[R-sig-Geo] Modeling SAR Origin-Destination flows with R?

[Guido Schulz]

Hi there,

I would like to model SAR Origin-Destination (migration) flows using the
model specification proposed by LeSage & Pace
(http://www4.fe.uc.pt/spatial/doc/lecture7.pdf, page8, equation [10]):

y = rho1*W_o*y + rho2*W_d*y + rho3*W_w*y + alpha*iota + X_d*beta_d +
X_o*beta_o + D*gamma + epsilon

Where "o" stands for origin and "d" for destination.

What is crucial about the model is, that we have the spatially lagged
responses of the origin AND the destination, as well as the spatially
lagged predictors of the origin AND the destination in the SAR equation.
-----------------------------------------
Do you have any ideas how to implement such a model in R?

Is there a way to specify a spdep::lagsarlm in that way or should I use
a different package/function? As far as I can see there spdep::lagsarlm
allows only for ONE "listw".

Best & thx,

Guido