garma - R package for estimation of Gegenbauer Seasonal/Cyclical long memory processes.

Lifecycle: experimental Licence

Overview & Introduction

This package fits a GARMA model (refer documentation) to a univariate time series.

GARMA models are extensions of ARIMA models which allow for both fractional differencing (like “fracdiff”) but also allow thta to happen at a non-zero frequency in the spectrum.

This package will estimate that frequency.

At time of writing several estimation methods are supports as well as a number of (non-linear) optimisation routines.

However only k=1 models may be fit (ie only a single Gegenbauer factor) and also only non-seasonal integer differencing of d=0 or d=1 is supported (however you can always manually implement further differencing yourself prior to calling these routines).

Installation.

Ensure you have the “devtools” package installed:

> install.packages('devtools')

After this you can install this package by typing:

> remotes::install_github('rlph50/garma')

Documentation

An Introduction to the “garma” packages is available here, and the reference documentation is available here here.