Weighted sharp Lebesgue law for product hyperinterpolation

Establish the conjectured asymptotic order of the Lebesgue constant for the Gauss–Jacobi/trapezoidal product hyperinterpolation rule on the truncated conic surface for all weights with gamma>-1, namely \Lambda_{n,\gamma}^{V,\prod}\asymp n^{1/2} for -1<\gamma<-1/2 and \Lambda_{n,\gamma}^{V,\prod}\asymp n^{\gamma+1} for \gamma\ge-1/2.

Background

For the truncated conic surface equipped with the Jacobi-type weight parameterized by gamma>-1, the paper studies the Lebesgue constant of degree-n hyperinterpolation generated by positive degree-2n cubature rules. The authors transfer estimates from weighted disk hyperinterpolation through the quadratic disk-to-cone quotient.

The paper proves the upper bound \Lambda_{n,\gamma}V\le C_\gamma n{\gamma+1} for nonnegative half-integer gamma and proves the matching linear upper and lower bounds when gamma=0. It does not establish the conjectured two-regime asymptotic law for the remaining weights, which requires a weighted boundary-kernel analysis or an equivalent estimate for the even restriction of disk hyperinterpolation.

References

This suggests, for the explicit Gauss--Jacobi/trapezoidal product rule, \Lambda_{n,\gamma}{V,\prod}\asymp \begin{cases} n{1/2},&-1<\gamma<-1/2,\ n{\gamma+1},&\gamma\ge-1/2. \end{cases} Equation eq:weighted-conjecture is a conjectural hyperinterpolation statement; the projection theorem alone does not imply it.

eq:weighted-conjecture:

Λn,γV,∏≍{n1/2,−1<γ<−1/2,nγ+1,γ≥−1/2.\Lambda_{n,\gamma}^{V,\prod}\asymp \begin{cases} n^{1/2},&-1<\gamma<-1/2,\\ n^{\gamma+1},&\gamma\ge-1/2. \end{cases}

— Positive Cubature Compression and Optimal Hyperinterpolation Stability on a Conic Surface  (2609.36430 - An et al., 29 Sep 2026) in Section 6, subsection “Weighted case and open sharp law”; equation (weighted-conjecture) and concluding Section