spec.true
is a matrix of coordinates for the spectral density of the
ARMA(3,1) model with AR and MA coefficients (-0.5, 0.4, 0.8) and (0.2)
respectively. This was the model used to simulate the example data sets
Dfexample
, Dpexample2
and Dpexample3
,
which consist of observations of the process at every time point,
at every second time point and every third time point, respectively.
See the package TSA
for functions to calculate values of the spectral density for different ARMA models, and the function arima.sim
for simulating time series from them.
Nason, G.P., Powell, B., Elliott, D. and Smith, P. (2016) Should We Sample a Time Series More Frequently? Decision Support via Multirate Spectrum Estimation. Journal of the Royal Statistical Society, Series A., 179, (to appear).