Abstract
Quadrature formulas (QFs) based on radial basis functions (RBFs) have become an essential tool for multivariate numerical integration of scattered data. Although numerous works have been published on RBF-QFs, their stability theory can still be considered as underdeveloped. Here, we strive to pave the way towards a more mature stability theory for global and function-independent RBF-QFs. In particular, we prove stability of these for compactly supported RBFs under certain conditions on the shape parameter and the data points. As an alternative to changing the shape parameter, we demonstrate how the least-squares approach can be used to construct stable RBF-QFs by allowing the number of data points used for numerical integration to be larger than the number of centers used to generate the RBF approximation space. Moreover, it is shown that asymptotic stability of many global RBF-QFs is independent of polynomial terms, which are often included in RBF approximations. While our findings provide some novel conditions for stability of global RBF-QFs, the present work also demonstrates that there are still many gaps to fill in future investigations.
Similar content being viewed by others
Availability of data and material
Not applicable.
Notes
Henceforth, we will refer to these as “RBF-QFs”.
\(X_N\) is \(\mathcal {S}_{M,d}\)-unisolvent, for instance, when the kernel \(\varphi \) is conditionally positive definite of order d and \(X_N\) is \(\mathbb {P}_{d}(\varOmega )\)-unisolvent, which is a common assumption to ensure uniqueness of RBF interpolants.
P having full rank means that P has full column rank, i.e., the columns of P are linearly independent. This is equivalent to the set of data points being \(\mathbb {P}_d(\varOmega )\)-unisolvent.
Assuming the sequence of points is dense in \(\varOmega \).
For a function from the appropriate native function space, the \(L^\infty (\varOmega )\)-error between the function and its RBF interpolant is in \(\mathcal {O}( \exp ( -C \log h_\textrm{max}(X_N) / h_\textrm{max}(X_N) ) )\); see [98].
In Fig. 10a the cubic PHS RBF-QF first shows third-order convergence before it then settles for second-order convergence. We believe that the observed initial third-order decrease in the error is a combination of the second-order approximation rate of the cubic PHS-RBF interpolant and the decreasing Lebesgue constant \(\Vert C_N\Vert _{\infty }\) in (11). Once the QF is stable (\(\Vert C_N\Vert _{\infty } = \Vert I\Vert _{\infty }\)), the second-order approximation rate dominates the error of the QF, and we thus start to observe second-order convergence.
References
Ames, W.F.: Numerical Methods for Partial Differential Equations. Academic Press, London (2014)
Aziz, I., Khan, W., et al.: Numerical integration of multi-dimensional highly oscillatory, gentle oscillatory and non-oscillatory integrands based on wavelets and radial basis functions. Eng. Anal. Bound. Elem. 36(8), 1284–1295 (2012)
Babuška, I., Melenk, J.M.: The partition of unity method. Int. J. Numer. Methods Eng. 40(4), 727–758 (1997)
Bayona, V.: Comparison of moving least squares and RBF + poly for interpolation and derivative approximation. J. Sci. Comput. 81(1), 486–512 (2019)
Bayona, V.: An insight into RBF-FD approximations augmented with polynomials. Comput. Math. Appl. 77(9), 2337–2353 (2019)
Bos, L., Caliari, M., De Marchi, S., Vianello, M., Xu, Y.: Bivariate Lagrange interpolation at the Padua points: the generating curve approach. J. Approx. Theory 143(1), 15–25 (2006)
Bos, L., De Marchi, S.: Univariate radial basis functions with compact support cardinal functions. East J. Approx. 14(1), 69 (2008)
Bos, L., De Marchi, S., Vianello, M., Xu, Y.: Bivariate Lagrange interpolation at the Padua points: the ideal theory approach. Numer. Math. 108(1), 43–57 (2007)
Brass, H., Petras, K.: Quadrature Theory: The Theory of Numerical Integration on a Compact Interval. Mathematical Surveys and Monographs, vol. 178. AMS, Providence (2011)
Briol, F.X., Oates, C.J., Girolami, M., Osborne, M.A., Sejdinovic, D.: Probabilistic integration: a role in statistical computation? Stat. Sci. 34(1), 1–22 (2019)
Brutman, L.: Lebesgue functions for polynomial interpolation: a survey. Ann. Numer. Math. 4, 111–128 (1996)
Buhmann, M.D.: Radial basis functions. Acta Numer. 9, 1–38 (2000)
Buhmann, M.D.: Radial Basis Functions: Theory and Implementations, vol. 12. Cambridge University Press, Cambridge (2003)
Caflisch, R.E.: Monte Carlo and quasi-Monte Carlo methods. Acta Numer. 1998, 1–49 (1998)
Cavoretto, R., De Rossi, A., Sommariva, A., Vianello, M.: RBFCUB: a numerical package for near-optimal meshless cubature on general polygons. Appl. Math. Lett. 125, 107704 (2022)
Cohen, A., Davenport, M.A., Leviatan, D.: On the stability and accuracy of least squares approximations. Found. Comput. Math. 13(5), 819–834 (2013)
Cohen, A., Migliorati, G.: Optimal weighted least-squares methods. SMAI J. Comput. Math. 3, 181–203 (2017)
Cools, R.: Constructing cubature formulae: the science behind the art. Acta Numer. 6, 1–54 (1997)
Cools, R.: An Encyclopaedia of cubature formulas. J. Complex. 19(3), 445–453 (2003)
Cools, R., Mysovskikh, I., Schmid, H.: Cubature formulae and orthogonal polynomials. J. Comput. Appl. Math. 127(1–2), 121–152 (2001)
Davis, P.J., Rabinowitz, P.: Methods of Numerical Integration. Courier Corporation, North Chelmsford (2007)
De Marchi, S.: On optimal center locations for radial basis function interpolation: computational aspects. Rend. Splines Radial Basis Funct. Appl. 61(3), 343–358 (2003)
De Marchi, S., Schaback, R.: Stability of kernel-based interpolation. Adv. Comput. Math. 32(2), 155–161 (2010)
Dick, J., Kuo, F.Y., Sloan, I.H.: High-dimensional integration: the quasi-Monte Carlo way. Acta Numer. 22, 133 (2013)
Engels, H.: Numerical Quadrature and Cubature. Academic Press, London (1980)
Fasshauer, G.E.: Solving partial differential equations by collocation with radial basis functions. In: Proceedings of Chamonix, vol. 1997, pp. 1–8. Vanderbilt University Press, Nashville (1996)
Fasshauer, G.E.: Meshfree Approximation Methods with MATLAB, vol. 6. World Scientific, Singapore (2007)
Flyer, N., Barnett, G.A., Wicker, L.J.: Enhancing finite differences with radial basis functions: experiments on the Navier–Stokes equations. J. Comput. Phys. 316, 39–62 (2016)
Folland, G.B.: How to integrate a polynomial over a sphere. Am. Math. Mon. 108(5), 446–448 (2001)
Fornberg, B., Flyer, N.: A Primer on Radial Basis Functions with Applications to the Geosciences. SIAM, Philadelphia (2015)
Fornberg, B., Flyer, N.: Solving PDEs with radial basis functions. Acta Numer. 24, 215–258 (2015)
Fuselier, E., Hangelbroek, T., Narcowich, F.J., Ward, J.D., Wright, G.B.: Kernel based quadrature on spheres and other homogeneous spaces. Numer. Math. 127(1), 57–92 (2014)
Genz, A.: Testing multidimensional integration routines. In: Proceedings of International Conference on Tools, Methods and Languages for Scientific and Engineering Computation, pp. 81–94 (1984)
Glasserman, P.: Monte Carlo Methods in Financial Engineering, vol. 53. Springer, Berlin (2013)
Glaubitz, J.: Shock Capturing and High-Order Methods for Hyperbolic Conservation Laws. Logos Verlag Berlin GmbH, Berlin (2020)
Glaubitz, J.: Stable high order quadrature rules for scattered data and general weight functions. SIAM J. Numer. Anal. 58(4), 2144–2164 (2020)
Glaubitz, J.: Stable high-order cubature formulas for experimental data. J. Comput. Phys. 447, 110693 (2021)
Glaubitz, J.: Construction and application of provable positive and exact cubature formulas. IMA J. Numer. Anal. (2022). https://doi.org/10.1093/imanum/drac017.Drac017
Glaubitz, J., Gelb, A.: Stabilizing radial basis function methods for conservation laws using weakly enforced boundary conditions. J. Sci. Comput. 87(2), 1–29 (2021)
Glaubitz, J., Le Meledo, E., Öffner, P.: Towards stable radial basis function methods for linear advection problems. Comput. Math. Appl. 85, 84–97 (2021)
Glaubitz, J., Nordström, J., Öffner, P.: Energy-stable global radial basis function methods on summation-by-parts form. arXiv:2204.03291 (2022)
Glaubitz, J., Öffner, P.: Stable discretisations of high-order discontinuous Galerkin methods on equidistant and scattered points. Appl. Numer. Math. 151, 98–118 (2020)
Haber, S.: Numerical evaluation of multiple integrals. SIAM Rev. 12(4), 481–526 (1970)
Halton, J.H.: On the efficiency of certain quasi-random sequences of points in evaluating multi-dimensional integrals. Numer. Math. 2(1), 84–90 (1960)
Hardy, R.L.: Multiquadric equations of topography and other irregular surfaces. J. Geophys. Res. 76(8), 1905–1915 (1971)
Hesthaven, J.S., Warburton, T.: Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications. Springer, Berlin (2007)
Hlawka, E.: Funktionen von beschränkter Variation in der Theorie der Gleichverteilung. Ann. Mat. Pura Appl. 54(1), 325–333 (1961)
Huybrechs, D.: Stable high-order quadrature rules with equidistant points. J. Comput. Appl. Math. 231(2), 933–947 (2009)
Ibrahimoglu, B.A.: Lebesgue functions and Lebesgue constants in polynomial interpolation. J. Inequal. Appl. 2016(1), 1–15 (2016)
Iske, A.: On the approximation order and numerical stability of local Lagrange interpolation by polyharmonic splines. In: Modern Developments in Multivariate Approximation, pp. 153–165. Springer (2003)
Iske, A.: Radial basis functions: basics, advanced topics and meshfree methods for transport problems. Rend. Sem. Mat. Univ. Pol. Torino 61(3), 247–285 (2003)
Iske, A.: Scattered data approximation by positive definite kernel functions. Rend. Sem. Mat. Univ. Pol. Torino 69(3), 217–246 (2011)
Iske, A., Sonar, T.: On the structure of function spaces in optimal recovery of point functionals for ENO-schemes by radial basis functions. Numer. Math. 74(2), 177–201 (1996)
Kansa, E., Hon, Y.: Circumventing the ill-conditioning problem with multiquadric radial basis functions: applications to elliptic partial differential equations. Comput. Math. Appl. 39(7–8), 123–138 (2000)
Kansa, E.J.: Multiquadrics: a scattered data approximation scheme with applications to computational fluid-dynamics: II solutions to parabolic, hyperbolic and elliptic partial differential equations. Comput. Math. Appl. 19(8–9), 147–161 (1990)
Karvonen, T., Kanagawa, M., Särkkä, S.: On the positivity and magnitudes of Bayesian quadrature weights. Stat. Comput. 29(6), 1317–1333 (2019)
Krommer, A.R., Ueberhuber, C.W.: Computational Integration. SIAM, Philadelphia (1998)
Krylov, V.I., Stroud, A.H.: Approximate Calculation of Integrals. Courier Corporation, North Chelmsford (2006)
Kuipers, L., Niederreiter, H.: Uniform Distribution of Sequences. Courier Corporation, North Chelmsford (2012)
Larsson, E., Fornberg, B.: A numerical study of some radial basis function based solution methods for elliptic PDEs. Comput. Math. Appl. 46(5–6), 891–902 (2003)
Lasserre, J.B.: Simple formula for integration of polynomials on a simplex. BIT Numer. Math. 61(2), 523–533 (2021)
Manly, B.F.: Randomization, Bootstrap and Monte Carlo Methods in Biology, vol. 70. CRC Press, Boca Raton (2006)
Maxwell, J.C.: On approximate multiple integration between limits of summation. In: Proceedings of the Cambridge Philosophical Society, vol. 3, pp. 39–47 (1877)
Mehri, B., Jokar, S.: Lebesgue function for multivariate interpolation by radial basis functions. Appl. Math. Comput. 187(1), 306–314 (2007)
Micchelli, C.A., Rivlin, T.J.: A survey of optimal recovery. In: Micchelli, C.A., Rivlin, T.J. (eds.) Optimal Estimation in Approximation Theory, pp. 1–54. Plenum Press, New York (1977)
Migliorati, G., Nobile, F.: Stable high-order randomized cubature formulae in arbitrary dimension. J. Approx. Theory 275, 105706 (2022)
Minka, T.P.: Deriving quadrature rules from Gaussian processes. Technical report, Statistics Department, Carnegie Mellon University (2000)
Murphy, K.P.: Machine Learning: A Probabilistic Perspective. MIT Press, Cambridge (2012)
Mysovskikh, I.: The approximation of multiple integrals by using interpolatory cubature formulae. In: Quantitative Approximation, pp. 217–243. Elsevier (1980)
Mysovskikh, I.P.: Cubature formulae that are exact for trigonometric polynomials. In: Cools, R., Schmid, H.J. (eds.) TW Reports (2001)
Niederreiter, H.: Random Number Generation and Quasi-Monte Carlo Methods. SIAM, Philadelphia (1992)
O’Hagan, A.: Bayes–Hermite quadrature. J. Stat. Plan. Inference 29(3), 245–260 (1991)
Olver, F.W.J., Olde Daalhuis, A.B., Lozier, D.W., Schneider, B.I., Boisvert, R.F., Clark, C.W., Miller, B.R., Saunders, B.V., Cohl, H.S., McClain, M.A.: NIST Digital Library of Mathematical Functions. Release 1.1.1, March 15, 2021 (2021). http://dlmf.nist.gov/
Platte, R.B., Trefethen, L.N., Kuijlaars, A.B.: Impossibility of fast stable approximation of analytic functions from equispaced samples. SIAM Rev. 53(2), 308–318 (2011)
Punzi, A., Sommariva, A., Vianello, M.: Meshless cubature over the disk using thin-plate splines. J. Comput. Appl. Math. 221(2), 430–436 (2008)
Quarteroni, A., Valli, A.: Numerical Approximation of Partial Differential Equations, vol. 23. Springer, Berlin (2008)
Reeger, J.A.: Approximate integrals over the volume of the ball. J. Sci. Comput. 83, 45 (2020)
Reeger, J.A., Fornberg, B.: Numerical quadrature over the surface of a sphere. Stud. Appl. Math. 137(2), 174–188 (2016)
Reeger, J.A., Fornberg, B.: Numerical quadrature over smooth surfaces with boundaries. J. Comput. Phys. 355, 176–190 (2018)
Reeger, J.A., Fornberg, B., Watts, M.L.: Numerical quadrature over smooth, closed surfaces. Proc. R. Soc. A Math. Phys. Eng. Sci. 472(2194), 20160401 (2016)
Rudin, W.: Real and Complex Analysis. McGraw-Hill Education, New York (1987)
Shankar, V.: The overlapped radial basis function-finite difference (RBF-FD) method: a generalization of RBF-FD. J. Comput. Phys. 342, 211–228 (2017)
Shu, C., Wu, Y.: Integrated radial basis functions-based differential quadrature method and its performance. Int. J. Numer. Meth. Fluids 53(6), 969–984 (2007)
Sommariva, A., Vianello, M.: Meshless cubature by Green’s formula. Appl. Math. Comput. 183(2), 1098–1107 (2006)
Sommariva, A., Vianello, M.: Numerical cubature on scattered data by radial basis functions. Computing 76(3–4), 295 (2006)
Sommariva, A., Vianello, M.: RBF moment computation and meshless cubature on general polygonal regions. Appl. Math. Comput. 409, 126375 (2021)
Sommariva, A., Womersley, R.: Integration by rbf over the sphere. Applied Mathematics Report AMR05/17. University of New South Wales (2005)
Stroud, A.H.: Approximate Calculation of Multiple Integrals. Prentice-Hall, Englewood Cliffs (1971)
Tominec, I., Larsson, E., Heryudono, A.: A least squares radial basis function finite difference method with improved stability properties. SIAM J. Sci. Comput. 43(2), A1441–A1471 (2021)
Tominec, I., Nazarov, M., Larsson, E.: Stability estimates for radial basis function methods applied to time-dependent hyperbolic PDEs. arXiv:2110.14548 (2021)
Trefethen, L.N.: Cubature, approximation, and isotropy in the hypercube. SIAM Rev. 59(3), 469–491 (2017)
Trefethen, L.N.: Approximation Theory and Approximation Practice, Extended Edition. SIAM, Philadelphia (2019)
Trefethen, L.N.: Exactness of quadrature formulas. arXiv:2101.09501 (2021)
van den Bos, L., Sanderse, B., Bierbooms, W.: Adaptive sampling-based quadrature rules for efficient Bayesian prediction. J. Comput. Phys. 417, 109537 (2020)
Watts, M.L.: Radial basis function based quadrature over smooth surfaces. Theses and Dissertations (2016). https://scholar.afit.edu/etd/249
Wendland, H.: Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree. Adv. Comput. Math. 4(1), 389–396 (1995)
Wendland, H.: Fast evaluation of radial basis functions: Methods based on partition of unity. In: Approximation Theory X: Wavelets, Splines, and Applications. Citeseer (2002)
Wendland, H.: Scattered Data Approximation, vol. 17. Cambridge University Press, Cambridge (2004)
Weyl, H.: Über die Gleichverteilung von Zahlen mod. Eins. Math. Ann. 77(3), 313–352 (1916)
Acknowledgements
We thank Toni Karvonen for pointing out the connection between RBF-QFs and Bayesian quadrature. We also thank the anonymous reviewers for their helpful comments on an earlier manuscript draft.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Not applicable.
Additional information
Communicated by Elisabeth Larsson.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was partially supported by AFOSR #F9550-18-1-0316 and ONR #N00014-20-1-2595 (Glaubitz).
Disclaimer: The views expressed in this academic research paper are those of the authors and do not reflect the official policy or position of the United States Government or Department of Defense. In accordance with the Air Force Instruction 51-303, it is not copyrighted, but is the property of the United States government.
A Moments
A Moments
Henceforth, we provide the moments for different RBFs. The one-dimensional case is discussed in Sect. A.1, while two-dimensional moments are derived in Sect. A.2.
1.1 A.1 One-dimensional moments
Let us consider the one-dimensional case of \(\varOmega = [a,b]\) and distinct data points \(x_1,\dots ,x_N \in [a,b]\).
1.1.1 A.1.1 Gaussian RBF
For \(\varphi (r) = \exp ( - \varepsilon ^2 r^2 )\), the moment of the translated Gaussian RBF,
is given by
Here, \(\textrm{erf}(x) = 2/\sqrt{\pi } \int _0^x \exp ( -t^2 ) \, \textrm{d}t\) denotes the usual error function, [73, Section 7.2].
1.1.2 A.1.2 Polyharmonic splines
For \(\varphi (r) = r^k\) with odd \(k \in \mathbb {N}\), the moment of the translated PHS,
is given by
For \(\varphi (r) = r^k \log r\) with even \(k \in \mathbb {N}\), on the other hand, we have
Note that for \(x_n = a\) the first term is zero, while for \(x_n = b\) the second term is zero.
1.2 A.2 Two-dimensional moments
Here, we consider the two-dimensional case, where the domain is given by a rectangular of the form \(\varOmega = [a,b] \times [c,d]\).
1.2.1 A.2.1 Gaussian RBF
For \(\varphi (r) = \exp ( - \varepsilon ^2 r^2 )\), the two-dimensional moments can be written as products of one-dimensional moments. In fact, we have
Here, the multiplicands on the right-hand side are the one-dimensional moments from (31).
1.2.2 A.2.2 Polyharmonic splines and other RBFs
If it is not possible to trace the two-dimensional moments back to the one-dimensional ones, we are in need of another approach. This is, for instance, the case for PHS. We start by noting that for a data points \((x_n,y_n) \in [a,b] \times [c,d]\) the corresponding moment can be rewritten as follows:
with translated boundaries \(\tilde{a} = a - x_n\), \(\tilde{b} = b - x_n\), \(\tilde{c} = c - y_n\), and \(\tilde{d} = d - y_n\). We are not aware of an explicit formula for such integrals for most popular RBFs readily available from the literature. That said, such formulas were derived in [78,79,80] (also see [95, Chapter 2.3]) for the integral of \(\varphi \) over a right triangle with vertices \((0,0)^T\), \((\alpha ,0)^T\), and \((\alpha ,\beta )^T\). Assuming \(\tilde{a}< 0 < \tilde{b}\) and \(\tilde{c}< 0 < \tilde{d}\), we therefore partition the shifted domain \({\tilde{\varOmega } = [\tilde{a},\tilde{b}] \times [\tilde{c},\tilde{d}]}\) into eight right triangles. Denoting the corresponding integrals by \(I_1, \dots , I_8\), the moment \(m(x_n,y_n)\) correspond to the sum of these integrals. The procedure is illustrated in Fig. 11.
The special cases where one (or two) of the edges of the rectangle align with one of the axes can be treated similarly. However, in this case, a smaller subset of the triangles is considered. We leave the details to the reader, and note the following formula for the weights:
Here, \(\delta _0\) denotes the usual Kronecker delta defined as \(\delta _0(x) = 1\) if \(x = 0\) and \(\delta _0(x) = 0\) if \(x \ne 0\). The above formula holds for general \(\tilde{a}\), \(\tilde{b}\), \(\tilde{c}\), and \(\tilde{d}\). Note that all the right triangles can be rotated or mirrored in a way that yields a corresponding integral of the form
More precisely, we have
Finally, explicit formulas of the reference integral \(I_{\text {ref}}(\alpha ,\beta )\) over the right triangle with vertices \((0,0)^T\), \((\alpha ,0)^T\), and \((\alpha ,\beta )^T\) for some PHS can be found in Table 3. Similar formulas are also available, for instance, for Gaussian, multiquadric and inverse multiquadric RBFs.
We note that the approach presented above is similar to the one in [85], where the domain \(\varOmega = [-1,1]^2\) was considered. Later, the same authors extended their findings to simple polygons [84] using the Gauss–Grenn theorem. Also see the recent work [86], addressing polygonal regions that may be nonconvex or even multiply connected, and references therein. It would be of interest to see if these approaches also carry over to computing products of RBFs corresponding to different centers or products of RBFs and their partial derivatives, again corresponding to different centers. Such integrals occur as elements of mass and stiffness matrices in numerical PDEs. In particular, they are desired to construct linearly energy stable (global) RBF methods for hyperbolic conservation laws [35, 39, 40].
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Glaubitz, J., Reeger, J.A. Towards stability results for global radial basis function based quadrature formulas. Bit Numer Math 63, 6 (2023). https://doi.org/10.1007/s10543-023-00956-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10543-023-00956-0
Keywords
- Numerical integration
- Quadrature
- Cubature
- Radial basis functions
- Stability
- Cardinal functions
- Discrete orthogonal polynomials