[FieldTrip] Trentool doubts

[Ana Rita Oliveira]

Hello,

I have some questions regarding TRENTOOL.
I don't know if it's ok to ask questions about this toolbox here, but since they don't have a mailing list and it is incorporated in fieldtrip, I wanted to give it a try anyway.

I am trying to compute delays between electrode pairs using the InteractionDelayReconstruction_calculate.
My data was sampled at 1024Hz, I have around 300 trials of a visual stimulus and I am saving the epochs between [-100, 500] ms. I am applying a band-pass filter between 0.1-40 Hz.

>From what I've read I still cannot understand if I have enough data points to compute a robust transfer entropy, and if I don't have, what can I do? Increase the epoch time from the stimulus onset forward or that would not contribute with relevant information because would be too far away from the stimulus? Also, should I consider the period before the stimulus onset ([-100, 500] ms) or should I just define my time of interest starting at the stimulus onset ([0, 500] ms)?

>From what I understood, the actthrvalue parameter should depend on my highpass filter ("if you highpass filter your data at 10 Hz and have a sampling rate of 1200Hz, you shouldn't find any autocorrelation above 120 samples" --> seen in other thread ). In my case, since I highpass at 0.1Hz, I shouldn't find any autocorrelation above 10240 samples, but since I don't even have that amount of samples and the act is lower than that anyway, this actthrvalue parameter shouldn't matter much, because the max act will be defined by the act values from my data. Is this correct?

Also, in my case, it is common that the tau and dimension that I get, are always the lower values that I give in the options. For example, with
cfgTEP.ragdim         = 2:6;
cfgTEP.ragtaurange    = [0.1 0.4];
I would get dim=2 and tau=0.1, but if I give a ragtaurange  starting at 0.05, I would get 0.05. Is this a normal behaviour?

Again, I am sorry if this is a bit out of the topic in fieldtrip, but if someone could give me some lights on this, I would appreciate it!

Thank you in advance,
Rita

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20200925/f8b0f15f/attachment.htm>