On a theorem of Gauss-Kuzmin-Lévy

DOI10.1007/BF01292925MaRDI QIDQ1193290zbMATH OpenFDO

Authors Alexander Durner

Publication date 27 September 1992

Published in Archiv der Mathematik (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/BF01292925

zbMATH Keywords

rate of convergence regular continued fraction theorem of Gauss-Kuzmin-Lévy

Mathematics Subject Classification ID

Metric theory of continued fractions (11K50)

Review text

Let \(\mu\) denote Lebesgue measure on \(J=[0,1]\) and define the distribution functions \[ F_ n(x)=\mu(t\in J:1/\zeta_{n+1}(t)<x)\quad\text{ and }\quad G_ n(x)=\mu(t\in J:\xi_ n(t)<x), \] associated with the complete quotients \(\zeta_ n\) and the incomplete quotients \(\xi_ n\) of the expansion of a number \(t\in J\) as a regular continued fraction. The classical theorem of Gauss- Kuzmin-Lévy states that both \((F_ n)\) and \((G_ n)\) converge uniformly to \(F(x)=\log(1+x)/\log 2\). By using Wirsing's result on the rate of convergence of \((F_ n)\), the author improves on Lévy's result on the rate of convergence of \((G_ n)\), showing \(\| G_ n-F\|_ \infty=O(((3-\sqrt 5)/2)^ n)\) which is best possible.

Recommendations

Cites work

Cited in

(23)

This page was built for publication: On a theorem of Gauss-Kuzmin-Lévy

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1193290)