cluster statistics on TFR data including channels

✉️ Mailing List Archive (read-only)
1

External Email - Use Caution

hi,

i hope i did not miss anything in the documentation or the mailing list
archives.

i have a couple of AverageTFR objects (one for each subject) with the
difference between two conditions. i would like to run cluster based
non-parametric statistics on those, searching for clusters in all three
relevant dimensions (i.e. channels x freq x time).

if i am not mistaken, this is not implemented in MNE right now. i could
cluster over freq x time using mne.stats.permutation_cluster_1samp_test
and doing it one channel at a time. or i could use
mne.stats.spatio_temporal_cluster_1samp_test which would let me include
all channels, but force me to either pre-determine a time-window or a
frequency band.

is that right? or did i miss something?

thanks a lot in advance,

thomas

2

External Email - Use Caution

A similar question was asked recently on the MNE issue list. Basically it's
doable but not easy, we have plans to make it easier but haven't
implemented them yet, and you can try the "eelbrain" package (which works
with MNE) in the meantime if you want. See:

https://github.com/mne-tools/mne-python/issues/5144

Eric