clustering parameters

Eric Maris maris at NICI.RU.NL
Thu May 10 12:37:32 CEST 2007


Christine,


> I'm wondering which parameters I should use for determining the way the
> clusters are formed in the freqstatistics;
>
> in the tutorials, I only found 'cfg.minnbchan' to specify the minimum
> number of neighborhood channels for a cluster, but in an older example
> script I also found 'cfg.neighbourdist' to determine the min. distance
> between neighborhood channels...

The two fields have very different functionalities.
1. The field cfg.neighbourdist is used to determine which channel pairs are
neighbouring channels. See neighbourselection.m.
2. The field cfg.minnbchan is used to determine whether a
(channel,frequency,time)-triplet (or (channel,time)-pair, or
channel-singleton) that exceeds the threshold should be included in the
clustering algorithm or not. Tip: start with the default value
(cfg.minnbchan=0) and watch what happens if you change its value. Don't
worry too much, cfg.minnbchan is a tuning parameter for which there is no
theoretical rationale (contrary to the permutation test itself, which has a
beautiful theoretical rationale). That's why I get so many questions about
it... I have seriously considered removing it from the Fieldtrip code, but
it still in, so you can play with it...


> and if 'cfg.neighbourdist' is used, is the value specified in mm or cm?

You should specify it in the scale as the coordinates of your channel
configuration. Remember that you must run neighbourselection.m to construct
your neighbourhood structure.


Greetings,


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 :www.ru.nl/master/cns/>
> Thanks,
> Christine
>
> ----------------------------------
> 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