On limits of PV k-tuples

From MaRDI portal





For a positive integer \(k\), a PV \(k\)-tuple is a set \(\theta=\{\alpha_{1},\ldots,\alpha_{k}\}\) of algebraic integers with \(|\alpha_{i}|>1\) for \(i=1,\ldots,k\), such that the monic polynomial of least degree \(P\in\mathbb Z[X]\) which vanishes at \(\alpha_{1},\ldots,\alpha_{k}\) has all its remaining roots in the open unit disk. For \(k=1\) this is the definition of a PV number. Assume further that \(P\) is irreducible and reciprocal. The author addresses the open question of whether \(\theta\) is then a limit of PV \(k\)-tuples. He gives a positive answer under the extra assumption \(|\alpha_{i}|>(3+\sqrt{5})/2\) for each \(i\) such that \(\alpha_{i}\not\in\mathbb R\). He also provides more information in the case \(k=3\). Denote by \(S_{3}\) the set of real algebraic integers \(\alpha>1\) with exactly two other conjugates outside the unit disk, and by \(S_{3}'\) (resp. \(S_{3}\)) the subset of those \(\alpha\) for which these two other conjugates are real (resp. not real). The author proves that the reciprocal units in \(S_{3}'\) (necessarily of degree \(6\)) are limit points of \(S_{3}'\). He also proves that \textit{most} \(\alpha\in S_{3}\) are limit points of \(S_{3}\) and suggests all of them are. The main tool is a result of \textit{D. G. Cantor} [Trans. Am. Math. Soc. 105, 391-406 (1962; Zbl 0135.09002)] which gives a necessary and sufficient condition for a PV \(k\)-tuple \(\theta\) with an irreducible \(P\) to be a limit point of PV \(k\)-tuples.











This page was built for publication: On limits of PV k-tuples

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4703817)