[Eeglablist] baseline correction problem

[Arnaud Delorme]

Dear Baris and others,

the filter at the STUDY level is applied for display only. 
It is not taken into account in the statistics.
As Baris mentions, you need to filter the datasets themselves to filter the data.
Best,

Arno

On Apr 8, 2012, at 6:34 PM, Baris Demiral wrote:

> If you filter the epoched data in or out of the the study display,
> this naturally influences the baseline. If EEGLAB developers put
> another function accompanying the filter function in the display GUI
> -which automatically takes the baseline out- this may solve the
> problem... but if they do not implement this, the only way is to
> filter the data and save it with a separate filename (i.e.,
> 'presentation.set') and then create the study file over these smoothed
> data. This will take more space in the hard disk of course.
> Baris
> 
> On Thu, Mar 29, 2012 at 3:46 PM, mirek <3mirek at gmail.com> wrote:
>> Hi,
>> 
>> Thank you for your remarks. I took them into consideration and also
>> found what Makoto mentioned, that the offset is actually caused by the
>> 'LP display fiter' in the 'Set ERP plotting parameters'. Turning it off
>> disables that nasty baseline effect. But still the question remains
>> (relatively smaller, happily): can this filter be used without seriously
>> affecting the baseline?
>> 
>> Thanks for all
>> 
>> Mirek
>> 
>> 
>> On 29.03.2012 20:37, Makoto Miyakoshi wrote:
>>> Dear Mirek and Arno,
>>> 
>>> If filter option for ERP is applied, it changes baseline. It looks
>>> like filter is applied AFTER baseline correction.
>>> Arno, could you confirm it?
>>> 
>>> Makoto
>>> 
>>> 2012/3/26 Arnaud Delorme<arno at ucsd.edu>:
>>>> Dear Mirek,
>>>> 
>>>> if you doubt about baseline correction, you can do baseline correction in every single dataset manually and then ignore the baseline correction option when you compute ERP at the STUDY level. You may also subtract components in individual datasets before computing ERPs in the STUDY interface. Finally, at the STUDY level, you can then plot the ERP of each individual subject. This way you might be able to resolve your dilemma and find out where the offset comes from.
>>>> Hope this helps,
>>>> 
>>>> Arno
>>>> 
>>>> On Mar 24, 2012, at 4:12 PM, mirek wrote:
>>>> 
>>>>> Hi all,
>>>>> 
>>>>> The question seems to be quite basic, but still unresolved for me...
>>>>> 
>>>>> I work (or trying to) using study/designs to compute and plot ERPs. The
>>>>> problem is with the baseline. The plotted ERPs are definitely not
>>>>> properly baseline corrected, however the prestimuls baseline interval is
>>>>> set when precomputing channel measures (i.e.: "[-100 0]", since I need
>>>>> -100 to 0 ms correction).
>>>>> 
>>>>> I've checked whether the correction works to find out, that it does, but
>>>>> in unpredictable (for me) manner. Depending on the interval I set for
>>>>> correction, the ERPs are really moved up and down relative to x axis.
>>>>> However, they are still not in the right place...
>>>>> 
>>>>> My ERPs are segmented from -1000 to 2000 ms, artifactual ICs are marked
>>>>> as bad and rejected in the stage of precomputing channel measures.
>>>>> 
>>>>> I thought that it has something to do with artifactuals ICs which could
>>>>> be removed AFTER baseline is corrected. However it is not the case,
>>>>> since there is even more unpredictable offset at the posterior
>>>>> electrodes, where artifacts (mostly ocular) are relatively weak.
>>>>> 
>>>>> The same problem is visible for component cluster ERPs and it persists
>>>>> with the EEGLAB 10 as well as 11.
>>>>> 
>>>>> Any help...?
>>>>> 
>>>>> 
>>>>> --
>>>>> Mirek Wyczesany
>>>>> Jagiellonian University
>>>>> Psychophysiology Lab
>>>>> Kraków, PL
>>>>> 
>>>>> 
>>>>> PS. In the code of std_precomp I've found a remark at the 'rmbase'
>>>>> parameter: "%deprecated, for backward compatibility purposes, not
>>>>> documented"
>>>>> 
>>>>> _______________________________________________
>>>>> 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
>>>> 
>>>> _______________________________________________
>>>> 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
>>> 
>>> 
>> 
>> _______________________________________________
>> 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
> 
> 
> 
> -- 
> Ş. Barış Demiral, PhD.
> Department of Psychiatry
> Washington University
> School of Medicine
> 660 S. Euclid Avenue
> Box 8134
> Saint Louis, MO 63110
> Phone: +1 (314) 747 1603
> 
> _______________________________________________
> 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