English

Computing Absolutely Normal Numbers in Nearly Linear Time

Data Structures and Algorithms 2020-07-17 v4

Abstract

A real number xx is absolutely normal if, for every base b2b\ge 2, every two equally long strings of digits appear with equal asymptotic frequency in the base-bb expansion of xx. This paper presents an explicit algorithm that generates the binary expansion of an absolutely normal number xx, with the nnth bit of xx appearing after nnpolylog(n)(n) computation steps. This speed is achieved by simultaneously computing and diagonalizing against a martingale that incorporates Lempel-Ziv parsing algorithms in all bases.

Keywords

Cite

@article{arxiv.1611.05911,
  title  = {Computing Absolutely Normal Numbers in Nearly Linear Time},
  author = {Jack H. Lutz and Elvira Mayordomo},
  journal= {arXiv preprint arXiv:1611.05911},
  year   = {2020}
}
R2 v1 2026-06-22T16:56:27.867Z