A counter example in the Pick-Nevanlinna interpolation problem (Q1114119)

Let \(\Lambda\) be an arbitrary nonempty set and let K be a positive- definite kernel on \(\Lambda\). Denote by \(H_ K\) the associated functional Hilbert space on \(\Lambda\) with K as its reproducing kernel, and assume for simplicity that \(1\in H_ K\). Let \(E\subseteq \Lambda\) be a set of uniqueness for \(H_ K\), and let g be a function on E so that the kernel \(K(x,y)(1-g(x)\overline{g(y)})\) is positive-definite in E. Let G be the (unique) extension of g to \(\Lambda\) such that \(G\in H_ K\). \textit{F. Beatrous} and \textit{J. Burbea} [``Reproducing kernels and interpolation of holomorphic functions'', Complex analysis, functional analysis, and approximation theory, (ed. J. Mujica), North-Holland, Amsterdam, 25-46 (1986)] have posed the question whether or not, in general, the kernel K(x,y)(1-G(x)\(\overline{G(y)})\) is positive-definite on \(\Lambda\). By means of a simple example, the author shows that this extended kernel is not necessarily positive-definite on \(\Lambda\).