PyKF.KalmanFiltering.UnscentedFilter

Package PyKF :: Module KalmanFiltering :: Class UnscentedFilter

[show private | hide private]

Class UnscentedFilter

Known Subclasses:: BevertonHoltFilter1, BevertonHoltFilter2

Unscented Kalman filter, using Beverton-Holt recruitment as an example.
Algorithms based on van der Merwe and Wan (2001)
    
Measurement equation:
    y[t] = self.measure(x[t], v[t]) where v[t] ~ N(0,V[t])
    
Process equation:
    x[t+1] = F(x[t], w[t]) where w[t] ~ N(0,W[t])

This algorithm also (optionally) updates the measurement variance
according to West and Harrison (1997), pp52-57.

Method Summary
	`__init__(self, process_covariance, measurement_covariance)`
	`filter(self, x_hat, P_hat, y, args, *kwargs)`
	`get_sigma_points(self, x, P)`
	`get_weights(self, L, type)`
	`Lambda(self, L)`
	`measurement(self, x, args, *kwargs)`
	`process(self, x, args, *kwargs)`

Generated by Epydoc 2.1 on Tue Mar 22 13:16:54 2005

http://epydoc.sf.net