Preconditioners

Many linear solvers can be accelerated by using what is known as a preconditioner, an approximation to the matrix inverse action which is cheap to evaluate. These can improve the numerical conditioning of the solver process and in turn improve the performance. LinearSolve.jl provides an interface for the definition of preconditioners which works with the wrapped packages.

Using Preconditioners

Mathematical Definition

Preconditioners are specified in the keyword arguments to init or solve: Pl for left and Pr for right preconditioner, respectively. The right preconditioner, $P_r$ transforms the linear system $Au = b$ into the form:

\[AP_r^{-1}(P_r u) = AP_r^{-1}y = b\]

which is solved for $y$, and then $P_r u = y$ is solved for $u$. The left preconditioner, $P_l$, transforms the linear system into the form:

\[P_l^{-1}(Au - b) = 0\]

A two-sided preconditioned system is of the form:

\[P_l A P_r^{-1} (P_r u) = P_l b\]

By default, if no preconditioner is given the preconditioner is assumed to be the identity $I$.

Using Preconditioners

In the following, we will use the DiagonalPreconditioner to define a two-sided preconditioned system which first divides by some random numbers and then multiplies by the same values. This is commonly used in the case where if, instead of random, s is an approximation to the eigenvalues of a system.

using LinearSolve, LinearAlgebra
s = rand(n)
Pl = Diagonal(s)

A = rand(n,n)
b = rand(n)

prob = LinearProblem(A,b)
sol = solve(prob,IterativeSolvers_GMRES(),Pl=Pl)

Preconditioner Interface

To define a new preconditioner you define a Julia type which satisfies the following interface:

  • Base.eltype(::Preconditioner) (Required only for Krylov.jl)
  • LinearAlgebra.ldiv!(::AbstractVector,::Preconditioner,::AbstractVector) and LinearAlgebra.ldiv!(::Preconditioner,::AbstractVector)

Curated List of Pre-Defined Preconditioners

The following preconditioners match the interface of LinearSolve.jl.

  • LinearSolve.ComposePreconditioner(prec1,prec2): composes the preconditioners to apply prec1 before prec2.
  • LinearSolve.InvPreconditioner(prec): inverts mul! and ldiv! in a preconditioner definition as a lazy inverse.
  • LinearAlgera.Diagonal(s::Union{Number,AbstractVector}): the lazy Diagonal matrix type of Base.LinearAlgebra. Used for efficient construction of a diagonal preconditioner.
  • Other Base.LinearAlgera types: all define the full Preconditioner interface.
  • IncompleteLU.ilu: an implementation of the incomplete LU-factorization preconditioner. This requires A as a SparseMatrixCSC.
  • Preconditioners.CholeskyPreconditioner(A, i): An incomplete Cholesky preconditioner with cut-off level i. Requires A as a AbstractMatrix and positive semi-definite.
  • AlgebraicMultiGrid: Implementations of the algebraic multigrid method. Must be converted to a preconditioner via AlgebraicMultiGrid.aspreconditioner(AlgebraicMultiGrid.precmethod(A)). Requires A as a AbstractMatrix. Provides the following methods:
    • AlgebraicMultiGrid.ruge_stuben(A)
    • AlgebraicMultiGrid.smoothed_aggregation(A)
  • PyAMG: Implementations of the algebraic multigrid method. Must be converted to a preconditioner via PyAMG.aspreconditioner(PyAMG.precmethod(A)). Requires A as a AbstractMatrix. Provides the following methods:
    • PyAMG.RugeStubenSolver(A)
    • PyAMG.SmoothedAggregationSolver(A)
  • ILUZero.ILU0Precon(A::SparseMatrixCSC{T,N}, b_type = T): An incomplete LU implementation. Requires A as a SparseMatrixCSC.
  • LimitedLDLFactorizations.lldl: A limited-memory LDLᵀ factorization for symmetric matrices. Requires A as a SparseMatrixCSC. Applying F = lldl(A); F.D .= abs.(F.D) before usage as a preconditioner makes the preconditioner symmetric postive definite and thus is required for Krylov methods which are specialized for symmetric linear systems.
  • RandomizedPreconditioners.NystromPreconditioner A randomized sketching method for positive semidefinite matrices A. Builds a preconditioner $P ≈ A + μ*I$ for the system $(A + μ*I)x = b$