Digitised Gaussian ARMA Models for Integer-valued time series: An R Package
Maintainer: Hannah Lennon
All statistical R codes used to implement a Monte Carlo Expectation Maximisation tent Class algorithm for maximum likelihood estimation of the dependence parameters of a digitised Gaussian ARMA model are provided in a package named copulaIVTS.
Details are described in the manuscript “Estimation of a digitised Gaussian ARMA model by Monte Carlo Expectation Maximisation”, and further described in PhD Thesis format here (“https://www.research.manchester.ac.uk/portal/files/54587851/FULL_TEXT.PDF“).
Lennon H, & Yuan J, Estimation of a digitised Gaussian ARMA model by Monte Carlo Expectation Maximisation, Computational Statistics & Data Analysis 2018;8:e020683. Available at https://doi.org/10.1016/j.csda.2018.10.015
Lennon, H., 2016. Gaussian copula modelling for integer-valued time series (Doctoral dissertation, University of Manchester).