SSsimple (version 0.6.6)

SS.stst.tv: Steady State

Description

Find steady state of time-varying system, i.e., locate when Kalman gain converges

Usage

SS.stst.tv(F, H, Q, R, P0, epsilon, verbosity=0)

Arguments

F

A list of d x d matrices.

H

A list of n x d matrices.

Q

A list of d x d matrices.

R

A list of n x n matrices.

P0

Initial a priori prediction error.

epsilon

A small scalar number.

verbosity

0, 1 or 2.

Value

A named list.

P.apri

A d x d matrix giving a priori prediction variance.

P.apos

A d x d matrix giving a posteriori prediction variance.

Details

Note: The test for convergence has been (very, very slightly) modified since v0.5.1. The current test has been implemented for rigor. Users who have results based on earlier releases may observe infinitesimal differences in the resulting prediction error.

Examples

# NOT RUN {
F.tv <- list()
for(i in 1:10000) {
	F.tv[[i]] <- diag( c(1/(i+10), 1/(i+10)) )
}

H <- matrix(1, 2, 2)

SS.stst.tv(F.tv, H, 1, 1, 10^5, 10^(-10), verbosity=2)
# }