Strange behaviour of find_zero() from Roots.jl

[Tamas_Papp]

Bisection is really simple to implement. Eg

Is there a faster bisection root solver (that uses atol)?

This should be type stable (but didn’t test it thoroughly): using Statistics: middle """ bisection(f, a, b; fa = f(a), fb = f(b), ftol, wtol) Bisection algorithm for finding the root ``f(x) ≈ 0`` within the initial bracket `[a,b]`. Returns a named tuple `(x = x, fx = f(x), isroot = ::Bool, iter = ::Int, ismaxiter = ::Bool)`. Terminates when either 1. `abs(f(x)) < ftol` (`isroot = true`), 2. the width of the bracket is `≤wtol` (`isroot = false`), 3. `maxiter` number of iterations is re…