User:D. S. McNeil
Let's see. I'm a computational planetary astrophysicist who still has an affection for recreational number theory. Undergraduate at University of Toronto, in astronomy, math, and physics; master's and doctorate at Queen's University, Kingston; later worked as a postdoc at Queen Mary, University of London in the School of Mathematical Sciences and then as a fellow in the Earth Sciences department at the University of Hong Kong. I mostly studied planetary origins, and to that end had to spend a lot of time thinking about how to squeeze every last cycle from a cluster. The planetary formation problem is much harder to model (computationally, anyway) than the galaxy formation problem because we have to integrate for billions of dynamical times, unlike those galaxy types who have it ludicrously easy. Kept me busy, anyway!
These days I've made the jump to full-time numerical modeling in industry, and am currently working at a Toronto risk analysis firm with some fellow physicists. Right now I'm working on pandemic response scenarios, which is kind of fun. (In my head I think of it as preparing for the inevitable zombie invasion..) I've had to learn to say "agent-based" instead of "Lagrangian", but apart from that I'm doing much the same kind of numerical programming I've always done, simply in a different domain, and now I get paid for it. :^)
Creighton Kenneth Dement posted a question on the seqfan mailing list about a variant of the classic cannibals-and-missionaries problem; a bit of code later I had the answer, as well as mild chagrin that a few minutes' thought should have given me the solution.
In any case, here's the (main 52-node connected component of the) graph where each node is a permitted state of the problem and each edge corresponds to a permissible move (sending the policeman and a thief across the river on the boat if that's allowed). The green and red nodes are the initial and final states.
D. S. McNeil 14:45, 10 January 2010 (UTC)