The ACM ICPC South Pacific Region was on yesterday, and was great fun (as usual). My team this year, the Mehffort Musketeers consisted of Alex Berry (who’ll be competing in the Google Code Jam regionals soon as well), Michael Ford and myself.

For the benefit of people who did the ICPC this year: I solved problem A, C and I; Michael solved B and D, and Alex solved E, F and H. Here’s some general commentary on the problems that I solved:

  • Problem A was very simple, and the shell of my solution was complete within three minutes of the contest starting. Unfortunately, the entire problem was not defined until halfway through the test data, which led me to finishing it a bit later.
  • Problem C, which seems to have been the problematic problem this year (as far as judging’s concerned) was relatively straightforward, though I had two rather annoying bugs that took me about an hour to week out… it happens, I suppose. Solved on the first submission, which I’m happy about.
  • Problem I was a longest path problem, that was relatively straightforward depending on what sort of algorithm you chose to solve it. I’ve heard reports of people using the Bellman-Ford Algorithm and failing — as far as I can tell, such an algorithm would work on problems except where there existed a cycle not involving the endpoints of the path taken in the problem. I used Floyd’s Algorithm and had it solved first time. Simple.

As alluded to earlier, we solved 8 problems, and we’re currently the only team to do so with the testing data used on the day (this means that we’re in a provisional first place), however, there are many teams who are likely to get problem C rejudged, and following that we’ll likely be third. More news to come.