[FieldTrip] valid workaround for nonparametric stats on cortical surface data?

[Maris Skujevskis]

Hello Fieldtrip community,

ft_statistics_montecarlo fails when it is called through
ft_sourcestatistics with cortical surface (as opposed to a 3D volumetric
grid) source data.
To avoid this limitation, I can change my source level time series avg.mom
into a sensor data structure (my electrodes are simply 'virtual' then),
provide the source level neighborhood structure, and run
ft_timelockstatistics on my 'virtual sensor' data set. As far as I can
tell, this effectively gives source level statistics, just through the
wrong function, so to speak.

Since I am reading that ft_statistics_montecarlo might behave differently
depending on which function is calling it....

FT_STATISTICS_MONTECARLO should not be called directly, instead you should
call the function that is associated with the type of data on which you
want to perform the test.
% Use as
%   stat = ft_timelockstatistics(cfg, data1, data2, data3, ...)
%   stat = ft_freqstatistics    (cfg, data1, data2, data3, ...)
%   stat = ft_sourcestatistics  (cfg, data1, data2, data3, ...)


...I wanted to know whether there are some additional things to be aware of
when using ft_timelockstatistics on source level virtual electrode time
series.

Thanks!
Maris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20160817/45878084/attachment-0001.html>