I think this construction is very clever, and I think a lot of other problems could take advantage of a similar setup. For example, suppose one wanted to simulate the transmission of a pathogen within some host population. The distribution of hosts could be used to build a contact hierarchy, and the probability that any infected host exposes any susceptible host to the pathogen could be found efficiently by walking though the tree. This would be much better than the naive approach of iterating over all pairs of hosts to check for transmission events.
I think this construction is very clever, and I think a lot of other problems could take advantage of a similar setup. For example, suppose one wanted to simulate the transmission of a pathogen within some host population. The distribution of hosts could be used to build a contact hierarchy, and the probability that any infected host exposes any susceptible host to the pathogen could be found efficiently by walking though the tree. This would be much better than the naive approach of iterating over all pairs of hosts to check for transmission events.
This comment was marked helpful 0 times.