ywx649999311
commented
11 months ago Hi! I am very sorry for the delayed response.
The CARMATerm does not have a variance parameter at the moment, but it could be something to add in the future. For your need, I think you could increase the variance by multiplying all MA parameters by a constant. You can also check the model variance using the get_rms_amp() method of the CARMATerm.
P.S. I might still be slow in responding in the next week or two, but please do let me know if it works or not.
Hi! I've been trying to use this package for a science project where I need CARMA (I'm currently trying to fit a CARMA(3,1) model), and it's been great!
One question I couldn't figure out: no matter the parameters I put in for the AR and MA process in
CARMATerm, the variance of the resulting GP I sample seems to be <0.1. The light curve I'm trying to model, however, has a much larger variance. The TinyGP implementation of CARMA has a variance parameter, but I don't think this one does. Any suggestions about what I'm doing wrong?