[Eeglablist] I would like to train a data set with event information attached using EEGLAB using BCILAB ...

[Kaelasha Tyler]

Hi there,

I have some limited experience with EEGlab and the OpenBCI system.

For starters it looks like your .set file doesn't have any of the channel
locations in it. This is because the OpenBCI .txt file doesn't specify
these.
Did you use any specific montage or electrode positioning?
These would need to be added to the .set file in EEGlab by going to
Edit>Channel Locations and then adding the channel labels according to
where ever you have the OpenBCI leads positioned on your participants head.

I wasn't sure from your email whether you have already successfully opened
the data in EEGlab and been able to performing a range of simple analysis?
Such as a time frequency plot or even just a scroll plot of the raw data?
Or topographic plots?
Obviously getting familiar with a range of simple analysis of the OpenBCI
data set would be the the first step, before going to predictive learning
with BCILab which would be a step up in complexity.

I haven't had a chance to look at BCIlab yet, but hope to do so in the near
future.

Cheers,
Kaelasha





On Thu, Feb 6, 2020 at 9:01 PM <s1611341 at coins.tsukuba.ac.jp> wrote:

> Dear list
>
> I'm using this mailing list for the first time, so I'm sorry if I did
> something wrong.
>
> Also, I am not good at English, so it may be difficult to read sentences.
> I want to use BCILAB to learn the measured EEG.
> I measured EEG using OpenBCI.
> In openbci, brain wave data is output as .txt.
> Then, I sent the data to EEGLAB via MATLAB.
> So I attached the event information and saved it as a .set.
>
> Here are the questions.
>
> I tried to learn the model using the data of the .set according to the
> instruction on the BCILAB page.
> (https://sccn.ucsd.edu/wiki/BCILAB (https://sccn.ucsd.edu/wiki/BCILAB)
> Learning a Model)
> However, I got an error.
>
> -----------------------------------------------------
> Error using fieldnames
> Invalid input argument of type 'double'. Input must be a structure or a
> Java or COM
> object.
>
> Error in set_infer_chanlocs (line 58)
>  if isempty(fieldnames(locs)) && ~isempty(data.data)
>
> Error in io_loadset (line 422)
>  res = set_infer_chanlocs(res,allopts.montage_disambiguation); end
>
> Error in hlp_wrapresults (line 52)
>  [a{1:len}] = f(varargin{:});
>
> Error in exp_eval (line 131)
>  varargout = hlp_wrapresults(x.head,x.parts{:});
>
> Error in exp_beginfun (line 428)
>  parts{i} = exp_eval(parts{i},inf); %#ok<AGROW>
>
> Error in set_targetmarkers (line 87)
> if ~exp_beginfun('editing') return; end %#ok<SEPEX>
>
> Error in hlp_wrapresults (line 52)
>  [a{1:len}] = f(varargin{:});
>
> Error in exp_eval (line 131)
>  varargout = hlp_wrapresults(x.head,x.parts{:});
>
> Error in hlp_scope>@(f,a,frame__f1)feval(f,a{:})
>
> Error in hlp_scope (line 53)
>  [varargout{1:nargout}] = func(f,varargin);
>
> Error in exp_eval_optimized (line 52)
>  [varargout{1:nargout}] =
>  hlp_scope({'memoize',{'memory',1,'disk',double(unpartitioned)}},
>  @exp_eval,x,iters);
>
> Error in utl_check_bundle (line 42)
>  str{s} = exp_eval_optimized(str{s});
>
> Error in bci_train (line 784)
>  opts.data{k} = utl_check_bundle(opts.data{k}); end
>
> -------------------------------------------------------------------------------
> I don't know what to do because i'm ignorant...
>
> How can I train without this error?
>
> I would be glad if you could answer this question.
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu
>