| dlmModARMA {dlm} | R Documentation |
The function creates an object of class dlm representing a specified univariate or multivariate ARMA process
dlmModARMA(ar = NULL, ma = NULL, sigma2 = 1, m0, C0, dV)
ar |
a vector or a list of matrices (in the multivariate case) containing the autoregressive coefficients. |
ma |
a vector or a list of matrices (in the multivariate case) containing the moving average coefficients. |
sigma2 |
the variance (or variance matrix) of the innovations. |
m0 |
returned as m0 in the created dlm object. It defaults
to a vector of zeroes. |
C0 |
returned as C0 in the created dlm object. It defaults
to a 1e7 * diag(k), where k is the dimension of the
state vector in the DLM representation of the specified ARMA model. |
dV |
the variance, or the diagonal elements of the variance
matrix, of the observational noise. V is assumed to be
diagonal and it defaults to zero. |
The returned DLM only gives one of the many possible representations of an ARMA process.
The function returns an object of class dlm representing the ARMA
model specified by ar, ma, and sigma2.
Giovanni Petris, GPetris@uark.edu
Durbin and Koopman, Time series analysis by state space methods, Oxford University Press, 2001.
dlmModPoly, dlmModSeas,
dlmModReg
## ARMA(2,3)
dlmModARMA(ar = c(.5,.1), ma = c(.4,2,.3), sigma2=1)
## Bivariate ARMA(2,1)
dlmModARMA(ar = list(matrix(1:4,2,2), matrix(101:104,2,2)),
ma = list(matrix(-4:-1,2,2)), sigma2 = diag(2))