June 8th, 2011

Control Tower

Three things make a post

The pool is open for the season, but this evening it was too cool to go swimming. We finished Disc 1 of Fringe Season 2 instead. I love this show!

I've taken to reading discussion reviews on my lunch hour of shows I've been watching. Most of them are Onion A.V. club blogs, but I'm also greatly enjoying Mark Watches Avatar.

At work I'm working on a fixed version of a function that I have now thoroughly proven has a bug in it. Sadly, my code is interpreted, and therefore considerably slower than compiled version, but I think if somebody can convert it to compiled code, it will be not only better than the existing buggy code, but quite a bit faster. Today I believe I managed to remove the last of the stupid from my code, so now I just need to code up the test cases that demonstrate (I hope) its superiority.

Time seems to be unfolding quickly of late.
Dream of Bingo

Math-ish blather

So, speaking of algorithms:

Does anyone know if you can prove that hill-climbing will work on a curvilinear grid?

Which is to say, I have a set of points that have the same combinatorial structure as a regular grid, though it may be all stretched out in lat/lon space. I want to find the point nearest to some target by starting at on point in the grid and repeatedly moving to the neighboring point that's closest to the target. This will work as long as there aren't any local minima in the distance field.

The question is, are there simple conditions under which you can be certain there aren't any local minima? Or certain that there are? Because it would be useful to check whether getting the right answer is guaranteed or not. (Of course, it has to be less work to check that condition than it would be to evaluate every candidate point, or there's no point in using a clever algorithm rather than the brute-force approach...)