Busy Saturday
Feb. 20th, 2010 08:05 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
stoutfellowToday is the annual math competition, and as usual I'm one of the judges. I've checked the bus schedule, so that I won't wind up in Maryville again, and I'll be leaving soon.
Because of the once-an-hour #16 bus, I'll be getting there quite a bit early, and probably leaving quite a bit after. I think I'm going to spend the extra time in my office. Now that I've got the algorithms working properly, I'm going to try to get them organized better. The first step will be to divide the work into three notebooks: one for the actual code, one for the data describing the polyhedra, and one for the results of the computations. After that... well, I can see quite a few inefficiencies - places where the code can be modularized more thoroughly, and places where I can automate some of the tedious hand-entries. For example, one requirement of the algorithm, to prevent duplications, involves data concerning which edges "look alike" - i.e., there's a rotation of the polyhedron which takes one edge to another. I've been giving it that information by hand, but it'll be easier to just encode the rotations and let the computer do the rest of the work. I'd also like to insert some error-traps, to avoid the kinds of problems we were having yesterday.
:cracks mental knuckles:
Because of the once-an-hour #16 bus, I'll be getting there quite a bit early, and probably leaving quite a bit after. I think I'm going to spend the extra time in my office. Now that I've got the algorithms working properly, I'm going to try to get them organized better. The first step will be to divide the work into three notebooks: one for the actual code, one for the data describing the polyhedra, and one for the results of the computations. After that... well, I can see quite a few inefficiencies - places where the code can be modularized more thoroughly, and places where I can automate some of the tedious hand-entries. For example, one requirement of the algorithm, to prevent duplications, involves data concerning which edges "look alike" - i.e., there's a rotation of the polyhedron which takes one edge to another. I've been giving it that information by hand, but it'll be easier to just encode the rotations and let the computer do the rest of the work. I'd also like to insert some error-traps, to avoid the kinds of problems we were having yesterday.
:cracks mental knuckles:
