An iterative method which usually converges faster than the bisection method. Say that f(a) > 0 > f(b) and we wish to approximate a root of f. The first iterate is x₁ = (af(b)−bf(a))/(f(b)−f(a)); this is the x-value where the line connecting (a,f(a)) and (b,f(b)) crosses the x-axis. Depending on the sign of f(x₁) this process is repeated on the interval [a, x₁] or [x₁,b] to find x₂ and so on.