ABSTRACT
Sometimes no representative individual level dataset is available as an input for population synthesis. In this case, the methods described in the previous chapters must be adapted accordingly. The challenge is still to generate spatial microdata that fits all the constraint tables, but based on a purely synthetic ‘seed’ input cross-tabulated contingency table. Many combinations of individual level data could correspond to these distributions. Depending on the aim of the spatial microsimulation model, simply finding one reasonable fit can be suffcient.
