For fun recently I started poking around in Linear Algebra Methods in Combinatorics, a manuscript by one of my former professors who was himself a student of Erdos, that's where I have my exposure to Bollobas's work.

The wifing unit is rather avidly watching the new Doctor Who on her own. I accidentally spoiled that Donna Noble is coming back, but it doesn't really ruin anything other than the show itself (least favorite companion ever). Unfortunately, her enthusiasm for the new series makes her less enthusiastic for the old.

Quick algorithms question: when you have a pile of things, like, books on a bookshelf, that you want to sort manually for some reason, what method do you typically gravitate toward? I often use variants of insertion sort (optimized by the fact that a human eye can see a bunch of things at once and be a little bit intelligent) if I have a small list, since it's not terribly intensive from a "shuffling of books" point of view even if it's "computationally" expensive in the average case, since "shuffling books" is the real "cost" and "computations" are cheap. Heap sort is impractical, for instance, because I can't build a heap. Bubble sort - slightly modified to fix "turtles" - also comes up, but there's a lot of shoving things around. None of these get done strictly.

I'm now reading The Master of Hestviken, a set of four books by Sigrid Undset. I'm on the second book. I will go back to George R R Martin when I get the book. Update [2011-10-12 22:36:10 by gzt]:And I got 100% on my data mining test. Apparently some people did not? Perhaps they were undergrads, as this is a cross-listed grad/undergrad course, and undergrads don't realize you're supposed to learn all the material and prepare for tests.