Fast discrete convolution in \(\mathbb{R}^2\) with radial kernels using non-uniform fast Fourier transform with nonequispaced frequencies (Q2287853): Difference between revisions

From MaRDI portal
Changed an Item
ReferenceBot (talk | contribs)
Changed an Item
 
(4 intermediate revisions by 3 users not shown)
Property / describes a project that uses
 
Property / describes a project that uses: Matlab / rank
 
Normal rank
Property / describes a project that uses
 
Property / describes a project that uses: DLMF / rank
 
Normal rank
Property / MaRDI profile type
 
Property / MaRDI profile type: MaRDI publication profile / rank
 
Normal rank
Property / full work available at URL
 
Property / full work available at URL: https://doi.org/10.1007/s11075-019-00670-5 / rank
 
Normal rank
Property / OpenAlex ID
 
Property / OpenAlex ID: W2919600450 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q5558293 / rank
 
Normal rank
Property / cites work
 
Property / cites work: The sparse cardinal sine decomposition and its application for fast numerical convolution / rank
 
Normal rank
Property / cites work
 
Property / cites work: Multidimensional binary search trees used for associative searching / rank
 
Normal rank
Property / cites work
 
Property / cites work: The complexity of finding fixed-radius near neighbors / rank
 
Normal rank
Property / cites work
 
Property / cites work: A fast adaptive multipole algorithm in three dimensions / rank
 
Normal rank
Property / cites work
 
Property / cites work: An Algorithm for the Machine Calculation of Complex Fourier Series / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fixed-radius near neighbors search algorithms for points and segments / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fast Fourier Transforms for Nonequispaced Data / rank
 
Normal rank
Property / cites work
 
Property / cites work: Geometrical Structure of Laplacian Eigenfunctions / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3996584 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Accelerating the Nonuniform Fast Fourier Transform / rank
 
Normal rank
Property / cites work
 
Property / cites work: Using NFFT 3---A Software Library for Various Nonequispaced Fast Fourier Transforms / rank
 
Normal rank
Property / cites work
 
Property / cites work: The type 3 nonuniform FFT and its applications / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q3618967 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fast convolution with radial kernels at nonequispaced knots / rank
 
Normal rank
Property / cites work
 
Property / cites work: Rapid solution of integral equations of scattering theory in two dimensions / rank
 
Normal rank
Property / cites work
 
Property / cites work: Diagonal forms of translation operators for the Helmholtz equation in three dimensions / rank
 
Normal rank
Property / cites work
 
Property / cites work: Fixed-radius near neighbors search / rank
 
Normal rank
Property / cites work
 
Property / cites work: Q4856171 / rank
 
Normal rank
Property / cites work
 
Property / cites work: Scattering theory for the d'Alembert equation in exterior domains / rank
 
Normal rank

Latest revision as of 13:30, 21 July 2024

scientific article
Language Label Description Also known as
English
Fast discrete convolution in \(\mathbb{R}^2\) with radial kernels using non-uniform fast Fourier transform with nonequispaced frequencies
scientific article

    Statements

    Fast discrete convolution in \(\mathbb{R}^2\) with radial kernels using non-uniform fast Fourier transform with nonequispaced frequencies (English)
    0 references
    0 references
    22 January 2020
    0 references
    Let \(z_k\in {\mathbb R}^2\) be given points of the unit disk and \(f_k\in \mathbb C\), \(k=1,\ldots,N\). Further, \(\|x\|\) is the Euclidean norm of \(x\in {\mathbb R}^2\). Let \(g:(0,\infty) \to \mathbb R\) be given. Later \(g(t)\) is chosen as \(\log t\), \(t^2 \log t\), and \(t^{-2}\), \(t>0\). In this paper, the author presents a new algorithm, the so-called \textit{efficient Bessel decomposition} (EBD), for the fast evaluation of discrete convolutions with the radial kernel \(g(\|x\|)\) of the form \[ q_k = \sum_{\ell=1}^N g(\|z_k -z_{\ell}\|)\,f_{\ell}\,, \quad k = 1,\ldots,N\,. \] Using a trigonometric approximation of \(g(\|x\|)\), the discrete convolution can be rapidly computed by non-uniform fast Fourier transforms, see [\textit{D. Potts} et al., Numer. Math. 98, No. 2, 329--351 (2004; Zbl 1056.65146)]. The new EBD method approximates \(g(\|x\|)\) by a Bessel series outside the origin by a small number \(P\) of terms \[ g(\|x\|) \approx g(1) + \sum_{n=1}^P \alpha_n\,J_0(\rho_n \|x\|)\,, \quad \delta \le \|x\|\le 1\,, \] where \(J_0\) is the Bessel function of first kind, \(\rho_n\) is the \(n\)th positive zero of \(J_0\), and \(\delta >0\) is a cutoff parameter. Applying the trapezoidal rule to the integral \[ J_0(\rho_n \|x\|) = \frac{1}{2\pi} \int_{\|y\|=1} {\mathrm e}^{{\mathrm i}\rho_n\,x\cdot y}\,{\mathrm d}y\,, \] the author obtains a trigonometric approximation of \(g(\|x\|)\). An important consequence of EBD is the fact that much less terms last for a given accuracy. This allows for a faster evaluation of the discrete convolution at the cost of longer precomputation. A Matlab code of this method is available online. The computational cost and the error of the EBD method are estimated. Numerical results are presented too.
    0 references
    0 references
    0 references
    0 references
    0 references
    fast discrete convolution
    0 references
    non-uniform fast Fourier transform
    0 references
    efficient Bessel decomposition
    0 references
    radial kernel
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references