Skip to main content
Log in

Extensions of Clenshaw-Curtis-type rules to integrals over a semi-infinite interval

  • Original Paper
  • Published:
Numerical Algorithms Aims and scope Submit manuscript

Abstract

The Clenshaw-Curtis (C-C) rule is a quadrature formula for integrals on an interval [− 1,1] and efficient for smooth integrands f(x). Analogous rules exist: Fejér’s first and second kind, Basu and corrected C-C rules. We attempt to extend these five rules to integrals over a semi-infinite interval \([0,\infty )\) to develop corresponding formulae. Developing contour integration representations of the errors of the formulae, we prove that for f(z) analytic in a region containing \([0,\infty )\) in the complex plane z, the errors are of O(hjec/h) (\(j=1,2,\dots ,5\)), respectively, with a constant c > 0 as step size h → + 0. The extension of Fejér’s second rule (the case j = 1) agrees with a formula based on the Sinc interpolation. Numerical experiments show that new formulae inherit nice features of the C-C rule and its four analogs. For large h, the convergence rates are twice as fast as the asymptotic rates for small h. The kink phenomenon that is explained in the C-C rule and Fejér’s first rule appears in the convergence curves.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

References

  1. Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions. Dover, New York (1965)

    MATH  Google Scholar 

  2. Ahlfors, L.V.: Complex Analysis; An Introduction to the Theory of Analytic Functions of One Complex Variable. International series in pure and applied mathematics, 3rd edn. McGraw-Hill, New York (1979)

    Google Scholar 

  3. Basu, N.K.: Evaluation of a definite integral using Tschebyscheff approximation. Mathematica 13, 13–23 (1971)

    MathSciNet  MATH  Google Scholar 

  4. Clenshaw, C.W., Curtis, A.R.: A method for numerical integration on an automatic computer. Numer. Math. 2, 197–205 (1960)

    Article  MathSciNet  Google Scholar 

  5. Davis, P.J., Rabinowitz, P.: Methods of Numerical Integration, 2nd edn. Academic Press, Orland (1984)

    MATH  Google Scholar 

  6. Filippi, S.: Angenäherte Tschebyscheff-Approximation einer Stammfunktion – eine Modifikation des Verfahrens von Clenshaw und Curtis. Numer. Math. 6, 320–328 (1964)

    Article  MathSciNet  Google Scholar 

  7. Hasegawa, T., Sugiura, H.: Error estimate for a corrected Clenshaw-Curtis quadrature rule. Numer. Math. 130, 135–149 (2015)

    Article  MathSciNet  Google Scholar 

  8. Mason, J.C., Handscomb, D.C.: Chebyshev Polynomials. Chapman & Hall/CRC, Boca Raton (2003)

    MATH  Google Scholar 

  9. Notaris, S.E.: On a corrected Fejér quadrature formula of the second kind. Numer. Math. 133, 279–302 (2016)

    Article  MathSciNet  Google Scholar 

  10. Stenger, F.: Numerical Methods Based on Sinc and Analytic Functions Springer Series in Computational Mathematics, vol. 20. Springer, New York (1993)

    Book  Google Scholar 

  11. Stenger, F.: Summary of Sinc numerical methods. J. Comput. Appl. Math. 121, 379–420 (2000)

    Article  MathSciNet  Google Scholar 

  12. Sugiura, H., Hasegawa, T.: A truncated Clenshaw-Curtis formula approximates integrals over a semi-infinite interval. Numer. Algorithms 86, 659–674 (2021)

    Article  MathSciNet  Google Scholar 

  13. Takahasi, H., Mori, M.: Error estimation in the numerical integration of analytic functions. Rep. Compt. Centre. Univ. Tokyo 3, 41–108 (1970)

    MathSciNet  Google Scholar 

  14. Temme, N.M.: DLMF: Chapter 6 Exponential, Logarithmic, Sine, and Cosine Integrals (accessed June 19, 2020). https://dlmf.nist.gov/6.12/. NIST (2020)

  15. Trefethen, L.N.: Is Gauss quadrature better than Clenshaw-Curtis? SIAM Rev. 50, 67–87 (2008)

    Article  MathSciNet  Google Scholar 

  16. Waldvogel, J.: Fast construction of the Fejér and Clenshaw-Curtis quadrature rules. BIT 46, 195–202 (2006)

    Article  MathSciNet  Google Scholar 

  17. Waldvogel, J.: Towards a general error theory of the trapezoidal rule. In: Gautschi, W., Mastroianni, G., Rassias, T. (eds.) Approximation and Computation In Honor of Gradimir V. Milovanović, Springer Optimization and its Applications, vol. 42, pp. 267–282. Springer, New York (2011)

  18. Weideman, J.A.C., Trefethen, L.N.: The kink phenomenon in Fejér and Clenshaw-Curtis quadrature. Numer. Math. 107, 707–727 (2007)

    Article  MathSciNet  Google Scholar 

Download references

Acknowledgments

We thank Professor Sotirios E. Notaris for his helpful comments for improving the presentation and for his suggestion on the test function. We are grateful to the referee for his/her suggestion on the trapezoidal rule in the reference [17].

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Hiroshi Sugiura.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix 1: Sinc integral

Appendix 1: Sinc integral

We show that the LFII formula (1.2) is derived from the Sinc interpolation and give the asymptotic convergence rate.

We start with a change of variables x = t2 in the integral Qf,

$$ Qf={\int}_0^{\infty} f(x) dx={\int}_0^{\infty} g(t) dt, $$
(A.1)

where g(t) = 2tf(t2). We then interpolate g(t) in terms of the Sinc function \(\text {Sinc}(z)=\sin \limits (\pi z)/(\pi z)\) (cf. [11]) as follows

$$ g(t)\approx\sum\limits_{k=-\infty}^{\infty} g(kh) \text{Sinc}((t-kh)/h). $$
(A.2)

Using the right-hand side of (A.2) in (A.1), we have the Sinc formula \(I_{h}^{(\text {Sinc})}f\),

$$ Qf\approx I_{h}^{(\text{Sinc})}f=\sum\limits_{k=-\infty}^{\infty} g(kh) {\int}_0^{\infty}\text{Sinc}\left( \frac{t-kh}{h}\right) dt. $$
(A.3)

Since by a change of variables u = (tkh)/h, we have

$$ {\int}_0^{\infty}\text{Sinc}\left( \frac{t-kh}{h}\right) dt =h{\int}_{-k}^{\infty}\frac{\sin \pi u}{\pi u} du =\frac{h}{\pi}{\int}_{-\pi k}^{\infty}\frac{\sin u}{u} du, $$

in view of (A.3), it follows that

$$ I^{(\text{Sinc})}_{h} f=\sum\limits_{k=-\infty}^{\infty} g(kh) \frac{h}{\pi}[\text{Si}(\infty) -\text{Si}(-\pi k)]. $$
(A.4)

In view of (A.4) and the relations g(−x) = −g(x) and Si(−x) = −Si(x), recalling that g(t) = 2tf(t2), we obtain

$$ I_{h}^{(\text{Sinc})}f=\frac{2h}{\pi}\sum\limits_{k=1}^{\infty} g(kh) \text{Si}(\pi k)= \frac{4h}{\pi}\sum\limits_{k=1}^{\infty} kh \text{Si}(\pi k) f(k^{2}h^{2}). $$

Then, for \(I_{h}^{(\text {LFII})}f\) in (1.2), we see that

$$ I_{h}^{(\text{Sinc})}f=I_{h}^{(\text{LFII})}f. $$

Proof of the asymptotic convergence rate for the Sinc (LFII) formula

The mapping z = w2 is a conformal map of the infinite strip domain of width 2d (d > 0), given by \(D_{d}=\{w\in \mathbb {C} | |\Im w|<d\}\) onto the domain Pd in (1.7). Since f(z) is a function analytic in Pd, the function g(w) = 2wf(w2) in (A.1) is analytic in Dd. Then, Lemma 3.6.4 in [10, p.171] gives the asymptotic convergence rate O(heπd/h) (h → 0).

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sugiura, H., Hasegawa, T. Extensions of Clenshaw-Curtis-type rules to integrals over a semi-infinite interval. Numer Algor 90, 3–30 (2022). https://doi.org/10.1007/s11075-021-01177-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11075-021-01177-8

Keywords

Mathematics Subject Classification (2010)

Navigation