Abel summability of Jacobi type series (Q1358733)

Let \(\{ P_n^{(\alpha, \beta)}\}\) be the sequence of Jacobi orthonormal polynomials corresponding to the weight \(J^{(\alpha, \beta)}(x) =(1-x)^\alpha(1+x)^\beta\) on \([-1,1]\). Then, the sequence of Jacobi functions \[ F_n^{(\alpha, \beta)} = P_n^{(\alpha, \beta)} \sqrt{J^{(\alpha, \beta)}} \] is orthonormal with respect to the Lebesgue measure on \([-1,1]\). In the \(m\)-dimensional setting take \[ P_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X)=\prod_{j=1}^m P_{n_j}^{(\alpha_j, \beta_j)}(x_j), \qquad F_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X)=\prod_{j=1}^m F_{n_j}^{(\alpha_j, \beta_j)}(x_j), \] \[ J^{(\vec{\alpha}, \vec{\beta})}(X)=\prod_{j=1}^m J^{(\alpha_j, \beta_j)}(x_j), \] where \(\vec{n}\), \(\vec{\alpha}\), \(\vec{\beta}\) are multiindices and \(X= (x_1, \dots, x_m) \in Q=[-1,1]^m\). Thus, two types of Fourier series can be considered: for \(f \in L^1(J^{(\vec{\alpha}, \vec{\beta})})\), \[ \sum c_{\vec{n}} P_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X) , \] and for \(f \in L^1(Q)\), \[ \sum c_{\vec{n}} F_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X) , \] where \(c_{\vec{n}}\) are the Fourier coefficients for \(f\). Correspondingly, Abel approximations \[ f(\vec{r},X)=\sum_{\vec{n}} r_1^{n_1}\cdots r_m^{n_m} c_{\vec{n}} P_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X), \qquad \tilde{f}(\vec{r},X)=\sum_{\vec{n}} r_1^{n_1}\cdots r_m^{n_m} c_{\vec{n}} F_{\vec{n}}^{(\vec{\alpha}, \vec{\beta})}(X), \] can be constructed. Convergence of \(f(\vec{r},X)\) to \(f(X)\) as \(r \to 1^-\) with some restrictions on the components of \(\vec{r}\) was studied by \textit{L. A. Caffarelli} and \textit{C. P. Calderón} [Colloq. Math. 30, 277-288 (1974; Zbl 0291.40016)]; there, \(L^1\)-weak type estimates for the restricted maximal operator played a leading role. Here, the authors consider an analogous problem for \(\tilde{f}(\vec{r},X)\). Continuing the cited work of Caffarelli and Calderón, they extend the weak type estimate to the restricted maximal operator of the Abel sum \(\tilde{f}(\vec{r},X)\) and generalize the main theorem to a family of weights \[ J^{(\vec{\alpha}^*, \vec{\beta}^*)}(X)=\prod_{j=1}^m J^{(\alpha_j^*, \beta_j^*)}(x_j), \] where \(\alpha_j^* \geq \alpha_j \geq 0\), etc. Finally, a unified approach to both problems is considered.