[Eeglablist] Computing ERSPs in parallel?

[Arnaud Delorme]

Dear Tom,

std_precomp is the one you would want to use.
I would change the "for" loop line 483 to a "parfor" loop and comment line 497 (because errors cannot be generated in a parfor loop). 
Then call matlabpool(x) before running the function (where x is your number of processor).
Use this code to get the number of processors: import java.lang.*; r=Runtime.getRuntime; ncpu=r.availableProcessors (although in some cases, you have to divide this number by 2 because of hyper-threading).

This might do the trick,

Arno

On May 5, 2015, at 1:18 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu> wrote:

> Dear Arno and Ramon,
> 
> Here is a question for the EEGLAB developers.
> 
> Makoto
> 
> On Fri, May 1, 2015 at 9:53 AM, Tom Bullock <thomas.bullock at psych.ucsb.edu> wrote:
> Hello EEGLAB community,
> 
> I'm using the STUDY function to compute ERSPs for a dataset with 18 subjects and 3 conditions per subject.  The ERSPs take a number of hours to compute, so I'd like to set up a parfor loop to create the "xxx.datersp" files in parallel using our distributed processing network.
> 
> This post indicates that it should be possible to set up the parfor loop: http://sccn.ucsd.edu/pipermail/eeglablist/2014/008914.html . However I'm not sure which of the functions involved in the ERSP computation is the one I should attempt to parallelize (std_precomp.m, std_ersp.m or newtimef.m?), and which of the for loops I should be editing.
> 
> Does anyone have any insight into how to set this up?
> 
> Thanks!
> 
> Tom
> 
> p.s. this is the function I'm currently using to create the ERSPs:
> 
> [STUDY ALLEEG] = std_precomp(STUDY, ALLEEG, {},'interp','on','recompute','on','ersp','on','erspparams',...
> {'cycles' cycleParam  'nfreqs' 27 'timesout' [0:100:3000]  'freqs' [4 30] },'itc','off');
> 
> -- 
> Tom Bullock, PhD
> 
> Postdoctoral Researcher, Department of Psychological and Brain Sciences,
> University of California, Santa Barbara, CA 93106-9660
> 
> bullock at psych.ucsb.edu
> 
> https://labs.psych.ucsb.edu/giesbrecht/barry/Attention_Lab/Tom_Bullock.html
> 
> www.linkedin.com/in/tomwbullock
> 
> _______________________________________________
> 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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150505/fc79053f/attachment.html>