Distributional Reinforcement Learning on Path-dependent Options