<div dir="ltr">Dear Snow,<div><br></div><div>Don't apply ICA on each condition separately. I mean, scientifically that is not wrong, but EEGLAB is not designed to work in that way. Instead, you should run ICA on a data set with all conditions, and conditions should be separated after ICA. In this way different conditions can have same independent components, and difference of activities can be tested <b>within</b> independent components.</div>
<div><br></div><div>Makoto</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, May 29, 2014 at 5:06 AM, 张雪 <span dir="ltr"><<a href="mailto:zhangxue@psych.ac.cn" target="_blank">zhangxue@psych.ac.cn</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Makoto<br> I was using EEGLAB's studyset analysis. I did ICA for every condition of every subject. All the ICA numbers are 66. But When I do the preclust in study, the program still remind me different ICA decomposition and I couldn't do the following analyses. <br>
So, I want to know, what kind of ICA decompositions are identical?<br> Best wishes<br> Zhang Xue<br><br><span></span><br><br><br></blockquote></div><br><br clear="all"><div><br></div>-- <br><div dir="ltr">
Makoto Miyakoshi<br>Swartz Center for Computational Neuroscience<br>Institute for Neural Computation, University of California San Diego<br></div>
</div>