IRIS Toolbox Reference Manual

sspace

State-space matrices describing the model solution

Syntax

[T,R,K,Z,H,D,U,Omg] = sspace(m,...)

Input arguments

Output arguments

Options

Description

The state-space representation has the following form:

[xf;alpha] = T*alpha(-1) + K + R*e

y = Z*alpha + D + H*e

xb = U*alpha

Cov[e] = Omg

where xb is an nb-by-1 vector of predetermined (backward-looking) transition variables and their auxiliary lags, xf is an nf-by-1 vector of non-predetermined (forward-looking) variables and their auxiliary leads, alpha is a transformation of xb, e is an ne-by-1 vector of shocks, and y is an ny-by-1 vector of measurement variables. Furthermore, we denote the total number of transition variables, and their auxiliary lags and leads, nx = nb + nf.

The transition matrix, T, is, in general, rectangular nx-by-nb. Furthremore, the transformed state vector alpha is chosen so that the lower nb-by-nb part of T is quasi upper triangular.

You can use the get(m,'xVector') function to learn about the order of appearance of transition variables and their auxiliary lags and leads in the vectors xb and xf. The first nf names are the vector xf, the remaining nb names are the vector xb.