,  
whereas the first two columns represent the begin and end sample of  
the epoch, fixed to the 'absolute recording time'. This means that if  
you only want to shift the local time axis, those first columns should  
not change.
Now, if you want to realign the time axes of your epochs of interest  
according to the vector you described (a negative value of -51 meaning  
that the sub-event occurred before the trigger event) this implies  
that the newly defined t=0 moves to the left, and that implies that  
you have to add 51 samples to the third element in the trl-row.

Hope this helps,

Jan-Mathijs

On Apr 16, 2010, at 2:42 PM, Natalia Grion wrote:

> Hi Michael,
> When I wrote possible solution i meant: -(abs(trl(:,3))+(-51)) =  
> -5958,
> which is conceptually different from -6060.
> In "definetrial" function when declaring trl(:,3)= - 6009 means that
> 1)trial start before trigger event (negative sign) and 2)the offset of
> trigger is 6009 samples away with respect to the "begining" of the  
> trial:
> in sum: that the trial starts 6009 steps before trigger. If the sub- 
> event
> to which i want to realign happens 51 steps before trigger, then  
> number of
> steps with respect to t=0 are "-" 51, and this is applied when  
> defining
> data.time in the code of redefinetrial. But when redefining trl(:,3),
> trigger  should get "closer" to beginning of trial, as beginning of  
> trial
> is still the old one. If what i' saying is correct, then also trl(:, 
> 1) and
> (:,2) has to be modified relative to the new "sub-event".
> Any reply will be great.
> Natalia
>
>
>> Hi Natalie,
>>
>> may be I am missing something but the shift by 51 samples implied by
>>
>> (-)6009 +(-51)
>>
>> is always to the "left" as you call it - irrespective of the sign of
> "6009":
>>
>> 6009 +(-51) = 5958 < 6009 (left shift)
>>
>> -6009 +(-51) = -6060 < -6009 (left shift)
>>
>>
>> Michael
>>
>> -----UrsprÃ¼ngliche Nachricht-----
>> Von: Natalia Grion <grion at SISSA.IT>
>> Gesendet: Apr 16, 2010 1:09:28 PM
>> An: FIELDTRIP at NIC.SURFNET.NL
>> Betreff: [FIELDTRIP] redefinetrial_offset option
>>
>>> Hi all,
>>>        When inspecting in detail "redefinetrial" function I found a
>>> striking point, I think there is an error on the code that could be
> easily solved but maybe i'm missing something:
>>> My trials are of fixed length: they go from -3sec to +3sec with 1  
>>> trigger
>>> event as t=0. I want to realign the trials to a new (sub)event, so I
> defined offset as: Nxsamples relative to t=0. For each trial, offset  
> has
> different signs as the event happened either before trigger event
> (-#samples) or after it (+#samples).
>>> In the code: when having for example -51 (samples relative to  
>>> trigger)
> data.time is shifted to the left, and this would be correct. But when
> correcting "trial definition" this offset is summed to trl(:,3); the
>>> point
>>> is: my trl(:,3) is negative since is indicating  that the trial  
>>> begins
> before the trigger, (-)6009 +(-51) results in shifting the offset of
> trigger to the right which is not the case: (possible solution:
>>> abs(trl(:,3))+(-51).). Then: trl(:,1),trl(:,2) correction is  
>>> omitted,
>>> why?
>>> as "event" changed, shouldn't beg and ensample follow this change?
> sticking to +/-3sec defined as star/end of trial? In fact, data.time  
> was
> shift. In the rest of the code i don't see any line related to this
> change.
>>> Any help will be welcome!
>>> Natalia
>>> .............................
>>> elseif ~isempty(cfg.offset)
>>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
>>> %%%%%%%%%%%
>>> % shift the time axis from each trial
>>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 
>>> %%%%%%%%%%%
>>> offset = cfg.offset(:);
>>> if length(cfg.offset)==1
>>>   offset = repmat(offset, Ntrial, 1);
>>> end
>>> for i=1:Ntrial
>>>   data.time{i} = data.time{i} + offset(i)/data.fsample;
>>> end
>>> % also correct the trial definition
>>> if ~isempty(trl)
>>>   trl(:,3) = trl(:,3) + offset;
>>> end
>>> ........................................
>>> ----------------------------------------------------------------
>>> SISSA Webmail https://webmail.sissa.it/
>>> Powered by SquirrelMail http://www.squirrelmail.org/
>>> ----------------------------------------------------------------
>>> SISSA Webmail https://webmail.sissa.it/
>>> Powered by SquirrelMail http://www.squirrelmail.org/
>>> ----------------------------------
>>> The aim of this list is to facilitate the discussion between users  
>>> of
> the
>>> FieldTrip  toolbox, to share experiences and to discuss new ideas  
>>> for
> MEG
>>> and EEG analysis. See also
>>> http://listserv.surfnet.nl/archives/fieldtrip.html and
>>> http://www.ru.nl/neuroimaging/fieldtrip.
>>
>> ----------------------------------
>> The aim of this list is to facilitate the discussion between users of
> the
>> FieldTrip  toolbox, to share experiences and to discuss new ideas for
> MEG
>> and EEG analysis. See also
>> http://listserv.surfnet.nl/archives/fieldtrip.html and
>> http://www.ru.nl/neuroimaging/fieldtrip.
>
>
> ----------------------------------------------------------------
>  SISSA Webmail https://webmail.sissa.it/
>  Powered by SquirrelMail http://www.squirrelmail.org/
>
>
>
>
>
> ----------------------------------------------------------------
>  SISSA Webmail https://webmail.sissa.it/
>  Powered by SquirrelMail http://www.squirrelmail.org/
>
> ----------------------------------
> The aim of this list is to facilitate the discussion between users  
> of the FieldTrip  toolbox, to share experiences and to discuss new  
> ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html 
>  and http://www.ru.nl/neuroimaging/fieldtrip.

Dr. J.M. (Jan-Mathijs) Schoffelen
Donders Institute for Brain, Cognition and Behaviour,
Centre for Cognitive Neuroimaging,
Radboud University Nijmegen, The Netherlands
J.Schoffelen at donders.ru.nl
Telephone: 0031-24-3668063

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/neuroimaging/fieldtrip.