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'.
MIT License
3.23k
stars
614
forks
source link
can rts_smoother return cov(\theta_t, \theta_{t-1}) rather than just cov(\theta_t)? #207
It seems like that when running rts_smoother, it can also calculate $Cov(\thetat, \theta{t-1})$, when $thetat, \theta{t-1}$ are all multidimensional yet same dimensional. However, rts_smoother only return the $Cov(\theta_t)$, which is the covariance of every dimension within a state vector itself, but not every pair of dimensions across states at consecutive time step. Can we add that functionality? Or is it already there?
It seems like that when running rts_smoother, it can also calculate $Cov(\thetat, \theta{t-1})$, when $thetat, \theta{t-1}$ are all multidimensional yet same dimensional. However, rts_smoother only return the $Cov(\theta_t)$, which is the covariance of every dimension within a state vector itself, but not every pair of dimensions across states at consecutive time step. Can we add that functionality? Or is it already there?