arma_loglikelihood: arma_loglikelihood
In hlennon/copulaIVTS: Dependence modelling of a digitised Gaussian ARMA(p,q) copula model for Integer-Valued Time Series

Description Usage Arguments Details Value Note Author(s) See Also Examples

Evaluates the ARMA(p,q) log-likelihood function at the set of arma parameters e.g. order = c(1,1) data = arma data parameters = arma parameters (phi_1, ..., phi_p, theta_1, ..., theta_q)

1	arma_loglikelihood(parameters, data, ord)

`parameters`	a vector of arma parameters
`data`	a vector of arma data

ord

ord

order of arma model

Evaluates the ARMA(p,q) logliklihood function. 'data' and 'order' must be specified as global variables

scalar value of the arma LOG-likelihood evaluated at vector (parameters) for the given data

Evaluates Reduced Gaussian log-likelihood g <- n*log(2*pi) + n*log(sigma2) + sum(log(kal$S)) + sum(kal$E^2/kal$S)/sigma2 g <- -g/2

Hannah Lennon <drhannahlennon@gmail.com>

arma_lik

library(forecast)

# Example 1
library(gamlss.data)
data(polio)
data  <- polio
order <- c(1,1)
arma_loglikelihood(c(0.7, -0.5), data, c(1,1))



# Example 2
data  <- forecast::arima.sim(model=list(ar =c(0.4, -0.5), ma=c(0.8, 0.2)), n=2000)
order <- c(2,2)
para  <- c(0,0,0,0)

# takes 1 minute to optimise
# optim(para, arma_loglikelihood,  data=data, ord=order, control=list(fnscale=-1))$par
# [1]  0.4569347 -0.5244361  0.7507188  0.1448333