[Eeglablist] epochconsist treats files with only one epoch as a continuous file

[Tarik S Bel-Bahar]

Hello Rich, some quick notes below. best wishes.


the STUDY should accept your files, but they need to be:
1 file per each condition for each subject.
Also each file needs to be epoched (so it cannot be continuous, and I
believe it needs to have at least 2 epochs, check the STUDY
documentation online).

Further you have to make sure that for each subject X condition file,.
the epochs within the file are the same length (epoch consistency
across files loaded into STUDY).
You may want to reread very closely the online eeglab STUDY
documentation, including how to look into the study structure and
where everything is stored.

For your current needs, I recommend you use eeg_regepoch (type help
eeg_regepoch)
to cut each of those 15 second chunks into 1 or 2 second chunks. Thus
there will be ~30 or 15 epochs within each file, instead of one long
chunk.
They will remain sequential and analyzable.

Further, unless you have a clear reason to use really long epochs
becuase of past published methods with a similar protocol,
it's probably best you cut up your 30 second blocks into smaller
periods, and thus you will have epochs, and thus STUDY should accept
them.
Try it this way first, and see how far you can get, then when that works,
think about how/why to work with long/single epochs.

You can also search on google scholar for people who used really long
single epochs like you with eeglab.

If you try/want to make eeglab do something it's not built to do, you
need a matlab-saavy person to build specific versions of the functions
for you.
You may also benefit from several other EEG analysis programs out
there that focus on "resting state continuous non-event-related data"







On Thu, Sep 29, 2016 at 9:54 AM, Ingram, Richard E - ingramre
<ingramre at jmu.edu> wrote:
> Hi all - I have a question about epoch consistency when using the STUDY
> feature.  I've seen 6 other posts on this but haven't seen a resolution.
> Can anyone help?
>
> I have collected continuous data throughout an experimental session.  I
> wrote software to display stimuli, capture performance data, and send marker
> from software to EEG data stream.
>
> The experiment structure has 8 30 second blocks per subject (120 blocks
> total).  (this is exploratory so subject N is low.)  I used EEGLAB epoch to
> extract the 30 sec from each block based on unique marker for a particular
> block.  So, for each subject, I have 15 30 sec blocks of continuous data.
> All are 3840 frames (30 x 128 sampling).
>
> I have tried several different iterations of setting up a STUDY but all say
> that epoch consistency is no.  I double checked allpnts etc as noted by Arno
> in another post and these parameters check out.  A couple of questions:
>
> Since I have continuous data only (no repeated epochs), does EEGLAB not
> recognize the data as an epoch and so determine epoch consistency = no based
> on that?  If so, would putting a copy of the 30 block data in the same file
> allow STUDY to see determine epoch consistency = yes.
>
> Also, while the data seem to be pretty clean overall, there are some
> segments identified for rejection.  However, if I reject the data, the
> epochs will be of different lengths and then would EEGLAB  determine epoch
> consistency = no because of unequal epoch length?  Sort of a catch 22.
>
> I know that EEGLAB is designed with ERP in mind but I'm trying to figure out
> how to best use it for analyzing continuous data.
>
> Thanks for any and all advice.
>
> Rich
> ________________________________
> From: eeglablist-bounces at sccn.ucsd.edu [eeglablist-bounces at sccn.ucsd.edu] on
> behalf of Makoto Miyakoshi [mmiyakoshi at ucsd.edu]
> Sent: Friday, March 18, 2016 3:32 AM
> To: Whitehead, Kimberley
> Cc: eeglablist at sccn.ucsd.edu
> Subject: Re: [Eeglablist] epochconsist treats files with only one epoch as a
> continuous file
>
> Dear Kim,
>
> Thank you for your suggestion. If you wish, would you please share your
> finding in EEGLAB Bugzilla as enhancement?
>
> https://sccn.ucsd.edu/bugzilla/enter_bug.cgi
>
> By the way I've encountered a similar situation. I learned that one cannot
> make a single epoch with EEGLAB. Unfortunately I forgot how I work around
> the issue (probably I repeated the epoch twice so that it is recognized as a
> one-time repeated two epochs?)
>
> Makoto
>
> On Wed, Jan 20, 2016 at 4:46 PM, Whitehead, Kimberley
> <k.whitehead at ucl.ac.uk> wrote:
>>
>> Hello
>>
>>
>>
>> I wonder if anybody could please offer some advice on altering the
>> eeg_checkset code so that it doesn’t automatically treat files with one
>> epoch as continuous. I think the issue is here:
>>
>>
>>
>> In eeg_checkset line 182:
>>
>> case 'epochconsist', % test epoch consistency
>>
>>                 % ----------------------
>>
>>                 res = 'no';
>>
>>                 datasettype = unique_bc( [ EEG.trials ] );
>>
>>                 if datasettype(1) == 1 & length(datasettype) == 1, return;
>> %
>>
>> continuous data
>>
>>                 elseif datasettype(1) == 1,                        return;
>> %
>>
>> continuous and epoch data
>>
>>                 end;
>>
>>
>>
>> The code says that if there is only 1 epoch, the file is continuous. This
>> means
>>
>> that if I try to include this file in a study in which most files have
>> several
>>
>> epochs, it says there is no epoch consistency and there is an error when I
>> try
>>
>> to do FFT analysis. However, the nature of my analysis means there are
>> many
>>
>> files with only 1 epoch. (My epochs are spontaneous transients in EEG
>> which
>>
>> I've zeropadded to make the same duration/same number of data points. I've
>>
>> split them up by topography so it's common that in 1 file there may be
>> just one
>>
>> with, for example, a right occipital topography.)
>>
>>
>>
>> Thanks, Kim
>>
>>
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> 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
>
>
>
>
> --
> Makoto Miyakoshi
> Swartz Center for Computational Neuroscience
> Institute for Neural Computation, University of California San Diego
>
> _______________________________________________
> 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