Overview

R is an open source statistics program, i.e. it is not only free of charge, but anyone who wants to can help improve the software. R can be used for many things. From the ground up it can already do some things. The best thing is, however, that everyone can contribute additional functions: so-called libraries/packages. The packages can be installed directly from within the program. This way, everyone can install what they need and the computer is not filled with functions that are never needed. So far I have programmed two packages myself and contributed to another package to improve it. More information below.

Once you know R, it’s pretty intuitive. At the beginning, however, it can be difficult to learn the basics. It is important that you try it and don’t give up too quickly. The success will come at some point and with it the fun of the program.

For help, I have created the web page politikwissenschaften.ch, which contains among others some R tutorials. Also very helpful is datacamp.com. Students of the Institute for Political Science at the University of Zurich can register for a free Premium Account. Students from other institutes or universities can contact one of their lecturers who can create a education group free of charge.

Students of the Institute for Political Science at the University of Zurich also have the opportunity to visit my R-Helpdesk for questions. Information on this can be found at politischewissenschaften.ch/rhelpdesk_registration.php.

Own Packages

  • glm.predict: Functions to calculate predicted probabilities and discrete changes for glm, ordinal logistic regression and multinomial logistic regression with confident intervals.
  • brant: Function to conduct the brant test. Tests the parallel regression assumption for orginal logistic regressions.

Constributed Packages

  • fastDummies: Function to create dummy variables from factors fast and easy.