Identical approximative sequence for various notions of universality (Q1763775)

The authors consider several notions of universality and prove that gene\-ri\-cal\-ly they occur simultaneously with the same approximative sequence. Their result improves or extends a number of earlier statements due to Chui, Parnes, Luh, Melas and Nestoridis, among others. Specifically, let \(G\) be a simply connected domain in the complex plane \(\mathbb{C}\) with \(G \neq C\). If \(t \in G\) and \(f \in H(G) :=\) [the Fréchet space of holomorphic functions on \(G\), endowed with the locally uniform convergence], then \(S_N(f,t)\) \((N \geq 0)\) denotes the partial sum of \(f\) with center \(t\). A sequence \((f^{(-n)})_{n \geq 0} \subset H(G)\) is called a strict sequence of antiderivatives of \(f\) if \(f^{(0)} = f\) and \((f^{(-n-1)})' = f^{(-n)}\) \((n \geq 0)\). For every compact set \(K \subset \mathbb{C}\), it is denoted by \(A(K)\) the class of continuous functions on \(K\) that are holomorphic in the interior of \(K\). Let \((a_n)_{n \geq 1}, \, (b_n)_{n \geq 1}\) be two sequences of complex numbers such that \(a_n \neq 0\) \((n \geq 1)\), every boundary point of \(G\) is an accumulation point of \((b_n)_{n \geq 1}\) and for every compact set \(M \subset \mathbb{C}\) there exists a natural number \(n_0 = n_0(M)\) such that \(a_nz + b_n \in G\) for every \(n \geq n_0\) and every \(z \in M\). The set \(V(G,(a_n)_{n \geq 1},(b_n)_{n \geq 1}) =: V\) is defined as the class of functions \(f \in H(G)\) such that for every choice of compact sets \(K, \, M, \, L_1, \, L_2, \, L_3, \, L_4, \, L_5\) with \(K \subset \Omega^c, \, M \subset \mathbb{C}, \, L_j \subset G \, (1 \leq j \leq 5)\), \(K^c, \, M^c, \, L_4^c, \, L_5^c\) connected, for every \(b \in \partial G\) and for every choice of functions \(h \in A(K), \, \omega \in A(M), \, \psi \in A(L_4), \, \varphi \in A(L_5)\), there exist a sequence \((n_j)_{j \geq 1}\) of natural numbers and a strict universal sequence \((f^{(-n)})_{n \geq 0}\) of antiderivatives of \(f\) such that \(b_{n_j} \to b\) \((j \to \infty )\) and \(\lim_{j \to \infty} [\sup_{z \in L_2} \sup_{z \in L_3} | S_{n_j}(f,t)(z) - f(z)| \,\, + \,\, \sup_{t \in L_1} \sup_{z \in K} | S_{n_j}(f,t)(z)\) \(- h(z)| \,\, + \) \( \sup_{z \in M} | f(a_{n_j}z + b_{n_j}) - \omega (z)| \,\, + \,\, \sup_{z \in L_4} | f^{(n_j)}(z) - \psi (z)| \,\, + \,\, \sup_{z \in L_5} | f^{(-n_j)}(z) - \varphi (z)| ]= 0\). Then the main result of the paper asserts that \(V\) is residual, hence nonempty.