Hou, Goodman, and I had our weekly meeting today to discuss exoplanet fitting and inference. Goodman suggested a possible statistical model for stellar oscillations that would permit us to treat them as a kind of structured Gaussian noise. Yet another project becomes a kind of Gaussian process! The idea is to drive a damped oscillator with a broad-band source. I assigned that to myself as homework for my flight to Amsterdam.
Hi David,
ReplyDeletehttp://arxiv.org/abs/0902.3907
:-)
- Brendon
Brendon: Great paper, and extremely relevant. But there is something I deeply don't understand. The function f(t) you propose is not only not continuous, none of its integrals or derivatives is continuous. How can *that* be the driving force? Don't you need the function to be something with a well-defined fourier transform?
ReplyDeleteHmm, interesting questions. I hadn't thought about the finer mathematical details, just tried it numerically with a discrete time version, saw that it worked, and then made up a formula for the resulting covariance function of the signal.
ReplyDeleteIt's true that f(t) is not continuous, but its integrals definitely are. The antiderivative is just a Brownian motion right? Does this help?