Fig. 1.

Outline of the DITRAS framework. Ditras combines two probabilistic models: a diary generator (e.g., MD${}^{(t)}$) and trajectory generator (e.g., d-EPR). The diary generator uses a typical diary $W^{(t)}$ to produce a mobility diary D. The mobility diary D is the input of the trajectory generator together with a weighted spatial tessellation of the territory L. From D and L the trajectory generator produces a sampled mobility trajectory S