Add brute force method for finding pure nash equilibria.

[shizejin]

Based on #266.

Add a function pure_nash_brute to find all pure nash equilibria of a normal form game using brute force method.

For example, consider the Prisoners' Dilemma game.

>>> PD_bimatrix = [[(1, 1), (-2, 3)],
...                [(3, -2), (0, 0)]]
>>> g_PD = NormalFormGame(PD_bimatrix)
>>> pure_nash_brute(g_PD)
[(1, 1)]

Mainly referred to Tools for Game Theory. @oyamad

[coveralls]

Coverage increased (+0.04%) to 83.405% when pulling 1d709ab on shizejin:brute_force_pure_nash into 72f9034 on QuantEcon:master.

[oyamad]

I have reviewed this, and it looks good to me.

[jstac]

It's a simple algorithm but a useful addition and very nicely written.

@shizejin Many thanks for your contribution.
@mmcky This looks good to me.

[mmcky]

thanks @shizejin @oyamad @jstac. I will merge this now.