LLFGN: Concentrated Loglikelihood Function for H
In FGN: Fractional Gaussian Noise and power law decay time series model fitting

Description Usage Arguments Value Author(s) References See Also Examples

The concentrated loglikelihood, that is, the loglikelihood function maximized over the innovation variance parameter, is computed.

1	LLFGN(H, z)

`H`	parameter
`z`	data vector, assumed to be mean corrected

the value of the loglikelihood

A.I. McLeod

McLeod, A.I., Yu, Hao, Krougly, Zinovi L. (2007). Algorithms for Linear Time Series Analysis, Journal of Statistical Software.

FitFGN, DLLoglikelihood

#compute loglikelihood for NileFlowCMS with H=0.9
data(NileFlowCMS)
z<-NileFlowCMS
z<-z-mean(z)
LLFGN(0.9, z)


#simulate Gaussian white noise and tabulate the loglikelihood for H=0.40, 0.45, 0.50, 0.55, 0.60
set.seed(4321)
h<-c(0.40, 0.45, 0.50, 0.55, 0.60)
z<-rnorm(500, 100, 50)
z<-z-mean(z)
LL<-numeric(length(h))
for (i in 1:length(h))
	LL[i]<-LLFGN(h[i],z)
matrix(c(h,LL),ncol=2)