Back to Eigenvalue Solvers (ARPACK)
ar.core.eigsh
Compute eigenvalues and eigenvectors of a sparse matrix.
Syntax
ar.core.eigsh(matvec, n, nev, options?)
Description
Compute eigenvalues and eigenvectors of a sparse matrix. This is the unified API that dispatches to the appropriate solver: - 'symmetric' (default): Uses Lanczos method (more efficient for symmetric matrices) - 'nonsymmetric': Uses Arnoldi method (handles non-symmetric matrices, complex eigenvalues)
Parameters
| Name | Description |
|---|---|
| matvec | - Function computing y = A*x |
| n | - Dimension of the matrix (A is n×n) |
| nev | - Number of eigenvalues to compute |
| options(optional) | - Solver options including problem type |
Returns
Promise<EigsResult | EignResult>