scientific article; zbMATH DE number 1741759

From MaRDI portal
Publication:4330088





Given a positive nondecreasing function \(\xi(t)\), \(p\geq 1\) and \(\beta\geq 0\), the set \(W_\beta(L^p,\xi(t))\) is defined as the set of all integrable functions of period \(2\pi\) for which \[ \|[f(\cdot+t)-f(\cdot)]\sin^\beta(\cdot)\|_{L^p[0,2\pi]}=O(\xi(t))\quad\text{as }t\to 0. \] If \(H\) is a regular Hausdorff method with a differentiable mass function and for \(f\in W_\beta(L^p,\xi(t))\) let \((H)_n\) denote the \(n\)th term Hausdorff summabilty of the Fourier series of \(f\), then the author states conditions on \(f\) in order that \(\|f-(H)_n\|=O(\xi(1/n)n^{\beta+1/p})\). As a corollary the author takes \(\beta=0\) and \(\xi(t)=t^\alpha\) and claims that \(\|f-(H)_n\|=O(n^{-\alpha-1/p})\). However, this is not compatible with the previous estimate, so one of them cannot be valid. The reviewer suspects that the mistake is in the former estimate and not in the latter. Also, there is no explantion why the conditions required in the theorem (conditions (8) and (9)), are satisfied when \(\xi(t)\) is so taken.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4330088)