def graphs():
for tech in range(0, 3):
for xpar in [0.01]:
par = SimulationParameter(Nx=50, Ny=10, Nz=50,
Lx=1, Ly=1.5, Lz=1.2, Cy=0,
Ax=0.8, Vx=4, Ay=0.8, Vy=0.9, Az=0.7, Vz=1.20,
Wli=1850, Wsh=850, Wsa=350,
Cr=0.02, Fr=1.15, rendiment=0.9,
Nli=3, Nsh=3, Nsa=2,
bay_level=0,
tech=tech, strat=1, strat_par_x=xpar, strat_par_y=1)
t_par = TraceParameter(sim_time=3600 * 6, types={"Type" + str(i): (i / 10) for i in np.repeat(1, 10)},
int_mean=25,
start_fullness=0.5,
seed=[1])
c_par = Monitor.CostParam(intended_time=3.154e+8, lift=20000, shuttle=50000, shuttle_fork=35000,
satellite=35000, transelevator=40000,
scaffolding=30, energy_cost=0.0356 / 1000)
res = Test.test(parameter=par, trace_parameter=t_par, cost_par=c_par, log=True)
d = par.__dict__
t = copy.copy(t_par)
t.seed = t_par.seed[0]
d.update(t.__dict__)
d.update(res.__dict__)
with open("../data/strategy_par_x.csv", mode="a+", newline='') as f:
w = csv.DictWriter(f,
fieldnames=(list(d.keys())))
if os.stat(f.name).st_size < 10:
w.writeheader()
w.writerow(d)
def get_average_results(self):
ret = Monitor.Results()
for p in ret.__dict__:
ret.__dict__[p] = 0
try:
results = self.res.get()
except Exception as err:
return np.inf
i = 0
for res in results:
if not self.saved:
d = self.opt_par.map(self).__dict__
t = copy.copy(self.t_par)
t.seed = self.t_par.seed[i]
i += 1
d.update(t.__dict__)
d.update(res.__dict__)
w = csv.DictWriter(open("results_old.csv", mode="a+"),
fieldnames=(list(d.keys())))
if os.stat("results_old.csv").st_size < 10:
w.writeheader()
w.writerow(d)
for p in res.__dict__:
if isinstance(res.__dict__[p], int) or isinstance(
res.__dict__[p], float):
ret.__dict__[p] += (res.__dict__[p]) / len(results)
else:
ret.__dict__[p] = 999999999
self.saved = True
return ret
def test():
area = 8 * 100
nz = 200
par = SimulationParameter(Nx=15, Ny=5, Nz=200,
Lx=5, Ly=5, Lz=5, Cy=0,
Ax=0.8, Vx=4, Ay=0.8, Vy=0.9, Az=0.7, Vz=1.20,
Wli=1850, Wsh=850, Wsa=350,
Cr=0.02, Fr=1.15, rendiment=0.9,
Nli=34, Nsh=4, Nsa=4,
bay_level=0,
tech=0, strat=1, strat_par_x=1, strat_par_y=1)
t_par = TraceParameter(sim_time=360 * 3, types={"Type" + str(i): (1 / 10) for i in range(0, 10)}, int_mean=25,
start_fullness=0.5,
seed=1023)
c_par = Monitor.CostParam(intended_time=3.154e+8, lift=20000, shuttle=50000, shuttle_fork=35000,
satellite=35000, transelevator=40000,
scaffolding=30, energy_cost=0.0356 / 1000)
res = Test.test(parameter=par, trace_parameter=t_par, cost_par=c_par, log=True)
print(res)
def only_plot():
if __name__ == '__main__':
with open("../resultsLowNz.json", "r") as f:
result = jsonpickle.loads(f.read())
c = "rgbyk"
for p in Monitor.Results().__dict__:
for i in range(0, len(result)):
lists = sorted(result[i].items(), key=lambda x: int(x[0]))
x, y = zip(*lists) # unpack a l
# ist of pairs into two tuples
y = [a.__dict__[p] for a in y]
plt.plot(x, y, c[i], label="tech" + str(i))
plt.ylabel(p)
plt.xlabel('Nz')
plt.legend()
plt.savefig("../output/lowNz/" + str(p) + ".png", bbox_inches='tight')
plt.close()
def test(parameter, trace_parameter, cost_par=None, log=False):
class Status:
def __init__(self, parameter, monitor):
self.parameter = parameter
self.monitor = monitor
# todo temporari
# parameter.Nx = round(4000 / (parameter.Nz * parameter.Ny))
# print("\n\n")
# print(parameter.Nx)
# print(parameter.Ny)
# print(parameter.Nz)
# print("\n\n")
sim = Simulation(Status(parameter, None))
sim.__status__.monitor = Monitor(sim)
sim.logger.enable(log)
# env = simpy.RealtimeEnvironment(0, 0.1, False)
# warehouse = Warehouse(env, parameter, trace_load("trace1.json"))
warehouse = Warehouse(sim, parameter, trace_parameter)
sim.run(until=trace_parameter.sim_time)
ret = sim.get_status().monitor.get_result(cost_param=cost_par)
return ret
def adjacency():
d = {k.strip(): {row['antecedents'].strip(): row["lift"] for i, row in v.iterrows()} for k, v in
pd.read_csv("ad_rule.csv").groupby(by="consequents")}
gen = pd.read_csv("Original_dataset.csv", encoding='unicode_escape').dropna()
gen["Description"] = gen["Description"].map(lambda x: x.strip())
gen = gen.groupby("Description",
as_index=False).count().rename(
{"Country": "count"}, axis=1)[["Description", "count"]]
gen = {row["Description"].strip(): row["count"] / gen["count"].sum() for index, row in gen.iterrows()}
par = SimulationParameter(Nx=100, Ny=20, Nz=100,
Lx=5, Ly=5, Lz=5, Cy=0,
Ax=0.8, Vx=4, Ay=0.8, Vy=0.9, Az=0.7, Vz=1.20,
Wli=1850, Wsh=850, Wsa=350,
Cr=0.02, Fr=1.15, rendiment=0.9,
Nli=8, Nsh=4, Nsa=4,
bay_level=0,
tech=2, strat=2, strat_par_x=1, strat_par_y=100, adjacency=d)
t_par = TraceParameter(sim_time=36000, types=gen, int_mean=402, start_fullness=0.7,
seed=1023)
c_par = Monitor.CostParam(intended_time=3.154e+8, lift=20000, shuttle=50000, shuttle_fork=35000,
satellite=35000, transelevator=40000,
scaffolding=30, energy_cost=0.0356 / 1000)
res = Test.test(parameter=par, trace_parameter=t_par, cost_par=c_par, log=True)
print(res)
def test():
for tech in [0, 1, 2]:
with open("cost_vs_energy_tech{tech}.csv".format(tech=tech),
"w") as f:
f.write("par,cost,energy\n")
for cost_f in np.arange(0.7, 1.05, 0.05):
print(str(tech) + " " + str(cost_f))
par = OptParameter(
Nx=OptRange(10, 50),
Ny=OptRange(3, 20),
Nz=OptRange(10, 50),
Lx=1,
Ly=1.5,
Lz=1.2,
Cy=0,
Ax=0.8,
Vx=4,
Ay=0.8,
Vy=0.9,
Az=0.7,
Vz=1.20,
# Ax=0.2, Vx=4, Ay=0.2, Vy=0.9, Az=0.2, Vz=1.20,
Wli=1850,
Wsh=850,
Wsa=350,
Cr=0.02,
Fr=1.15,
rendiment=0.9,
Nli=OptRange(1, 5),
Nsh=OptRange(1, 3),
Nsa=OptRange(1, 3),
bay_level=0,
tech=tech,
strat=1,
strat_par_x=OptRange(0, 10, decimal=True),
strat_par_y=OptRange(0, 1))
t_par = TraceParameter(
sim_time=1000,
types={"Type" + str(i): (1 / 10)
for i in range(0, 10)},
int_mean=25,
start_fullness=0.5,
seed=[1])
f_par = FitnessParameter(completeness=-1000000000,
cost=cost_f,
energy_consumed=(1 - cost_f) / 1000)
c_par = Monitor.CostParam(intended_time=3.154e+8,
lift=20000,
shuttle=50000,
shuttle_fork=35000,
satellite=35000,
transelevator=40000,
scaffolding=30,
energy_cost=0.0356 / 1000)
r1, r2 = Opt.optimization(par, t_par, f_par, c_par, 10000, 2)
with open(
"casi_studio/ost_vs_energy_tech{tech}.csv".format(
tech=tech), "a+") as f:
# f.write(
# "tech " + str(tech) + " cost_f " + str(cost_f) + ":\n" + str(r1.__dict__) + "\n\n\n\n" + str(
# r2) + "\n\n\n\n\n")
f.write("{par},{cost},{energy}\n".format(
par=cost_f, cost=r2.cost, energy=r2.energy_consumed))