[Eeglablist] ERSP using baseline from a different condition

[Makoto Miyakoshi]

Dear Nikola,

Probably XXX_erspbase. I don't think "powbase" takes raw values for
computation, but most likely they just specify the time range(s) used for
baseline.

Well, however, it's actually more than pain in the neck to figure out the
link between those files and the combination of subject and condition on
STUDY.

If I were you I would do this.
1. go to edit/plot cluster and show all ERSPs and press ok to exit from
edit/plot cluster; this updates your STUDY on the workspace.
2. Click STUDY-clusters- {1,n} to open nth cluster in the workspace.
3. There you should be seeing something like 'icaersp' which has the size
of {x,y}; this corresponds to your STUDY design so carefully check which is
which if you have two independent variables in your design.
4. Also you should be finding 'ersptime' 'freqs'.
5. With these you can identify cluster, combination of conditions, time
scale, and freq scale. Thus you can compute your baseline period.
6. To visualilze your ERSP,
figure; imagesc(x); axis xy
You should be also able to feed your time and freq scales to the plot.

Try this hack. I think this is the easiest way to do what you want to do.

Makoto


2013/5/9 Nikola Vukovic <vukovicnikola at gmail.com>

> Dear Makoto,
>
> Thank you for your reply and advice. Let me see if I understand you
> correctly - I should load the reference condition's .icaersp files for all
> participants into Matlab, and then extract and average the compX_erspboot
> variables (http://d.pr/i/aWPm)? This variable could then be used as the
> argument of "powbase" in subsequent ERSP calculations?
>
> Nikola
>
>
> On Fri, May 10, 2013 at 2:06 AM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>wrote:
>
>> Dear Nikola,
>>
>> This requires manual file operation outside EEGLAB.
>> When you precompute ersp, EEGLAB generates design1_xxxxxx.icaersp. You
>> can open them by 'load disign1_xxxxxx.icaersp -mat' and can find baseline
>> values. Sorry this is only minimum information for the solution.
>>
>> Makoto
>>
>>
>> 2013/5/2 Nikola Vukovic <vukovicnikola at gmail.com>
>>
>>> Dear EEGlab community,
>>>
>>> I am a beginner EEGlab user, and have a problem which was brought up a
>>> couple of times before on the list, but for which I wasn't able to find a
>>> clear solution.
>>>
>>> I'm working with a STUDY design, which contains datasets decomposed
>>> using ICA, and their various components already clustered. I would like to
>>> find a way to calculate ERSPs for a cluster where the baseline for my
>>> experimental conditions of interest is taken from a separate condition (the
>>> "reference" condition).
>>>
>>> So far, looking at the mailing list archives, I've been able to gather
>>> that one can use the powbase variable to specify a baseline derived from
>>> some other condition. However, all the posts I found also report issues
>>> with being unable to get the format of the powbase matrix right, or
>>> otherwise getting Matlab errors with their customized script.
>>>
>>> I would be extremely grateful to anyone who can provide hints or scripts
>>> which would help me solve this issue. Saying that I'm inexperienced with
>>> eeglab coding would probably be the understatement of the year, but I am
>>> willing to follow any leads as to how to proceed. I presume there must be a
>>> way to calculate ERSP for a component cluster, selecting only my reference
>>> condition in the Study design, and then from this somehow extract the
>>> baseline values which could be used for calculating ERSP in other
>>> conditions?
>>>
>>> Best wishes,
>>> Nikola
>>>
>>> _______________________________________________
>>> 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
>>
>
>
>
> --
> Nikola Vukovic
> vukovicnikola at gmail.com
> www.vukovicnikola.info
> +447412032927
>



-- 
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/20130510/c547aac9/attachment.html>