Fast implementations of root finding algorithms in Julia that satisfy the SciML common interface.
High Level Examples
using NonlinearSolve, StaticArrays f(u, p) = u .* u .- 2 u0 = @SVector[1.0, 1.0] probN = NonlinearProblem(f, u0) solver = solve(probN, NewtonRaphson(), abstol = 1e-9) ## Bracketing Methods f(u, p) = u .* u .- 2.0 u0 = (1.0, 2.0) # brackets probB = IntervalNonlinearProblem(f, u0) sol = solve(probB, Falsi())
v1.0 Breaking Release Highlights!
v1.0 has been released for NonlinearSolve.jl, making it a decentralized solver library
akin to DifferentialEquations.jl. For simple implementations of nonlinear solvers,
you can now use SimpleNonlinearSolve.jl.
implementations designed for scalar and static vector inputs have all moved to the
lower dependency version. NonlinearSolve.jl is thus designed for the larger scale
more complex implementations, with
NewtonRaphson now sporting support for
LinearSolve.jl and soon SparseDiffTools.jl to allow for preconditioned Newton-Krylov and
exploitation of sparsity. The two pieces will continue to grow in this direction,
with NonlinearSolve.jl gaining more and more wrapped solver libraries and support
for more complex methods, while SimpleNonlinearSolve.jl will keep a lower dependency
version with implementations for small scale problems that do not need all of the
NonlinearProblem was split into
i.e. the bracketing versions now have their own problem definition, rather than using
u0 in a
NonlinearProblem. This helps for finding problem-algorithm
pairing errors at type time and overall improves the documentation / makes the roles