Do we have any autoregressive (AR) modeling experts here? My time-series is in binary because my work involves working with brain cells. Basically, a 'zero' means the brain cell is inactive, while a 'one' means the brain cell fired a pulse.
So, for each cell, my recordings look like 00100000000000000000010000000100000000010000000000 00000001001000000000001000000000001000...etc. and I have hundreds of cells recorded at once.
I'd like to apply AR to these time series so that I can compare them using Granger Causality.
Can someone help me out? I've posted this question in
http://www.mymathforum.com/ as well.