[R-sig-Geo] Origin and Destination-specific doubly constrained model

[Ticiana Grecco Zanon]

Hello everybody,


I would like to share a difficulty I have that it can be very easy for somebody in this forum.


I have calculated a distance parameter for the whole system in a doubly constrained model :

Model01 <- glm(Flow~Origin+Destination+log(Dij), family=poisson(link="log"),data=Base)


Now I would  like to do it, now origin-specific in the sense of Fotheringham (1981, 1983 and so one), that is, for each origin, a distance decay paramater with the purpose to see the spatial structure of the flows.


I suppose it is as bellow meanwhhile I am not sure.


Model02 <- glm(Flow~Origin+Destination+factor(Origin)xlog(Dij), family=poisson(link="log"),data=Base)


Thanks in advance,

Ticiana


	[[alternative HTML version deleted]]