McCall overwrites the anderson m setting

[jlperla]

Code right now as a

    v_star = fixedpoint(T, v_iv, iterations = iterations, ftol = ftol,
                        m = 0).zero # (5)

which should just be

v_star = fixedpoint(T, v_iv; iterations, ftol , m).zero # (5)

[jlperla]

In this section: https://julia.quantecon.org/dynamic_programming/mccall_model.html#implementation

[jlperla]

@jbrightuniverse Can you take a look at this? Try to make sure it works properly? It should converge significantly faster with anderson vs. just iteration.

[jlperla]

Looks like the default value of m=5 doesn't converge. That can happen with anderson. Setting to m=1 is almost 10x faster though! Fixing in #197