clusterrandanalysis input format

Sameer Walawalkar sameer at ANDREW.CMU.EDU
Mon Sep 24 20:39:14 CEST 2007


Dear Eric,

Following your suggestion I used labelcmb field from the output of
freqanalysis to choose combinations with the refchannel in both first and
second position.

cfg.channelcmb was a 305x2 cell array.

Yet, after clusterandanalysis,

posclusterslabelmat: [205x13 double]
negclusterslabelmat: [205x13 double]

(and not with dimensions [306xnumber of freqs]).
It reflects the fact (I think) that the first 205 combinations in
cfg.channelcmb have the refchannel in the first position and the rest of
the 100 combinations  have it in the second position.

This size of posclusterslabelmat and negclusterslabelmat (205x13) is what
I get from using
cfg.channelcmb = {refchannel  MEG}

So I am not sure your suggestion is the solution.  (Also, from your email,
I dont think this requires a newer corrected version of
clusterandanalysis.m).

Best,
sameer




On Sun, 23 Sep 2007, Eric Maris wrote:

> Dear Rodrigo Salazar and Sameer Walawalkar,
>
>
>
>
>
> You are both involved in similar analyses, and therefore I address you both.
>
>
>
> 1.	There was a bug in the latest version of freqanalysis_mtmfft, which
> caused the error message in Rodrigo's analysis. Sameer used an older version
> of freqanalysis_mtmfft, in which this bug was not yet present. I have
> corrected the bug and within a couple of days you can download the new
> m-file from the FT homepage.
> 2.	There were a few bugs in the private function
> clusterrandstatistics.m related to the postprocessing of the results of a
> onesided statistical test. I have corrected these bugs. I now obtain
> identical p-values for the elements of posclusters in (1) a twosided
> analysis with cfg.alphathresh=0.05 and (2) a onesided analysis
> (cfg.onetwo=onesided_1<2) with cfg.alphathresh=0.025, as it should be.
> Similarly, I now obtain identical p-values for the elements of negclusters
> in (1) a twosided analysis with cfg.alphathresh=0.05 and (2) a onesided
> analysis (cfg.onetwo=onesided_2<1) with cfg.alphathresh=0.025, again as it
> should be.
> 3.	Sameer has selected channel combinations by specifying
> cfg.channelcmb={refchan 'MEG'}. This only gives you the channel pairs in
> which your reference channel is in the first position. As a result, you will
> not select the channel pairs in which the reference channel is in the second
> position. I must admit that I was not aware of this problem. I advise you to
> use the labelcmb-field in the output of freqanalysis. From this field, you
> can easily select those channel pairs in which the reference channel is in
> either the first or the second position, and pass this selection via the
> cfg.channelcmb-field.
>
>
>
>
>
> Good luck,
>
>
>
> Eric Maris
>
>
>
> dr. Eric Maris
>
> NICI/Biological Psychology and
>
> F.C. Donders Center for Cognitive NeuroImaging
>
> University of Nijmegen
>
> P.O. Box 9104
>
> 6500 HE Nijmegen
>
> The Netherlands
>
> T:+31 24 3612651 (NICI)
>
> T:+31 24 3610754 (FCDC)
>
> F:+31 24 3616066 (NICI)
>
> E: maris at nici.ru.nl
>
> MSc Cognitive Neuroscience : <http://www.ru.nl/master/cns/>
> www.ru.nl/master/cns/
>
>
>
>
>
>
>
>
>
>  _____
>
> From: FieldTrip discussion list [mailto:FIELDTRIP at NIC.SURFNET.NL] On Behalf
> Of Rodrigo F. Salazar
> Sent: Friday, September 21, 2007 7:16 PM
> To: FIELDTRIP at NIC.SURFNET.NL
> Subject: Re: [FIELDTRIP] clusterrandanalysis input format
>
>
>
> Hi Eric,
>
>
>
> Thanks for your quick reply. I scanned the tutorials and I could not find
> one that is using clusterrandanalysis.m. I only found a matlab script named:
>
>
>
> "Apply clusterrandanalysis on TFRs of power that were computed with BESA"
>
>
>
> but as it is only a script I cannot access the data.
>
>
>
> So far I was capable to run freqanalysis.m by adapting my data to the format
> described in the documentation. However, when I run clusterranalaysis.m with
> the output of freqanalysis.m being 'freq'
>
>
>
> [clusrand] = clusterrandanalysis(cfg2,freq,freq); % I am using twice 'freq'
> as a test run with cfg2.statistic: 'indepsamplesZcoh'
>
>
>
> , I keep on getting errors such as :
>
>
>
> ??? Reference to non-existent field 'dof'.
>
>
>
> Error in ==> fieldtrip-lite-20070920/private/clusterrandstatistics at 680
>
> dofc1=reshape(sum(data.dof(replselc1,:,:),1),1,nfreq,ntime);
>
>
>
> And if I had the option:
>
>
>
> cfg.calcdof = 'yes'
>
>
>
> to get the degree of freedom, I get this other error:
>
>
>
> ??? Undefined function or variable "cumtapcnt".
>
>
>
> Error in ==> freqanalysis_mtmfft at 389
>
> if calcdof && ~exist(cumtapcnt)
>
>
>
> Could you please tell if I am using a wrong configuration variable or if
> there is a problem in the code?
>
>
>
> Thanks,
>
>
>
> Rodrigo.
>
>
>
>
>
> On Sep 20, 2007, at 9:56 AM, Rodrigo F. Salazar wrote:
>
>
>
>
>
> Hi,
>
>
>
> I am new to fieldtrip and I am working with intracortical data. I have never
> used any function of fieldtrip and I would like to use clusterrandanalysis.m
> using my own data format. I cannot find a clear description of the input
> needed for that function. Can someone quickly describe the input format or
> even send me an example of input data?
>
>
>
> Thanks,
>
>
>
> Rodrigo.
>
>
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Rodrigo F. Salazar, Ph.D
>
> Post-doctoral Fellow
>
> Lewis Hall 1,
>
> Center for Computational Biology
>
> Montana State University
>
> Bozeman, 59717 MT, USA
>
> Phone: ++1 (406) 994 71 90,
>
> Fax: ++1 (406) 994 74 38
>
> email: rsalazar at nervana.montana.edu
>
> http://cns.montana.edu/~rsalazar/
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> ----------------------------------
>
> The aim of this list is to facilitate the discussion between users of the
> FieldTrip toolbox, to share experiences and to discuss new ideas for MEG and
> EEG analysis.
>
> http://listserv.surfnet.nl/archives/fieldtrip.html
>
> http://www.ru.nl/fcdonders/fieldtrip/
>
>
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Rodrigo F. Salazar, Ph.D
>
> Post-doctoral Fellow
>
> Lewis Hall 1,
>
> Center for Computational Biology
>
> Montana State University
>
> Bozeman, 59717 MT, USA
>
> Phone: ++1 (406) 994 71 90,
>
> Fax: ++1 (406) 994 74 38
>
> email: rsalazar at nervana.montana.edu
>
> http://cns.montana.edu/~rsalazar/
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> ----------------------------------
>
> The aim of this list is to facilitate the discussion between users of the
> FieldTrip toolbox, to share experiences and to discuss new ideas for MEG and
> EEG analysis.
>
> http://listserv.surfnet.nl/archives/fieldtrip.html
>
> http://www.ru.nl/fcdonders/fieldtrip/
>
>
> ----------------------------------
> The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.
>

----------------------------------
The aim of this list is to facilitate the discussion between users of the FieldTrip  toolbox, to share experiences and to discuss new ideas for MEG and EEG analysis. See also http://listserv.surfnet.nl/archives/fieldtrip.html and http://www.ru.nl/fcdonders/fieldtrip.



More information about the fieldtrip mailing list