Yxilon

Abstract

Statistical research has always been at the edge of available computing power. Huge datasets, e.g in DataMining or Quantitative Finance, and computationally intensive techniques, e.g. bootstrap methods, always require a little bit more computing power than is currently available. But the most popular statistical programming language R, as well as statistical programming languages like S or XploRe, are interpreted which makes them slow in computing intensive areas. The common solution is to implement these routines in low-level programming languages like C/C++ or Fortran and subsequently integrate them as dynamic linked libraries (DLL) or shared object libraries (SO) in the statistical programming language.