A simple procedure for reducing numerical integration errors near singularities
Vincent P. Manno
Abstract
Vincent P. Manno
Abstract
Abstract Numerical integration of functions near singularity points presents a challenge to even advanced quadrature algorithms. Often it would be advantageous to have a simple technique with which to test the accuracy of these estimates. One such technique is described. It consists of subtracting an integrable complementary function from the original integrand, which reduces the curvature of the function to be integrated. This allows simpler numerical integration techniques to be employed for verification. In addition, the technique can be used directly to improve the accuracy of the baseline quadrature. The method provides a means of estimating the order of the functional singularity, which may be unknown.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Abstract Numerical integration of functions near singularity points presents a challenge to even advanced quadrature algorithms. Often it would be advantageous to have a simple technique with which to test the accuracy of these estimates. One such technique is described. It consists of subtracting an integrable complementary function from the original integrand, which reduces the curvature of the function to be integrated. This allows simpler numerical integration techniques to be employed for verification. In addition, the technique can be used directly to improve the accuracy of the baseline quadrature. The method provides a means of estimating the order of the functional singularity, which may be unknown.
Key concepts: Numerical integration, Quadrature (astronomy), Singularity, Gravitational singularity, Simple (philosophy), Curvature, Mathematics, Applied mathematics