General information

Robust Statistics deals with a very real problem in statistical applications: the effect of violations to the model used to analyze the data. The last 40 years have seen tremendous advancements in the theory of Robust Statistics, but unfortunately many of these procedures are not widely used in practice yet. One reason that contributes to the limited use of Robust Statistics is the heavy computational cost of many of these techniques. The lack of easy to use and well documented computer code does not help either.

In the last few years the consolidation of the R-project as a widely available, powerful and versatile computer program for statistical analysis has resulted in many people simultaneously developing and publishing R code that implements Robust Statistics techniques.

One of the main goals of this project is to organize the development of tools in R that would implement Robust Statistics methods for many widely used models. We plan to agree on a set of guidelines that can unify to some extent the many different software development practices currently used by multiple research groups around the world.