drl-supply-chain-env-sa.py

class State(object):
    def __init__(self, warehouse_num, T, demand_history, t = 0):
        self.warehouse_num = warehouse_num
        self.factory_stock = 0
        self.warehouse_stock = np.repeat(0, warehouse_num)
        self.demand_history = demand_history
        self.T = T   # Length of one episode
        self.t = t

    def to_array(self):
        return np.concatenate( ([self.factory_stock], 
                                self.warehouse_stock, 
                                np.hstack(self.demand_history), 
                                [self.t]) )

    def stock_levels(self):
        return np.concatenate( ([self.factory_stock], self.warehouse_stock) )

class Action(object):
    def __init__(self, warehouse_num):
        self.production_level = 0
        self.shippings_to_warehouses = np.zeros(warehouse_num)