-
An optimal complexity spectral solver for the Poisson equation
Authors:
Ouyuan Qin
Abstract:
We propose a spectral solver for the Poisson equation on a square domain, achieving optimal complexity through the ultraspherical spectral method and the alternating direction implicit (ADI) method. Compared with the state-of-the-art spectral solver for the Poisson equation \cite{for}, our method not only eliminates the need for conversions between Chebyshev and Legendre bases but also is applicab…
▽ More
We propose a spectral solver for the Poisson equation on a square domain, achieving optimal complexity through the ultraspherical spectral method and the alternating direction implicit (ADI) method. Compared with the state-of-the-art spectral solver for the Poisson equation \cite{for}, our method not only eliminates the need for conversions between Chebyshev and Legendre bases but also is applicable to more general boundary conditions while maintaining spectral accuracy. We prove that, for solutions with sufficient smoothness, a fixed number of ADI iterations suffices to meet a specified tolerance, yielding an optimal complexity of $\mathcal{O}(n^2)$. The solver can also be extended to other equations as long as they can be split into two one-dimensional operators with nearly real and disjoint spectra. Numerical experiments demonstrate that our algorithm can resolve solutions with millions of unknowns in under a minute, with significant speedups when leveraging low-rank approximations.
△ Less
Submitted 23 February, 2025;
originally announced February 2025.
-
A new banded Petrov--Galerkin spectral method
Authors:
Ouyuan Qin,
Lu Cheng,
Kuan Xu
Abstract:
We propose a Petrov--Galerkin spectral method for ODEs with variable coefficients. When the variable coefficients are smooth, the new method yields a strictly banded linear system, which can be efficiently constructed and solved in linear complexity. The performance advantage of our method is demonstrated through benchmarking against Mortensen's Galerkin method and the ultraspherical spectral meth…
▽ More
We propose a Petrov--Galerkin spectral method for ODEs with variable coefficients. When the variable coefficients are smooth, the new method yields a strictly banded linear system, which can be efficiently constructed and solved in linear complexity. The performance advantage of our method is demonstrated through benchmarking against Mortensen's Galerkin method and the ultraspherical spectral method. Furthermore, we introduce a systematic approach for designing the recombined basis and establish that our new method serves as a unifying framework that encompasses all existing banded Galerkin spectral methods. This significantly addresses the ongoing challenge of developing recombined bases and sparse Galerkin spectral method. Additionally, the accelerating techniques presented in this paper can also enhance the performance of the ultraspherical spectral method.
△ Less
Submitted 17 February, 2025;
originally announced February 2025.
-
Computing the $p$-Laplacian eigenpairs of signed graphs
Authors:
Chuanyuan Ge,
Ouyuan Qin
Abstract:
As a nonlinear extension of the graph Laplacian, the graph $p$-Laplacian has various applications in many fields. Due to the nonlinearity, it is very difficult to compute the eigenvalues and eigenfunctions of graph $p$-Laplacian. In this paper, we establish the equivalence between the graph $p$-Laplacian eigenproblem and the tensor eigenproblem when $p$ is even. Building on this result, algorithms…
▽ More
As a nonlinear extension of the graph Laplacian, the graph $p$-Laplacian has various applications in many fields. Due to the nonlinearity, it is very difficult to compute the eigenvalues and eigenfunctions of graph $p$-Laplacian. In this paper, we establish the equivalence between the graph $p$-Laplacian eigenproblem and the tensor eigenproblem when $p$ is even. Building on this result, algorithms designed for tensor eigenproblems can be adapted to compute the eigenpairs of the graph $p$-Laplacian. For general $p>1$, we give a fast and convergent algorithm to compute the largest eigenvalue and the corresponding eigenfunction of the signless graph $p$-Laplacian. As an application, we provide a new criterion to determine when a graph is not a subgraph of another one, which outperforms existing criteria based on the linear Laplacian and adjacency matrices. Our work highlights the deep connections and numerous similarities between the spectral theories of tensors and graph $p$-Laplacians.
△ Less
Submitted 14 January, 2025;
originally announced January 2025.
-
Solving nonlinear ODEs with the ultraspherical spectral method
Authors:
Ouyuan Qin,
Kuan Xu
Abstract:
We extend the ultraspherical spectral method to solving nonlinear ODE boundary value problems. We propose to use the inexact Newton-GMRES framework for which an effective preconditioner can be constructed and a fast Jacobian-vector multiplication can be effected, thanks to the structured operators of the ultraspherical spectral method. With a mixed-precision implementation, the inexact Newton-GMRE…
▽ More
We extend the ultraspherical spectral method to solving nonlinear ODE boundary value problems. We propose to use the inexact Newton-GMRES framework for which an effective preconditioner can be constructed and a fast Jacobian-vector multiplication can be effected, thanks to the structured operators of the ultraspherical spectral method. With a mixed-precision implementation, the inexact Newton-GMRES-ultraspherical framework exhibits extraordinary speed and accuracy, as we show by extensive numerical experiments.
△ Less
Submitted 30 June, 2023;
originally announced June 2023.
