sigex.add: Build the model by adding on another latent component

In jlivsey/sigex: Ecce Signum: Multivariate Signal Extraction

Build the model by adding on another latent component

Description

# Background: A sigex model consists of process x = sum y, for stochastic components y. Each component process y_t is either stationary or is reduced to stationarity by application of a differencing polynomial delta(B), i.e. w_t = delta(B) y_t is stationary. We have a model for each w_t process, which is specified through the ranks (indices of non-zero Schur complements, cf. background for sigex.param2gcd) of the white noise covariance matrix; also there is the model type, which denotes the specification of the t.s. model for w_t; all the regressors, which are specified by individual time series rather than by latent component, and must have length T; pre-specified bounds for cyclical parameters, for each component, if applicable.

Usage

sigex.add(mdl, vrank, class, order, bounds, name, delta)

Arguments

mdl	The specified sigex model, a list object
vrank	Vector of integers between 1 and N, corresponding to indices of non-zero Schur complements in the GCD of the innovations' covariance matrix for the new latent component
class	Character string of t.s. model type for the new latent component
order	Vector of model order
bounds	Four numbers, gives bounds for rho and omega, the cycle parameters of the new latent component rho lies in (bounds[1],bounds[2]) omega lies in (bounds[3],bounds[4])
name	Character string giving the latent component's name
delta	Differencing polynomial (corresponds to delta(B) in Background) written in format c(delta0,delta1,...,deltad)

Value

mdl: the updated sigex model, a list object

jlivsey/sigex documentation built on March 20, 2024, 3:17 a.m.