Implement the Clenshaw-Curtis quadrature (use a fast cosine transform or the fft) and verify its fast convergence for the integral of f(x) = e in [0, 1]. For what value of n do you get machine precision?