[Eeglablist] If you want to run 3 Hz 3 cycle wavelet but only have epochs shorter than 3 second...

[Baris Demiral]

I think you are talking about the sandwich method;  attaching the 
non-continuous neighbouring epochs this way may create boundary effects. 
At the first glance, it seems fine, since you are interested in the far 
away time points. But, if you are normalizing your spectrum later, this 
might change the values. Maybe, in that plug-in the following can be 
implemented:  rather than attaching the epochs, take the mirror/reverse 
amplitude values and append it; that is, for the first time point of the 
epoch at, say -300ms, write it down, then take the next time point at 
-298ms, and write it down, till say 500ms, then flip these values and 
append them to the preceding part of the epoch as -302ms being the 
original 300ms, then 304ms being the original 208ms. FLIPLR function is 
doing that. This may be more natural.
Baris

On 03/20/2013 10:23 AM, Arnaud Delorme wrote:
> We had debate about the use of this plugin. I personally think that if you have short epochs (and have lost the continuous data), then the only thing you could do to artificially make them larger is pad them with 0. However, 0-padding can add artifacts at boundaries.
>
> Makoto's solution was to add real data (from other data epochs). I think this is debatable and you need to be aware that your ERSP solution will be valid only within a specific time range at low frequencies.
>
> Makoto, maybe you could implement the 0-padding as well so people have more choices?
>
> Arno
>
> On 13 Mar 2013, at 05:58, Makoto Miyakoshi wrote:
>
>> Dear list,
>>
>> Recently we have seen more than one questions on the list asking how to have an 'extended' epochs-to perform a good wavelet analysis-without too much hustles. I was unable to provide a reasonable solution at that time, but now alpha version (i.e. use at your own risk) of it was created. This is a plugin for EEGLAB. Sorry I don't remember who asked the question. Hopefully they find this post. If you find a bug let me know.
>>
>> -- 
>> Makoto Miyakoshi
>> JSPS Postdoctral Fellow for Research Abroad
>> Swartz Center for Computational Neuroscience
>> Institute for Neural Computation, University of California San Diego
>> <sandwichPlugin.zip>_______________________________________________
>> 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