2007 October 2
Most of today was SF homework. Happy with the result, though, as always, took longer than it felt it should have.
Had a tiny little bit of time to look at my broadband spectra data. Annoyingly, I have to spend some time thinking about network logistics: there are about 30 gigs of data at Hat Creek, and limited bandwidth. The usual practice is to average down the spectra before downloading, but there are clear single-channel outliers that will distort the averaged data. That can be avoided if I flag the outliers, but I don’t know of a way to do that automatedly. I could write a tool to do it, but that’d require my sketchy Python library, which can only run here. And the data are at Hat Creek … etc. I’ll just average them and worry about things later. It makes me sad to mess up 24 channels’ worth of data when only 1 is the problem, but that’s life. Hell, I’ve got 1024 channels; I can afford to waste a few.
Then a lot of lab. Still very enjoyable to work with the students. Some of us chatted and agreed that it’d be useful to give them an overview not so much of IDL but of general sound programming practices. I’ll try to put that together, though it’ll probably take a while to do so. But I’ll be pretty happy if it comes together in a way that I like. Well, that’s kind of tautological. But yeah.