Synopsis: Reproducibility in statistical computing means that other people can repeat what you have done. Sounds pretty easy, but in practice it often isn’t. Ask yourself

  • Do you know exactly which script / code version you have used for each result in your work
  • More generally, could you repeat each step from the raw data to all the outputs you have produced
  • Have you recorded the R version, the version of all packages that you were using
  • Have you recorded the random seed?

Here are a few tips to achieve all that:

http://www.revolutionanalytics.com/webinars/reproducibility-revolution-r-open-and-checkpoint-package