Work precision diagram for nonlinearproblem

[TorkelE]

Enable WP-diagrams to be created from NonlinearProblems.

Sample code:

# Fetch pakages.
using NonlinearSolve, StaticArrays, DiffEqDevTools, Plots

# Prepares NonlinearProblem.
f(u, p) = u .* u .- p
u0 = @SVector[1.0, 1.0]
p = 2.0
static_prob = NonlinearProblem(f, u0, p)
real_sol = solve(static_prob, NewtonRaphson(), reltol = 1e-15, abstol = 1e-15)

# Sets WP input.
abstols = 1.0 ./ 10.0 .^ (4:12)
reltols = 1.0 ./ 10.0 .^ (4:12)
setups = [Dict(:alg=>NewtonRaphson())
          Dict(:alg=>TrustRegion())]
solnames = ["NewtonRaphson";"TrustRegion"]

# Plots WP-diagram.
wp = WorkPrecisionSet(static_prob, abstols, reltols, setups; names=solnames, numruns=100, appxsol=real_sol, error_estimate=:l2) # Only l2 currently supported.
plot(wp)

The output here is trivial (since all cases solves the problem perfectly, giving all errors = 0.0). However, if I give the wrong input real solution we can get something like:

[TorkelE]

is it currently safe to run using JulaiFormater; format("."); to format repos? I know we currently have some problems with that for Catalyst.

[avik-pal]

Format check is broken on master, so I would hold off doing that here