Grayson, 1970

Looking at some old floor plan generating algorithms from the 70's. One that gets cited a lot is Grason's Thesis : "Methods for the computer-implemented solution of a class of "Floor Plan" design problems"

A summary of it is in his paper:

http://portal.acm.org/ft_gateway.cfm?id=805070&type=pdf

Carried out my own interpretation of the work using an adjacency matrix and a grid which is sized by the number of nodes connected to the direction (N, S, E, W) notes.

Strange to think that back then, it took over an hour to generate the solution, while now this can be calculated in real-time.