External Email - Use Caution
Hi all,
I have been used *spatio_temporal_cluster_1samp_test* on my
time-frequency data (power (EpochsTFR) with a shape:[epochs X chan X freqs
X time] ). I have read the cited paper (MarisOostenveld07) for the
permutation tests. However, it is still not clear to me what these
functions are exactly doing. I have confused about the cluster analysis
after seeing my plots. I have attached my code here. I used a public
dataset (eegbci) as an example. My code could seem long, but most parts are
related to plotting. There are two main functions and I made more major
parts bold and with bigger font sizes.
More specifically, I am wondering if transforming participants' power
data to *EpochsTFR*" and "*AverageTFR*" is the correct way of analyzing
multiple participants? (e.g, after getting the power/ itc data for each
participant saving it in a multidimensional array and then transform
it to EpochsTFR
or AverageTFR).
Additionally, what do cluster_p_values tell us? Don't they show the
significance of one point in a power/itc data? So if I look at one point in
power/itc plot, its corresponding point in cluster_p_values plot should
tell me whether or not that point is significant?
I would be very thankful if someone can help me with these questions.
Thank you,
-Mary
"""