UnscentedTransform¶
- class probnum.filtsmooth.gaussian.approx.UnscentedTransform(dimension, spread=0.0001, priorpar=2.0, special_scale=0.0)¶
Bases:
object
Used for unscented Kalman filter.
See also p. 7 (“Unscented transform:”) of 1.
- Parameters
dimension (int) – Spatial dimensionality
spread (float) – Spread of the sigma points around mean
priorpar (float) – Incorporate prior knowledge about distribution of x. For Gaussians, 2.0 is optimal (see link below)
special_scale (float) – Secondary scaling parameter. The primary parameter is computed below.
References
- 1
Wan, E. A. and van der Merwe, R., The Unscented Kalman Filter, http://read.pudn.com/downloads135/ebook/574389/wan01unscented.pdf
Methods Summary
estimate_statistics
(proppts, sigpts, covmat, …)Computes predicted summary statistics, predicted mean/kernels/crosscovariance, from (propagated) sigmapoints.
propagate
(time, sigmapts, modelfct)Propagate sigma points.
sigma_points
(rv)Sigma points.
Methods Documentation
- estimate_statistics(proppts, sigpts, covmat, mpred)[source]¶
Computes predicted summary statistics, predicted mean/kernels/crosscovariance, from (propagated) sigmapoints.
Not to be confused with mean and kernels resulting from the prediction step of the Bayesian filter. Hence we call it “estimate_*” instead of “predict_*”.
- propagate(time, sigmapts, modelfct)[source]¶
Propagate sigma points.
- Parameters
time (float) – Time \(t\) which is passed on to the modelfunction.
sigmapts (np.ndarray, shape=(2 N+1, N)) – Sigma points (N is the spatial dimension of the dynamic model)
modelfct (callable, signature=(t, x, **kwargs)) – Function through which to propagate
- Returns
M is the dimension of the measurement model
- Return type
np.ndarray, shape=(2 N + 1, M),