Optimal recovery of interpolation operators in Hardy spaces (Q2491219)

Let \(p \geq 1\) and let \(\mathcal{B}\) denote the unit ball in the Hardy space \(H^p\). If \(U\) is a bounded linear functional on \(H^p\) and if \(\alpha\) is an algorithm using some information \(I(f)\), then \(\alpha^*\) is an optimal recovery algorithm for \(U\) if \[ \sup\{| Uf-\alpha^*(I(f))| : f \in {\mathcal{B}}\}=\inf_\alpha\sup\{| Uf-\alpha(I(f))| : f \in {\mathcal{B}}\}\;. \] Let \(0<r<1<R\) and consider \(\omega\) to be a primitve root of unity. The authors identify the optimal algorithm in the case of \(Uf\) being the \((n-1)\)-st Lagrange interpolant at the zeros of \(z^n - r^n\) evaluated at a point \(R\omega^k\) and for \(I(f)= \{f(s\omega^j)\), \(j=0,\ldots,n-1\}\), where \(1>s\not=r\). They also consider optimal recovery in the \(H^2\) case for \(Uf\) being a Lagrange interpolant or a Taylor section of \(f\) evaluated in a certain point.