On inhomogeneous continued fraction expansions and inhomogeneous diophantine approximation (Q676290)

From MaRDI portal
scientific article
Language Label Description Also known as
English
On inhomogeneous continued fraction expansions and inhomogeneous diophantine approximation
scientific article

    Statements

    On inhomogeneous continued fraction expansions and inhomogeneous diophantine approximation (English)
    0 references
    0 references
    5 January 1998
    0 references
    Given any pair of real numbers \(\vartheta,\varphi\), define the inhomogeneous approximation constants as \[ {\mathcal M}(\vartheta,\varphi)= \min\{{\mathcal M}_+(\vartheta,\varphi), {\mathcal M}_-(\vartheta,\varphi)\}, \] where \({\mathcal M}_+(\vartheta,\varphi)= \liminf_{q\to\infty} (q|q\vartheta- \varphi|)\) and \({\mathcal M}_-(\vartheta,\varphi)= {\mathcal M}_+(\vartheta,-\varphi)\). There are several known algorithms for inhomogeneous diophantine approximation which have been or can be used to deal with these approximation constants [see \textit{J. W. S. Cassels}, Math. Ann. 127, 288-304 (1954; Zbl 0055.04401); \textit{R. Descombes}, Ann. Sci. Éc. Norm. Supér., III. Sér. 73, 283-355 (1956; Zbl 0072.03802); \textit{V. Turán-Sos}, Acta Math. Acad. Sci. Hung. 9, 229-241 (1958; Zbl 0086.03902); \textit{K. Nishioka}, \textit{I. Shiokawa} and \textit{J.-I. Tamura}, J. Number Theory 42, 61-87 (1992; Zbl 0770.11039); \textit{S. Itô} and \textit{H. Tachii}, Tokyo J. Math. 16, 261-289 (1993; Zbl 0795.11027); \textit{J. M. Borwein} and \textit{P. B. Borwein}, J. Number Theory 43, 293-318 (1993; Zbl 0778.11039); \textit{T. W. Cusick}, \textit{A. Rockett} and \textit{P. Szüsz}, J. Number Theory 48, 259-283 (1994; Zbl 0820.11042)]. Each of these algorithms expresses \(\varphi\), in some way, in terms of the approximation denominators \(q_n\) provided by the ordinary continued fraction expansion of \(\vartheta\). The author exhibits a new algorithm which allows in principle for the computation of the above constants. He proves the formula \[ {\mathcal M}_+(\vartheta,\varphi)= \liminf_{n\to\infty} (k_n|k_n\vartheta-\varphi|) \] where \(k_n\) is an integer which realises the minimal value of \(|k\vartheta-\varphi|\), \(0\leq k<q_n\). He uses this for the effective computation of \({\mathcal M}(\vartheta^*,1/a)\), \({\mathcal M}(\vartheta^*,1/(2a))\), and \({\mathcal M}(\vartheta^*,1/(a^2+4)^{1/2})\) when \(\vartheta^*= ((a^2+4)^{1/2}-a)/2\).
    0 references
    inhomogeneous approximation constants
    0 references
    inhomogeneous diophantine approximation
    0 references
    continued fraction expansion
    0 references
    algorithm
    0 references

    Identifiers