calc_reward passes in last_state and previous_state with the same current hp values

[thekevalian]

I noticed while writing my reward function that the last state and current state passed into calc_reward always gave me the same value for the current_hp, even though there was a difference. Changing the shallow copy in openai_api.py from

battle = copy.copy(self.current_battle)

to a deep copy

battle = copy.deepcopy(self.current_battle)

fixed the issue.

I'm not sure if it's only me experiencing this, but I've attached a screenshot of the edit just in case anyone else has experienced this:

[hsahovic]

Hi @thekevalian,
Thanks for reporting this. I'll investigate.