Stochastic geometry and the London underground

Way back when I was analysing London house price data for the Summer Data Challenge, I made a histogram of the distances from a random point in London to the nearest tube station. I noted that it peaked around half a kilometre, but ignored the shape of the distribution itself. This is an unfortunate faux pas for the accomplished procrastinator, so let’s right that wrong with the help of some stochastic geometry.

Continue reading


Annealing the Underground

I was reading the other day about the Tube Challenge, where people attempt to visit all 270 stations on the London Underground in the quickest possible time. Apparently successful routes are a closely-guarded secret, which got me thinking about how difficult an optimisation problem it must be to find the best route.

Continue reading