[Eeglablist] std_erpimage error mesage

Thu Jun 18 10:35:38 PDT 2015

Dear Lucia,

> (I ask because the code you've written seems to be extracting the event
markers (ALLEEG.event.type) from all entries into a study, and I'm not sure
how that helps?)

Oh sorry you are right. It should be sum instead of find.

numTrialList = zeros(length(ALLEEG),3)) % if there are 3 conditions
for n = 1:length(ALLEEG)
tmpEEG = ALLEEG(1,n);
allEventTypes = {tmpEEG.event.type}';
numTrialList(n,1) = sum(strcmp(allEventTypes, 'event1')));
numTrialList(n,2) = sum(strcmp(allEventTypes, 'event2')));
numTrialList(n,3) = sum(strcmp(allEventTypes, 'event3')));
end
min(numTrialList(:))

> From your reply, it sounds like I need to know how many trials I have per
condition per subject (e.g. subject1 has 25 trials for condition1, subject2
has 20 trials for condition1 etc.).

Unless you are absolutely sure that any single condition from any subjects
has at least X trials and you'll use less than X number of lines for
erpimage.

This is simply because if you have 20 trials you can't visualize the
erpimage with > 20 lines. Subsampling is fine, but you can't upsample it.

> Does this mean that there is a minimum number of trials required for
doing ERP analysis in eeglab? What is the minimum number?

It's a separate question and as far as I know if you have a single trial
you can't perform 'erp analysis' i.e. can't compute average etc with EEGLAB.

> One of my subjects only has 5 trials for a condition - does that mean I
can't calculate ERPs using for that condition?

Because of this subject this condition, your STUDY ERPimage is limited to
use maximum of < 5 lines.

If you have minimum of 50 trials across all subjects all conditions, the
max ERPimage line is <50, and I'm not sure if it is 50-3 or 50-4...
something around it. It's not 50 or 50-1 for sure, I know one of my
colleagues tested it a few years ago.

Makoto

On Thu, Jun 18, 2015 at 9:22 AM, Li, Lucia M <lucia.li at imperial.ac.uk>
wrote:

>  Dear Makoto,
>
>  Thank you for taking the time to answer my questions.
>
>  From your reply, it sounds like I need to know how many trials I have
> per condition per subject (e.g. subject1 has 25 trials for condition1,
> subject2 has 20 trials for condition1 etc.). Does this mean that there is a
> minimum number of trials required for doing ERP analysis in eeglab? What is
> the minimum number? One of my subjects only has 5 trials for a condition -
> does that mean I can't calculate ERPs using for that condition?
>
>  Have I understood correctly? (I ask because the code you've written
> seems to be extracting the event markers (ALLEEG.event.type) from all
> entries into a study, and I'm not sure how that helps?)
>
>  Many thanks again & kind regards,
> Lucia
>  ------------------------------
> *From:* Makoto Miyakoshi [mmiyakoshi at ucsd.edu]
> *Sent:* Monday, June 15, 2015 6:23 PM
> *To:* Li, Lucia M; Ramón Martinez
> *Cc:* eeglablist at sccn.ucsd.edu
> *Subject:* Re: [Eeglablist] std_erpimage error mesage
>
>   Dear Lucia,
>
>  You need to know (or the application should figure it out) the minimum
> number of available trials across all conditions and subjects. To code it,
> it should be something like this:
>
>  numTrialList = zeros(length(ALLEEG),3)) % if there are 3 conditions
> for n = 1:length(ALLEEG)
> tmpEEG = ALLEEG(1,n);
> allEventTypes = {tmpEEG.event.type}';
> numTrialList(n,1) = find(strcmp(allEventTypes, 'event1')));
>  numTrialList(n,2) = find(strcmp(allEventTypes, 'event2')));
>  numTrialList(n,3) = find(strcmp(allEventTypes, 'event3')));
>  end
> min(numTrialList(:))
>
>  I wrote it without opening Matlab so there may be mistake.
>
>  I believe this should be really implemented in EEGLAB so that the
> application suggests the upper bound. Ramon, would you please add it to
> Bugzilla?
>
>  Makoto
>
>
>
>
> On Mon, Jun 8, 2015 at 8:40 AM, Li, Lucia M <lucia.li at imperial.ac.uk>
> wrote:
>
>>  Dear eeglab users & experts,
>>
>>  I was wondering if someone might be able to help me decipher this error
>> message.
>>
>>  I have datasets from 12 people which I wish to combine into a study.
>> There are 15 different bin types.
>>
>>  Having created my study by loading all the datasets in, I tried to
>> "precompute channel measures". I attempted to pre-compute: ERP, Power
>> spectrum, ERP-image, ERSP & ITCs. I clicked 'Test' as well to make sure
>> that the parameters I entered produced sensible results. All other options
>> I left as default ('Spherical interpolation' ticked, 'Remove ICA
>> artifactual components pre-tagged' and 'Remove artifactual ICA cluster'
>> unticked).
>>
>>  Unfortunately, the pre-compute throws up this error message:
>>
>>  "The default number of ERPimage lines is too large for one of the
>> dataset, {Error occurred in function std_erpimage() at line 169'
>>
>>  I'm not entirely sure what the problem is. I have also tried to
>> pre-compute with fewer bin types and it's worked perfectly.
>>
>>  My two thoughts about why the full study won't precompute properly:
>> - some of the bins overlap e.g. a bin for responses from the first half
>> of the task, and a bin for left hand responses from the first half of the
>> task
>> - some bins have <10 trials
>>
>>  Any thoughts and help on this matter greatly appreciated!
>>
>>  Kind regards,
>> Lucia
>>
>>  ---
>> L M Li
>> Wellcome Trust Clinical Fellow
>> Imperial College London
>>
>> _______________________________________________
>> 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
>

-- 
Makoto Miyakoshi
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150618/17b0b7f0/attachment.html>