Depends R (> = 2.3.0) Imports MASS Suggests RUnit, knitr

This package provides functions for estimation, testing and diagnostic checking of generalized lin-ear autoregressive moving average (GLARMA) models for discrete valued time series with regres-sion variables. These are a class of observation driven non-linear non-Gaussian state space mod-els. The state vector consists of a linear regression component plus an observation driven compo-nent consisting of an autoregressive-moving average (ARMA) filter of past predictive residu-als. Currently three distributions (Poisson, negative binomial and binomial) can be used for the re-sponse series. Three options (Pearson, score-type and unscaled) for the residuals in the observa-tion driven component are available. Estimation is via maximum likelihood (conditional on ini-tializing values for the ARMA process) optimized using Fisher scoring or Newton Raphson iter-ative methods. Likelihood ratio and Wald tests for the observation driven component allow test-ing for serial dependence in generalized linear model settings. Graphical diagnostics includ-ing model fits, autocorrelation functions and probability integral transform residuals are in-cluded in the package. Several standard data sets are included in the package