Hello Sean:
There should be a 1-1 correspondence between time points in the stim files
and time points in the time series data. (There is one exception: if the
"-stim_nptr" option is used, then there should be multiple stim fnc. time
points for each data point.)
Therefore, line 0 of the stim file corresponds to time point 0 of the time
series data, line 1 of the stim file corresponds to time point 1 of the time
series data, etc. etc. ("-stim_nptr" option excluded).
Important: Although using "-nfirst 8" means that the first 8 data points are
ignored, this does not mean that the first 8 input stim fnc. points are
ignored. Specifically, if maxlag > 0, then some (or all) of the first 8
input stim fnc. points will be used in modeling the data (as a time-delayed
response to previous inputs).
Doug Ward
