🌐 Website · 📚 Documentation
Sparse matrices, fill-reducing orderings and iterative solvers in pure Go (CGO=0), on go-ndarray.
sparse provides the COO, CSR, CSC and DOK formats over float64 and
complex128, Matrix Market I/O, the RCM, AMD and COLAMD orderings, and the
iterative solvers CG, GMRES, BiCGSTAB, MINRES, LSQR and LSMR with their
preconditioners, following scipy.sparse.linalg. Its tests replay answers
recorded from SciPy 1.18.1 (and SuiteSparse AMD as a binary oracle): iteration
counts of CG, GMRES, MINRES, LSQR and LSMR equal SciPy's. Sparse direct
factorisations, nested dissection, eigensolvers and algebraic multigrid are
not provided yet; the documentation says so. On one core, SpMV, CG and
GMRES(30) on 2-D Laplacians take 0.56 to 0.98 of SciPy's time.
a := gallery.Laplacian2D(100, 100) // *sparse.CSR[float64]
b := make([]float64, a.Rows)
ic, _ := iter.NewIC0(a) // incomplete Cholesky
res, _ := iter.CG(a, b, &iter.Options{RTol: 1e-8, M: ic})
fmt.Println(res.Info, res.Iterations) // 0 = converged, as in SciPy| Repo | What it is |
|---|---|
| sparse | released v0.1.0: sparse formats and operations, Matrix Market I/O, RCM/AMD/COLAMD orderings, iterative solvers and preconditioners, matrix gallery |
| docs | the documentation site (Hugo + Hextra), served at /docs/ |
| go-sparse.github.io | the landing page (Hugo) |
| brand | logo, icons and social images |
It is part of an ecosystem with go-ndarray (N-dimensional arrays), go-numerics (optimisation, quadrature, ODE and special functions), go-statistics (distributions and tests), go-plot (figures) and go-notebook (Go notebooks).