def agent_ids(self):
agents = []
for f_id in self.world.facilities.keys():
agents.append(Utils.agentid_producer(f_id))
for f_id in self.world.facilities.keys():
agents.append(Utils.agentid_consumer(f_id))
return agents
def _retailer_profit(self, env, step_outcome):
# 终端（Retailer）营业额
wc = self.env_config['global_reward_weight_consumer']
parent_facility_balance = dict()
# 计算SKU的Reward的时候，将其所属的Store的Reward也计算在内（以一定的权重wc）
for facility in env.world.facilities.values():
if isinstance(facility, ProductUnit):
parent_facility_balance[
facility.id] = step_outcome.facility_step_balance_sheets[
facility.facility.id]
else:
parent_facility_balance[
facility.id] = step_outcome.facility_step_balance_sheets[
facility.id]
consumer_reward_by_facility = {
f_id: wc * parent_facility_balance[f_id] + (1 - wc) * reward
for f_id, reward in
step_outcome.facility_step_balance_sheets.items()
}
rewards_by_agent = {}
for f_id, reward in step_outcome.facility_step_balance_sheets.items():
rewards_by_agent[Utils.agentid_producer(f_id)] = reward / 1000000.0
for f_id, reward in consumer_reward_by_facility.items():
rewards_by_agent[Utils.agentid_consumer(f_id)] = reward / 1000000.0
return rewards_by_agent
def world_to_state(self, world):
state = {}
for facility_id, facility in world.facilities.items():
f_state = self._state(facility)
self._add_global_features(f_state, world)
state[Utils.agentid_producer(facility_id)] = f_state
state[Utils.agentid_consumer(facility_id)] = f_state
return self._serialize_state(state), state