[Eeglablist] FW: About an export of mean band power etc

[Tarik S Bel-Bahar]

Small modification:
Go the line that starts with v2,
And change the v near the end of this line of code to v2. Cheers
On Feb 26, 2012 9:37 PM, "Tarik S Bel-Bahar" <tarikbelbahar at gmail.com>
wrote:

> Hi Guy, due to your time urgency, and current level of awareness of eeglab
> and matlab,
> perhaps the following code derived for your situation may help. Do let us
> know of your final solution please.
> This solution uses spectopo to compute info for a specific component at a
> specific frequency for a series of files, components, and frequencies.
> See annotations for what to change or modify. You should end up with
> unique and easily addressable variables
> in your workspace, one for each "file, component, and frequency
> combination".
> Averaging and computing scores within and across participants I leave to
> you. Check the brief notes in the code.
> Best wishes, Tarik
> [just paste the below to your matlab editor, modify, and give it a go] [if
> there is a hiccup when attempting to run, let me know].
>
>
> *************************BEGIN SAMPLE CODE
>
>
> %this should run on epoched and unepoched files
> %I did not modify settings further, nor examine function specifications
>
> clear
> clc
> eeglab
> cd ( '/Homedirectory/DirectoryNameYouwantTouseThatHasTheFiles/');
> %directory should already exist and have all EEGLAB files
> filepath = pwd; %makes filepath the same as the directory with the files
> %here you set up your lists of components, files, and frequencies to loop
> through
> myListofComponents = [1,2,3,4,5];
> myListofFiles = ['file1.set','file2.set','file3.set'];
> myListOfFrequencies = [5,10,20]
> percentageofFiletoAnalyze = 2;
> for LoopThroughFiles = 1: length(myListofFiles)
>     %begin file loop
> currentfilename = myListofFiles(LoopThroughFiles);
> EEG = pop_loadset('filename','currenfilename','filepath',currentfilepath);
> [ALLEEG, EEG, CURRENTSET] = eeg_store( ALLEEG, EEG, 1 );
>  for LoopThroughComponents = 1: length(myListofComponents)
>      %begin component loop
> currentcompA = myListofComponents(LoopThroughComponents);
> %set number of component of interest in current single subject single
> session file
> for LoopThroughFrequencies = 1: length(myListofFrequencies)
>     %begin frequency loop
> currentfrequency = myListOfFrequencies(LoopThroughComponents);
> % set current frequency of interest
> figure;  [spectra freqs speccomp] = pop_spectopo(EEG, 0, [0      EEG.xmax
> *1000], 'EEG' ...
>  , 'freq', [currentfrequency], 'plotchan', 0, 'percent',
> percentageofFiletoAnalyze, 'icacomps', [currentcompA]...
> , 'nicamaps', 1, 'icamaps', [1], 'icamode',  'normal',  'freqrange',[8
> 12],'electrodes','off'); close(gcf);
> %ignore figure I could not manage to turn it off
> %freqs returns a column of numbers (freqs1),  %freqs(2), etc.... which
> specify the Hz for %each column in speccomp
> newspecs=speccomp;newfreqs=freqs; %dump output to dummy holding variables
> currentfilename1=currentfilename(1:5); %take first n characters of each
> filename
> currentcompA1= num2str(currentcompA);
>  v = genvarname([ 'speccomp' currentfilename1 currentcompA1 'hz']);
> eval([v ' = newspecs']);
> v2 = genvarname([ 'speccomp' currentfilename1 currentcompA1 'hz']);
> eval([v ' = newfreqs']);
> end %end frequency loop
> end %end component loop
> end %end file loop
>
> %NOTES TO GUY
> %the above should provide a file in the workspace named
> [FileName_Component Number_ Frequency]
> %the top row of each "speccompfileXhz" file contains the output for that
> %component at that particular frequency for that particular file
> %you should be able to find the columns in this one row that
> %represent alpha (try plotting that one row)
> %then take an average of the values in those columns of the one row
> %and save them out as a new variable or add them to a statistics speadsheet
> %am not sure about getting frequency bands, but the dirty way
> %would be to simply have run the code for all sub-frequencies of interest
> %and average up from there. Also, apologies, but I can't speak to or
> %provide caveats about function settings,
> % nor about power vs. amplitude, or using newtimef, or using non-eeglab
> % computation options. Good luck and let us know
> %what solution ends up working best for you
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20120226/380ae17a/attachment.html>