[Eeglablist] Problem reading channel locations

arno arno at salk.edu
Tue Dec 13 18:01:02 PST 2005


Dear Alexander,

this error may also possibly arise because the neural network toolbox is 
instaled on your computer (and contains a function named "cell2mat") and 
it is before the EEGLAB function "cell2mat" in the path. In the next 
release, this function in EEGLAB has been renamed "celltomat" to avoid 
this conflict. Similarly the function "h" which is causing a problem 
with an "eval" statement in the brainstorm toolbox has been renamed "eegh".

If your problem persist after you have removed commas, try also changing 
the path order.

Best,

Arno

Alexander Roth wrote:

> Hi all,
>
> I have a channel position file that consists of the following entries:
>
> channel number, radius, theta, phi
>
> Here’s an extract of the first three lines of this file as an example:
>
> 1,1,105,60
>
> 2,1,84,63
>
> 3,1,63,75
>
> When I’m trying to import this data with the command
>
> readlocs('F:\RohEEG\SZ\NetzPos.pos','filetype','custom','format',{'channum' 
> 'radius' 'theta' 'sph_phi'})
>
> I’m getting the following error message:
>
> readlocs() warning: Fewer columns in the input than expected.
>
> See >> help readlocs
>
> readlocs(): Inserting electrode labels automatically.
>
> ??? Subscripted assignment dimension mismatch.
>
> Error in ==> cell2mat at 46
>
> M(i,j) = C{i,j};
>
> Error in ==> readlocs at 588
>
> allchannum = cell2mat( { eloc.channum } );
>
> What am I doing wrong? I specified four columns in the readlocs() 
> command, so why does eeglab complain about to few columns in the input 
> file?
>
> Regards
>
> Alexander Roth
>



More information about the eeglablist mailing list