Figure 14.

Illustration of likelihoods, hazards and survivals for moderately restrictive kernels (moderate human movement, most movement under 50 km). Plots showing how the pairwise likelihoods, survivals and hazards vary with time and distance under different model structures. The first row of plots shows the pairwise likelihoods, the second row shows the pairwise survival and the third row shows the pairwise hazard values for different combinations of distance (in kilometres) and time between symptom onset (days). The first column shows the results for a time-only version of the algorithm. The second column shows results for an exponential kernel and the third column shows results for a Gaussian kernel. In this example values for beta, the shaping parameter for the distance kernels have been chosen to represent a context where there is more some movement of parasites, but where little movement is expected beyond 50–75 km. The likelihood for the Gaussian Kernel is more concentrated, which could represent shorter range movement e.g. commutes, whereas the Exponential has a longer tail so could represent a mixture of short and longer range parasite movement.