The index of Sturmian sequences (Q1348765): Difference between revisions

If \(u\), \(v\), \(w\) are words on a (finite) alphabet, with respective lengths \(|u|\), \(|v|\), \(|w|\), such that \(w=uv\), the fractional power \(w^r\), where \(r= n+ \frac{|u|}{|w|}\) (\(n\) an integer), is defined to be \(w^r:= (uv)^n u\) [see for example \textit{F. Dejean}, J. Comb. Theory, Ser. A 13, 90-99 (1972; Zbl 0245.20052)]. If \(x\) is an infinite word on a finite alphabet and \(w\) a factor (subword) of \(x\), the index of \(w\) is defined by \(\operatorname {ind}(w)= \sup\{r\in \mathbb{Q}\); \(w^r\) is a factor of \(x\}\). The index of \(x\) is the sup of \(\operatorname {ind}(w)\) for all factors of \(x\). The main result of this paper is a formula giving the index of a Sturmian sequence in terms of the continued fraction expansion of its slope. Note that a recent preprint of \textit{W. T. Cao} and \textit{Z.-Y. Wen}, ``Some properties of the Sturmian sequences'' give inter alia the integer index of Sturmian sequences, using singular words. Note that Reference [12] has appeared: \textit{M. Lothaire}, Algebraic combinatorics on words, Cambridge University Press (2002; Zbl 1001.68093).