Creating reproducible publication quality graphics with R: A tutorial

The publication of results of an empirical analysis often includes graphical representations. One of the particular strengths of the statistical computing language and environment R are the graphical capabilities. Beside the relatively simple base graphic system a number of more sophisticated alternative graphic systems are available within R. Compared to these the base graphic system is easy to use but nevertheless highly customizable and therefore advantageous for many research projects. However, even in the base graphic system the number of graphical functions and parameters is large. Due to this it can be hard to find the functions and parameters which must be used to customize a graphic for publication purposes. This paper provides a brief tutorial for this along with a set of simple rules how to structure the code to achieve a sufficient level of reproducibility for a data analysis in general and for associate graphical results in specific. The paper summarizes experiences from a R tutorial hold at the University of Wuppertal for a number of times. The code snippets in the text and the code to reproduce the given examples in the appendix is copy-and-paste-ready and can easily be adopted for own tasks.