I want to solve (A - c_i) * x = b for multiple shifts c_i. A is a Hermitian matrix, where A*x is efficient than obtaining A directly. A - c_i * I is positive definite. I know that one can use Krylov subspace methods (Lanczos iteration) to solve this problem for multiple shifts efficiently. Is there…

@Jae-Mo_Lihm I just released the vesion 0.8.0 of Krylov.jl. All solvers are able to solve complex linear systems now !!! Don’t hesitate to try this new feature :slight_smile:

Linear problem with multiple shifts

Specific Domains Numerics

rafael.guerra October 4, 2021, 5:11am 4

Have you also checked KrylovKit.jl?
Please read here.

1 Like

Topic		Replies	Views
Solving shifted linear system Numerics linearalgebra	19	669	April 15, 2024
Which Krylov package solves A*x=b where x and b can be any "vector-like" object? New to Julia	1	270	December 28, 2023
[ANN] KrylovKit: Community announcement	8	1679	September 7, 2018
How to solve this Ax=b faster? Numerics linearalgebra , linearsolve	24	2433	June 7, 2022
Lanczos tridiagonalization algorithm in Julia? Numerics question , package , linearalgebra , algorithm	19	3077	March 10, 2022

Linear problem with multiple shifts

Related topics