Weak abelian periodicity of infinite words (Q504987)

We start by introducing some definitions necessary for understanding the paper's contribution. For a finite word \(w\), the frequency \(\rho_{a}\left( w\right) \) of a letter \(a\) in \(w\) is defined as \(\rho_{a}\left( w\right) \) = \(\frac{\left| w\right| _{a}}{\left| w\right| }\) where \(\left| w\right| _{a}\) is the number of occurrences of \(a\) in \(w\) and \(\left| w\right| \) is the length of \(w\). In an infinite word \(w\), a letter \(a\) has frequency \(\rho_{a}\left( w\right) \)= \(\lim_{n\rightarrow\infty}\rho_{a}(\mathrm{pref}_{n}(w))\) where pref\(_{n}(w)\) is the prefix of length \(n\) of \(w\). An infinite word is said to have letter frequencies if this limit exists for each letter. An infinite word \(w\) is called weakly abelian periodic (WAP) if it is an infinite concatenation of words with identical frequencies of letters. Such infinite word is called bounded WAP if the length of all factors in this concatenation is bounded by a constant. An infinite word is recurrent if each of its factors occurs infinitely many times. It is uniformly recurrent if the the length of all gaps between two consecutive occurrences of the same factor is limited by a constant (depending on this factor). In the case of a binary alphabet, the authors associate an infinite word \(w=a_{1}a_{2}\cdots\) with a path consisting of linear segments connecting integer points in \(\mathbb{Z}_{2}\). (They claim that a similar construction is possible for larger alphabets in spaces of higher dimensions.) The path starts at the point \(\left( x_{0},y_{0}\right) =\left( 0,0\right) \). Each segment connects a point \(\left( x_{n},y_{n}\right) \) with a point \(\left( x_{n+1},y_{n+1}\right) \), where \(x_{n+1}=x_{n}+1\), and \(y_{n+1}=y_{n}-1\) if \(a_{n}=0\) and \(y_{n+1}=y_{n}+1\) if \(a_{n}=1\). An infinite word \(w\) is of bounded width if there exist two lines with the same rational slope, so that this path lies between them. From multiple results on infinite WAP words presented in the paper, let us mention the following sample.\newline1. Every bounded WAP word is of bounded width; every word of bounded width is WAP, but not necessarily bounded. \newline2. Every uniformly recurrent word of bounded width is a bounded WAP while not every uniformly recurrent WAP word is of bounded width.\newline3. If a binary word is not WAP, then it has frequencies of letters.\newline4. If a word is a fixed point of a binary uniform morphism then it is bounded WAP if and only if it is abelian periodic.\newline5. The shuffle of two bounded WAP words is WAP while there exist two WAP words such that their shuffle is not WAP.