> When it tells me to save my path matrices at the end, it gives me 3 options (with a lag of 1, 2, and 3).
Did you specify a model with 3 lags? With FMRI data, a log of 1 is probably good enough.
> Lapack routine dgesv: system is exactly singular: U[1,1] = 0
With 3 regions, the upper bound for the number of paths is 3. In your case, the degrees of freedom would be 0, and it seems the system is singular, which means insoluble. When you see the following choice:
Numerical scheme (0: Iterative; 1: Direct)?
Choose 1 (Direct), and see if that helps.
Gang