An Introduction to Discrete-Valued Time Series

Appendix C

Computational Aspects

For illustrating the models and methods described in this book, a number of data examples and numerical examples have been presented. The corresponding computations would not have been possible without software support. The unique software solution used was the program R (R Core Team, 2016). This choice was mainly motivated by the fact that R is freely available to everyone and it also runs on every common operating system. It is widely used anyway. Other computational software packages could be used, for example Matlab or Mathematica. To allow the reader to easily do the computations on her/his own, and to modify and extend the computations, all R codes are provided on the companion website: www.wiley.com/go/weiss/discrete-valuedtimeseries as a password-protected zip file (DiscrValTS17). The datasets that are freely available can be found there too. An overview of the R codes used for this book is given in Section C.2, while the datasets are listed in Section C.3.

To simplify code translation into other programming languages, the R codes were written by using only basic commands; that is, commands available in base R (and analogously in other common languages) without the need for special packages (with a few exceptions, as described in Section C.1 below). Another advantage of this approach is that “black boxes” are avoided to some degree, and the code can be better understood and related to the formulae used in this book. On the other hand, ...