Function Reconstruction Using Rank-1 Lattices and Lower Sets

Our study focuses on constructing optimal rank-1 lattices that enable exact integration and reconstruction of functions in the Chebyshev space, based on finite index sets. We introduce novel theoretical lower bounds on the minimum number of integrands needed for reconstruction, show equivalence between different plans for reconstruction under certain conditions and propose an innovative algorithm for generating the optimal generator vector of rank-1 lattices. By leveraging the inherent structure of the set of interpolators, our approach ensures admissibility conditions through exhaustive search and verification, outperforming existing methods in terms of computation time and memory usage. Numerical experiments validate the efficiency and practical applicability of our algorithm.