Python Kalman filtering and optimal estimation library. Implements Kalman filter, particle filter, Extended Kalman filter, Unscented Kalman filter, g-h (alpha-beta), least squares, H Infinity, smoothers, and more. Has companion book 'Kalman and Bayesian Filters in Python'.
If I have a state transition function where the noise term is scaled by the state:
x[t] = f(x[t-1]) + f(x[t-1])*N(0,sigma)
Will your implementation of ukf still work? It's not obvious from the documentation since all your example f() and h() lack noise.
That's not how noise works. The attribute R implements measurement noise, and Q implements process nose. Both are in the examples, and my book explains all of this in detail.
If I have a state transition function where the noise term is scaled by the state:
x[t] = f(x[t-1]) + f(x[t-1])*N(0,sigma)
Will your implementation of ukf still work? It's not obvious from the documentation since all your example f() and h() lack noise.