PNAR: Poisson Network Autoregressive Models

Quasi likelihood-based methods for estimating Poisson Network Autoregression with p lags, PNAR, following generalized linear models are provided. PNAR models with the identity and with the logarithmic link function are allowed. The inclusion of exogenous covariates is also possible. Moreover, it provides tools for testing the linearity of linear PNAR model versus several nonlinear alternatives. Finally, it allows generating multivariate count distributions, from linear and nonlinear PNAR models, where the dependence between Poisson random variables is generated by suitable copulas. References include: Armillotta, M. and K. Fokianos (2022a). Poisson network autoregression. <arXiv:2104.06296>. Armillotta, M. and K. Fokianos (2022b). Testing linearity for network autoregressive models. <arXiv:2202.03852>.

Package details

AuthorMichail Tsagris [aut, cre], Mirko Armillotta [aut, cph], Konstantinos Fokianos [aut]
MaintainerMichail Tsagris <mtsagris@uoc.gr>
LicenseGPL (>= 2)
Version1.1
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:
install.packages("PNAR")

Try the PNAR package in your browser

Any scripts or data that you put into this service are public.

PNAR documentation built on Aug. 23, 2022, 1:05 a.m.