btcmodel: Modelo estrutural tendencia e ciclo bayesiano

In santoscs/mecnost: Modelos Estruturais de Componentes Nao Observados para Series Temporais

Description

Estima o modelo estutrural de componentes nao observados para a tendencia estocastica e ciclo (gap) estacionario de forma bayesiana com MCMC e Gibbs sample. Para detalhes ver secao 4.6.1 de Petris e Petrone (2009).

Usage

btcmodel(y, a.theta, b.theta, dV = 1e-07, m0 = rep(0, 4), C0 = diag(x =
  c(rep(1e+07, 2), rep(1, 2))), n.sample = 1, thin = 0,
  save.states = FALSE)

Arguments

y	data vector or univariate time series
a.theta	prior mean of system precisions (recycled, if needed)
b.theta	prior variance of system precisions (recycled, if needed)
dV	the variance, or the diagonal elements of the variance matrix in the multivariate case, of the observation noise. V is assumed to be diagonal and it defaults to zero
m0	the expected value of the pre-sample state vector
C0	the variance matrix of the pre-sample state vector
n.sample	requested number of Gibbs iterations
thin	discard thin iterations for every saved iteration
save.states	should the simulated states be included in the output?

Value

a list of Markov chain Monte Carlo (MCMC) simulated values.

phi	simulated values of the AR parameter
vars	simulated values of the unknown variance
theta	simulated values of the state vectors

References

PETRIS, G.; PETRONE, S.; CAMPAGNOLI, P. Dynamic Linear Models with R. New York: Springer Science, 2009.

santoscs/mecnost documentation built on May 29, 2019, 1:48 p.m.