Numerical approximation of vector-valued highly oscillatory integrals
read more
Citations
Clenshaw–Curtis–Filon-type methods for highly oscillatory Bessel transforms and applications
Numerical approximation of highly oscillatory integrals
Fast integration of highly oscillatory integrals with exotic oscillators
Fast, numerically stable computation of oscillatory integrals with stationary points
Numerical approximations to integrals with a highly oscillatory Bessel kernel
References
Handbook of Mathematical Functions
Harmonic Analysis: Real-variable Methods, Orthogonality, and Oscillatory Integrals
Asymptotics and Special Functions
Errata: Milton Abramowitz and Irene A. Stegun, editors, Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, National Bureau of Standards, Applied Mathematics Series, No. 55, U.S. Government Printing Office, Washington, D.C., 1994, and all known reprints
Related Papers (5)
On the Evaluation of Highly Oscillatory Integrals by Analytic Continuation
Efficient quadrature of highly oscillatory integrals using derivatives
Frequently Asked Questions (10)
Q2. What is the common usage of f?
In other words, ∥∥f (m)∥∥ = O(f̃), for m = 0, 1, . . ..The most common usage is f = O(1), which states that f and its derivatives are bounded in [a, b] for increasing ω.
Q3. What is the asymptotic order of the n n matrix?
An n × n matrix A satisfies the right-hand regularity condition for a nonsingular matrix W depending on ω if it can be written as A = P + GW , where G is a nonsingular matrix such that G−1 = O(1) and P is a matrix such that PW−1 = o(1).
Q4. What is the way to compute a Levin-type method?
A Levin-type method retains the asymptotic behaviour of the expansion, while increasing the accuracy of the approximation for fixed frequency.
Q5. What is the order of the Levin-type method?
In [7] it was noted that for the eiωg oscillator, using the functions σk from the asymptotic expansion as a basis caused the order of the resulting Levin-type method to increase with each additional node point.
Q6. What is the basis for the asymptotic behaviour of matrices?
Let A = (aij) and à = (ãij) be two n×m matrices which depend on a real parameter ω, such that the entries of à are always nonnegative.
Q7. What is the simplest way to compute a highly oscillatory integral?
the asymptotic order depended on using Gauss-Laguerre quadrature, which exploits the exponential nature of an oscillator, which will not work for the Airy oscillator case.
Q8. What is the way to compute a highly oscillatory integral?
Using a generalization of an asymptotic expansion, the accuracy of the approximation in fact improves as the frequency of oscillations increases.
Q9. What is the simplest example of a Levin-type method?
To combat this issue, the authors will derive a Levin-type method that has the same asymptotic behaviour as the asymptotic expansion, whilst providing the ability to decrease error further.
Q10. What is the order of the error predicted by the preceding corollary?
The asymptotic order of the error predicted by the preceding corollary is equivalent to that of the asymptotic expansion for both the case where f̃ = 1 and f̃ = (1, 0)>.