Linear recurrent sequences and irrationality measures

From MaRDI portal





Consider the recurrence relation \((n+1)u_{n+1}-(2n+1+\alpha)u_ n+x(n+\alpha)u_{n-1}=0\) for all \(n\in {\mathbb{Z}}_{>0}\) and \(\alpha\),x fixed rational numbers. Consider the solutions \((a_ n)=(1,1+\alpha,...)\) and \((b_ n)=(0,1,...)\). Then, as \(n\to \infty\), the numbers \(a_ n/b_ n\) converge to the number \(\ell =_ 2F_ 1(1,1/2,(\alpha +3)/2,x)/(1+\alpha),\) where \({}_ 2F_ 1\) is Gauss' hypergeometric function. This convergence is for small x strong enough to provide an irrationality proof of \(\ell\) together with irrationality measure. Although the author claims that there is almost no intersection with Huttner's work, the truth is, that the author's results form a subset of results that can be derived using Padé-approximations. For example, the \(a_ n\) are Gegenbauer polynomials in x. It is well known that Gegenbauer polynomials are special cases of Jacobi polynomials and these, in their turn occur as denominators in the Padé approximations to \({}_ 2F_ 1(1,b,c,x)\). For Gegenbauer polynomials we have \(b=1/2\). For the general derivation of Padé approximations to hypergeometric functions see \textit{M. Huttner} [C. R. Acad. Sci., Paris, Sér. I 302, 603-606 (1986; Zbl 0607.10026)]. The case just mentioned is a degenerate case of it.











This page was built for publication: Linear recurrent sequences and irrationality measures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q807666)