To be released when the accompanying paper is accepted.

This Julia package provides a basic set of fast linear algebra routines (including matrix-vector multiplication, Cholesky factorization, LDL factorization, log-determinants, and linear system solves) for sequentially semiseparable (SSS) matrices. The package documentation includes various examples for efficiently performing Gaussian process regression for many well-known kernels (e.g., Matérn, Celerite, Spline, and stable spline kernels).

BoundaryIntegralEquations.jl

A Julia package for solving boundary integral equations (BIEs) using the (collocation) Boundary Element Method (BEM). This package provides tools for discretizing and solving BIEs. In particular the focus has been on acoustical applications including viscothermal losses. For large-scale simulations approximate approaches such as the Fast Multipole Method (FMM) and hierarchical matrices ($\mathrm{\mathscr{H}︀}$-matrices) have been included.

BlockDiagonalMatrices.jl

A small Julia package for working with block diagonal matrices. As opposed to JuliaArrays/BlockDiagonals.jl, this package supports non-square blocks and fixes the issue for linear scaling for indexing.

SymSemiseparableMatrices.jl

A Julia package for working with symmetric generator representable semiseparable matrices.

FMM2D.jl

A Julia wrapper for the Flatiron Institute’s 2D Fast Multipole Method (FMM) libraries.