On a conjecture of F. Móricz and X. L. Shi (Q2644885): Difference between revisions

Let f(x,y) be a continuous function, \(2\pi\)-periodic in x and y, \[ \omega_ x(f,\delta)=\sup_{| u| \leq \delta}\max_{x,y}| f(x+u,y)-f(x,y)| \] its partial modulus of continuity, \[ Lip(1,1)=\{f:\;\omega_ x(f,\delta)=O(\delta)\text{ and } \omega_ y(f,\delta)=O(\delta)\}, \] and \[ \tilde f^{(1,1)}(x,y)=\frac{1}{\pi^ 2}\int^{\pi}_{0}\frac{1}{2} \cot \frac{v}{2}dv\cdot \] \[ \cdot \int^{\pi}_{0}\{f(x+u,y+v)-f(x-u,y+v)- f(x+u,y-v)+f(x-u,y-v)\}\frac{1}{2} \cot \frac{u}{2}du \] the conjugate function. The author proves that there exists a function \(f\in Lip(1,1)\) such that \[ \limsup_{\delta \to +0}(\delta \ln^ 2\frac{1}{\delta})\omega_ x(\tilde f^{(1,1)},\delta)>0. \] This is an affirmative answer to one of the conjectures raised in J. Approximation Theory 49, 346-377 (1987; Zbl 0617.42013) by \textit{X. L. Shi} and the reviewer.