Munich Personal RePEc Archive

Numerical accuracy of Ox 6.2

Kulaksizoglu, Tamer (2012): Numerical accuracy of Ox 6.2.

[img]
Preview
PDF
MPRA_paper_38219.pdf

Download (341kB) | Preview

Abstract

This article evaluates the numerical accuracy of Ox object-oriented matrix programming language. It uses the Standard Reference Datasets (StRD) for estimation accuracy, the ELV DOS program for statistical distribution accuracy, and the DIEHARD battery of randomness tests for random number generation accuracy. The main finding of the paper is that Ox is quite accurate in almost every area it is tested. However, it should also be stressed that there is room for improvement in its optimization library.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.