SDtoMAinf: Find the moving average representation of a time series based...

In gregorkb/tscourse: Contains several functions for time series analysis.

Find the moving average representation of a time series based on the spectral density.

Description

Find the moving average representation of a time series based on the spectral density.

Usage

SDtoMAinf(f, lambda, trun = 500, tol = 1e-04)

Arguments

f	a vector with evaluations of the spectral density
lambda	the frequencies to which the values of f correspond. Should be evenly spaced and dense between -pi and pi.
trun	the number of terms to keep in the moving average representation of the time series
tol	controls the accuracy. Smaller values require more computation time.

Value

a list containing a vector containing the coefficients of the moving average representation of the time series as well as the standard deviation of the white noise in the MA-infinity representation. This is based on the work of Pourahmadi (1984).

References

Pourahmadi, M. (1984). Taylor expansion of and some applications. The American Mathematical Monthly, 91(5), 303-307.

gregorkb/tscourse documentation built on Oct. 3, 2022, 5:31 p.m.