[FieldTrip] custer based non parametric permutation test for coherence difference

[ishita basu]

I am very interested in using your firldtrip code for doing nonparametric
cluster based permutation testing for coherence differences.

I tried reading through the fieldtrip tutorials on doing similar analysis
on time-frequency spectral data, but could not find anything on coherence
except for the coherence calculations for the cortico-muscular data
(doesn’t involve any statistical testing). I also saw that the main
function being used , i.e the ft_freqstatistics function requires input
from ft_freqanalysis  and not ft_connectivityanalysis (the one for
calculating coherence).




Another doubt I had was, it seems like I need some neighborhood information
for doing the cluster analysis. If I wanted to cluster only in the time
frequency grid and not spatially (I am using depth electrode data ) or I
wanted the algorithm to find spatial clusters for me in the absence of
prior knowledge, is that feasible ?


Thanks


Ishita
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20150806/258caca2/attachment.html>