Dear Makoto,<div><br></div><div>I checked my raw data and found that there is significant large power in about 0.5 Hz,just as shown in the link below.</div><div><a href="https://www.dropbox.com/s/24fhfxcnqyvqyuc/raw_rythm.jpg">https://www.dropbox.com/s/24fhfxcnqyvqyuc/raw_rythm.jpg</a> </div>
<div>then I high-pass filtered at 1 Hz,the resulted power-spectra plot showed that although frequency below 1 Hz was well attenuated, the power remained significant in 1~1.5 Hz,which became the largest power in the spectra after filter done.,just as shown in the link below.</div>
<div><a href="https://www.dropbox.com/s/u66n4qu43ww4mfo/1Hzfiltered.jpg">https://www.dropbox.com/s/u66n4qu43ww4mfo/1Hzfiltered.jpg</a> </div><div>So I think such a filter-left 1 Hz activity is the cause of the 1 Hz component in my ICA decomposition.And such a 1 Hz component seems to do harm to my other components,just as shown in the links below:This is a comparatively "good" component,but as you can see from the picture,the baseline before 0 time point shows a slight 1 Hz activity,which makes the baseline looks uneven.</div>
<div><a href="https://www.dropbox.com/s/woaacj0426jjqel/good_component.jpg">https://www.dropbox.com/s/woaacj0426jjqel/good_component.jpg</a>
</div><div>So,I have 2 questions about it:</div><div>1. The slow rhythms,such as 1 Hz,seems to do harm to ICA decomposition.If so,what's the reason behind this?Is it because slow rhythms<br> increase non-stationarity?</div>
<div>2.What can I do to resolve such a problem I met?</div><div><br></div><div>Thank you very much!</div><div><br></div><div><div class="gmail_quote">2012/5/23 Makoto Miyakoshi <span dir="ltr"><<a href="mailto:mmiyakoshi@ucsd.edu" target="_blank">mmiyakoshi@ucsd.edu</a>></span><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Yuan-Fang,<br>
<br>
It's not the 1Hz high-pass filter introduced it additively, but the<br>
filter left the 1Hz activity as a result (I mean that's the only<br>
possibility I can think of). Imagine you have a data with a lot of<br>
sub- and near-1Hz activity as a broad spectrum (and this wave does not<br>
beat at 1Hz). What happens when you filter out sub-1Hz activity? 1Hz<br>
activity appears as a result, which beats at 1Hz.<br>
<br>
I'm still wondering what the artifact is. Please tell us what you find.<br>
<div class="HOEnZb"><div class="h5"><br>
Makoto<br>
<br>
2012/5/22 Yuan-fang Chou <<a href="mailto:distancejay@gmail.com">distancejay@gmail.com</a>>:<br>
> Dear Makoto,<br>
><br>
> Thanks for your comment.I will check my original dataset soon.But could you<br>
> please explain to me how 1 Hz high pass filter could make such a strange<br>
> component? I'm really naive on this.<br>
><br>
> 2012/5/23 Makoto Miyakoshi <<a href="mailto:mmiyakoshi@ucsd.edu">mmiyakoshi@ucsd.edu</a>><br>
>><br>
>> Dear Yuan-Fang,<br>
>><br>
>> That is probably because you used a 1Hz high-pass filter. Probably<br>
>> your original dataset has a very strong sub-1Hz activity. Please check<br>
>> that possibility.<br>
>><br>
>> Makoto<br>
>><br>
>> 2012/5/22 Yuan-fang Chou <<a href="mailto:distancejay@gmail.com">distancejay@gmail.com</a>>:<br>
>> > Dear list,<br>
>> > I filtered my 64 channels data at 1~30Hz,and epoched it before running<br>
>> > ICA.Then I got one of the last a few components in the link below.It<br>
>> > appears that this component oscillates steadily at about 1 Hz<br>
>> > rhythm,which<br>
>> > looks rather strange. Could anyone tell me how to explain this<br>
>> > component's<br>
>> > activity?<br>
>> ><br>
>> > <a href="https://www.dropbox.com/s/mgadt1q7scn9vsf/component.PNG" target="_blank">https://www.dropbox.com/s/mgadt1q7scn9vsf/component.PNG</a><br>
>> ><br>
>> > --<br>
>> > Yuan-Fang Chao<br>
>> > School of Psychology<br>
>> > SouthWest University<br>
>> > Beibei,Chongqing,China<br>
>> ><br>
>> ><br>
>> ><br>
>> ><br>
>> > _______________________________________________<br>
>> > Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" target="_blank">http://sccn.ucsd.edu/eeglab/eeglabmail.html</a><br>
>> > To unsubscribe, send an empty email to<br>
>> > <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu">eeglablist-unsubscribe@sccn.ucsd.edu</a><br>
>> > For digest mode, send an email with the subject "set digest mime" to<br>
>> > <a href="mailto:eeglablist-request@sccn.ucsd.edu">eeglablist-request@sccn.ucsd.edu</a><br>
>><br>
>><br>
>><br>
>> --<br>
>> Makoto Miyakoshi<br>
>> JSPS Postdoctral Fellow for Research Abroad<br>
>> Swartz Center for Computational Neuroscience<br>
>> Institute for Neural Computation, University of California San Diego<br>
><br>
><br>
><br>
><br>
> --<br>
> Yuan-Fang Chao<br>
> School of Psychology<br>
> SouthWest University<br>
> Beibei,Chongqing,China<br>
><br>
><br>
><br>
<br>
<br>
<br>
--<br>
Makoto Miyakoshi<br>
JSPS Postdoctral Fellow for Research Abroad<br>
Swartz Center for Computational Neuroscience<br>
Institute for Neural Computation, University of California San Diego<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>Yuan-Fang Chao <br>School of Psychology<br>SouthWest University<br>Beibei,Chongqing,China<br><br><br><br>
</div>