[FieldTrip] problem with ft_multiplot and planar gradient 

[Michael Wibral]

Dear Luisa,

I was also wondering whether you were aware of the fact that planar gradients of group data can look very different depending on when you do the averaging and the megplanar/combineplanar steps. This is because combineplanar is a nonlinear step that cannot be exchanged with the other steps without changing the results.
So creating planar gradients in single trials and then using combineplanar and then averaging trials within subjects and then over subjects will look pretty different from doing all averaging first (trials, subjects) and then using megplanar/combineplanar to get an planar gradient representation of this average.

Michael


-----Ursprüngliche Nachricht-----
Von: "Luisa Frei" <l.frei at psy.gla.ac.uk>
Gesendet: Feb 3, 2011 7:07:44 PM
An: "Email discussion list for the FieldTrip project" <fieldtrip at donders.ru.nl>
Betreff: [FieldTrip] problem with ft_multiplot and planar gradient  ?

>Hi everybody,
>I have a problem that impacts several analysis steps and I'm hoping  
>someone can point me to the right direction.
>
>I have 13 recording sessions, that I want to average, and then I want  
>to compute the planar gradients of the average. I first average the  
>sessions with ft_timelockanalysis for each session. Then I use  
>ft_timelockgrandaverage to combine the sessions.
>Then, just to check the data, I plot the ERFs with ft_multiplotER.  
>This is when the first problem arises. The x and y axes of the plots  
>are shifted in space somehow (see picture 3.png).
>Next, I compute the planar gradients. When I plot the topographies  
>for the planar gradient, I get a very strange activation pattern,  
>that has nothing to do with the ERF topography (which, btw has no  
>obvious flaws) and I also get several warnings when I make the plot  
>(Picture4.png).
>
>I also know for a fact that the grand average planar gradient should  
>look different, as I computed it once before by hand, and it looked  
>much more sensible.
>
>If anyone has ever seen this before and can give me a hint as to  
>where to look for the error, I would be very grateful.
>
>Luisa
>
>
>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Michael Wibral.vcf
Type: text/x-vcard
Size: 637 bytes
Desc: not available
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20110204/147f7834/attachment-0001.vcf>