class TradingEnv:
  def _long(self,): # buy
     ...
  def _long_cover(self, current_price_mean, current_mkt_position, action):  # sell possession
    ...
  def step(self, action):
    ...
    # process buy and sell action
    # update position of the agent
    # return next_state and reward
    return self.obs_return, self.chg_reward[0], done, self.info