Jun 23, 2015: Arend Rensink: Model-Driven Paper Writing, or: Being Real about your Data

June 23, 2015Model-Driven Paper Writing, or: Being Real about your Data
Room: HB 2AArend Rensink

When we write a paper, we often include listings and experimental values from actual cases and runs. Let's just call all of that "data". Obviously we have a scientific obligation to ensure that the data in the paper is real; i.e., the listings are the actual ones used to run your experiments, the reported values are actually obtained by running your program on an actual machine. However, there is a danger that your paper will get out of sync with the real data, if (as is often the case) your experiments are still being tweaked as you write it.

Since formatting a table in LaTeX is an absolute pain and updating it even more so, it is easy to forget things or make errors.


In this talk, I want to show you some, mostly LaTeX-based, tips and tricks that can help you in maintaining the connection between your real data and your paper. Essentially, you can compile your paper from the actual data sources, possibly after applying some transformation. Though it takes some time and thought to set this up properly, I firmly believe that this or similar techniques should become part of our standard way of working.