drl-supply-chain-gym.py

class SimpleSupplyChain(gym.Env):
    def __init__(self, config):
        self.reset()
        self.action_space = Box(low=0.0, high=20.0)            # Continuous space
        self.observation_space = Box(low=-10000, high=10000)

    def reset(self):
        self.supply_chain = SupplyChainEnvironment()
        self.state = self.supply_chain.initial_state()
        return self.state.to_array()

    def step(self, action):
        action_obj = Action(self.supply_chain.warehouse_num)
        action_obj.production_level = action[0]
        action_obj.shippings_to_warehouses = action[1:]
        self.state, reward, done = self.supply_chain.step(self.state, action_obj)
        return self.state.to_array(), reward, done, {}