frequency-domain granger prediction and model order

[John Stout]

Hey Mike,

I was wondering what you thought about adjusting the model-order to control for cycle-counts when utilizing the frequency-domain granger prediction method from your book.

For example, using LFP data that was down-samped to 250hz, I estimated that a model order of 33 should provide a little more than 1 cycle of 8Hz theta, while a model order of 8 would provide a little more than 1 cycle of 40Hz gamma. I guess I'm wondering if it's better to control for cycle count by changing the model-order (per frequency) or to hold the model-order constant across all frequencies.

Thanks!

John

[Mike X Cohen]

Hi John. It's a good question, and I'm not really sure what the best thing to do is. But I generally try to keep the parameters constant. If the order differs between 8 and 40 Hz, then it will be difficult to compare the results across those frequencies, which is otherwise something that you might want to do.

Mike

--
You received this message because you are subscribed to the Google Groups "AnalyzingNeuralTimeSeriesData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to analyzingneuraltimes...@googlegroups.com.
Visit this group at https://groups.google.com/group/analyzingneuraltimeseriesdata.
To view this discussion on the web visit https://groups.google.com/d/msgid/analyzingneuraltimeseriesdata/e1b4737e-1c5d-4c7d-9d7f-051fba6dbda6%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Mike X Cohen, PhD
Fresh look: mikexcohen.com

[John Stout]

Thank you Mike!

To view this discussion on the web visit https://groups.google.com/d/msgid/analyzingneuraltimeseriesdata/CAABvs%2Bs3XnCePCUR3dtPkP6S7OB7VK3QSVih2G1WMggJa-GbXw%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

John Stout

M.S. Neuroscience

Griffin Lab - Neurobiology of Learning and Memory

Department of Psychological and Brain Sciences

University of Delaware