def get_coefs(self, X, Y, inds):
from sko.SA import SA_TSP
X_train, X_valid, Y_train, Y_valid = train_test_split(X,
Y,
test_size=0.3)
self.fit(X_train, Y_train, inds, search=False)
def goal(coefs):
self.coefs = coefs.reshape((self.n_estimators, self.n_classes_))
return self.score(X_valid, Y_valid)
sa_tsp = SA_TSP(func=goal,
x0=self.init_coefs.ravel(),
T_max=100,
T_min=1,
L=10)
best_points, _ = sa_tsp.run()
return best_points.reshape((self.n_estimators, self.n_classes_))
模拟退火算法
'''
sa = SA(func=func3,
x0=[1, 1, 1],
T_max=1,
T_min=1e-9,
L=300,
max_stay_counter=150)
best_x, best_y = sa.run()
print('\n模拟退火算法用于多元函数优化\n')
print('best_x:', best_x, 'best_y', best_y)
# plt.plot(pd.DataFrame(sa.best_y_history).cummin(axis=0))
# plt.show()
sa_tsp = SA_TSP(func=getTotalDistance,
x0=range(NUM_POINT),
T_max=100,
T_min=1,
L=10 * NUM_POINT)
best_points, best_distance = sa_tsp.run()
print('\n模拟退火算法解决TSP问题（旅行商问题）\n')
print(best_points, '\nBest Distance:', best_distance)
# from matplotlib.ticker import FormatStrFormatter
# fig, ax = plt.subplots(1, 2)
# best_points_ = np.concatenate([best_points, [best_points[0]]])
# best_points_coordinate = pointsCoordinate[best_points_, :]
# ax[0].plot(sa_tsp.best_y_history)
# ax[0].set_xlabel("Iteration")
# ax[0].set_ylabel("Distance")
# ax[1].plot(best_points_coordinate[:, 0], best_points_coordinate[:, 1],
# marker='o', markerfacecolor='b', color='r', linestyle='-')
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = routine.shape
return sum([
distance_matrix[routine[i % num_points], routine[(i + 1) % num_points]]
for i in range(num_points)
])
# %%
from sko.SA import SA_TSP
sa_tsp = SA_TSP(func=cal_total_distance,
x0=range(num_points),
T_max=100,
T_min=1,
L=10 * num_points)
best_points, best_distance = sa_tsp.run()
print(best_points, best_distance, cal_total_distance(best_points))
# %% Plot the best routine
from matplotlib.ticker import FormatStrFormatter
fig, ax = plt.subplots(1, 2)
best_points_ = np.concatenate([best_points, [best_points[0]]])
best_points_coordinate = points_coordinate[best_points_, :]
ax[0].plot(sa_tsp.best_y_history)
ax[0].set_xlabel("Distance")
ax[0].set_ylabel("Iteration")
distance_matrix = spatial.distance.cdist(points_coordinate, points_coordinate, metric='euclidean')
print('distance_matrix is: \n', distance_matrix)
def cal_total_distance(routine):
num_points, = routine.shape
return sum([distance_matrix[routine[i % num_points], routine[(i + 1) % num_points]] for i in range(num_points)])
# test:
points = np.arange(num_points) # generate index of points
cal_total_distance(points)
# %%
print('-------------')
print('starting SA for TSP problem ...')
from sko.SA import SA_TSP
sa_tsp = SA_TSP(func=cal_total_distance, x0=range(num_points))
best_points, best_distance = sa_tsp.fit()
print(best_points, best_distance, cal_total_distance(best_points))
fig, ax = plt.subplots(1, 1)
best_points_ = np.concatenate([best_points, [best_points[0]]])
best_points_coordinate = points_coordinate[best_points_, :]
ax.plot(best_points_coordinate[:, 0], best_points_coordinate[:, 1], 'o-r')
plt.show()
return totalDis
problem_set = 'kroC100'
tmax = 100
tmin = 1
setting = '%s_SA' % problem_set
start = time.process_time()
nodeMap, nodeDis = readData('tsp_data/%s.tsp' % problem_set)
opt = readOpt('tsp_data/%s.opt.tour' % problem_set)
opt_distance = calculateTotalDis(opt)
opt = np.concatenate([opt, [opt[0]]])
fig, ax = plt.subplots(1, 3)
sa_tsp = SA_TSP(func=calculateTotalDis, x0=range(len(nodeDis)), T_max=tmax, T_min=tmin, L=10 * len(nodeDis))
best_points, best_distance = sa_tsp.run()
best_points_ = np.concatenate([best_points, [best_points[0]]])
ax[0].plot([nodeMap[p][0] for p in best_points_], [nodeMap[p][1] for p in best_points_], 'o-r')
ax[0].set_title('SA solution')
finish = time.process_time()
print('time cost', finish-start)
ax[1].plot(sa_tsp.best_y_history)
ax[1].set_title('Distance Curve')
ax[2].plot([nodeMap[p][0] for p in opt], [nodeMap[p][1] for p in opt], 'o-r')
ax[2].set_title('Optimal solution')
fig.suptitle('Exp. %s' % setting)
plt.savefig('figures/%s_tmax%dtmin%dtime%.1f.jpg' % (setting, tmax, tmin, finish-start))
plt.show()
def cal_total_distance(routine):
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = routine.shape
return sum([distance_matrix[routine[i % num_points], routine[(i + 1) % num_points]] for i in range(num_points)])
from sko.SA import SA_TSP
# L * np.log(Tmin/Tmax, decay_coe)
tsp_obj = SA_TSP(func=cal_total_distance, x0=range(num_points), T_max=280, T_min=1e-3, L=100*num_points)
tsp_obj.cool_down = MethodType(operators.cool_down, tsp_obj)
tsp_obj.get_new_x = MethodType(operators.get_new_x, tsp_obj)
best_points, best_distance = tsp_obj.run()
X = data_cities[best_points, 1]
Y = data_cities[best_points, 2]
plt.figure(figsize=(12, 8))
# plt.title('TSP Graph')
plt.title('GA: ' + str(best_distance))
plt.xlabel('X')
plt.ylabel('Y')
plt.scatter(X, Y, s=140)
def load_data(f1=1, f2=1, f3=1, f4=1, f5=1, f6=1, f7=1):
file_name = 'location_data.csv'
data = np.loadtxt(file_name, delimiter=',')
transport_data = []
if f1 == 1:
transport_data.append(data[0])
if f2 == 1:
transport_data.append(data[1])
if f3 == 1:
transport_data.append(data[2])
if f4 == 1:
transport_data.append(data[3])
if f5 == 1:
transport_data.append(data[4])
if f6 == 1:
transport_data.append(data[5])
if f7 == 1:
transport_data.append(data[6])
points_coordinate = np.zeros((len(transport_data), 2))
volume = np.zeros((len(transport_data)))
solids = np.zeros((len(transport_data)))
cattle = np.zeros((len(transport_data)))
pigs = np.zeros((len(transport_data)))
chicken = np.zeros((len(transport_data)))
truck_vol = 18 #truck has capacity of 18m3
for n in range(0, len(transport_data)):
points_coordinate[n][0:2] = transport_data[n][0:2]
volume[n] = transport_data[n][2]
solids[n] = transport_data[n][3]
cattle[n] = transport_data[n][4]
pigs[n] = transport_data[n][5]
chicken[n] = transport_data[n][6]
max_vol = np.argmax(volume, axis=0)
digestor_loc = [points_coordinate[max_vol]]
num_points = points_coordinate.shape[0]
distance_matrix = spatial.distance.cdist(points_coordinate,
points_coordinate,
metric='euclidean')
distance_matrix = distance_matrix * 111 # 1 degree of lat/lon ~ = 111km
print(distance_matrix)
distance_home = spatial.distance.cdist(
points_coordinate, digestor_loc, metric='euclidean') * 111
def cal_total_distance(routine):
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = routine.shape
trip_vol = 0
dist = 0
for i in range(num_points):
trip_vol = trip_vol + volume[routine[i % num_points]]
trips = 0
dist_home = 0
if trip_vol > truck_vol:
trips = trip_vol // truck_vol
trip_vol = trip_vol % truck_vol
dist_home = dist_home + int(
distance_home[routine[i % num_points]])
dist += distance_matrix[routine[i % num_points], routine[
(i + 1) % num_points]] + 2 * dist_home
return dist
sa_tsp = SA_TSP(func=cal_total_distance,
x0=range(num_points),
T_max=100,
T_min=1,
L=10 * num_points)
best_points, best_distance = sa_tsp.run()
def best_points_route(best_points, start):
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = best_points.shape
new_route = [start]
trip_vol = 0
dist = 0
for i in range(num_points):
trip_vol = trip_vol + volume[best_points[i % num_points]]
trips = 0
dist_home = 0
new_route.append(best_points[i])
if best_points[i % num_points] == max_vol:
trips = 0
trip_vol = 0
if (trip_vol > truck_vol) & (best_points[i % num_points] !=
max_vol):
trips = trip_vol // truck_vol
trip_vol = trip_vol % truck_vol
dist_home = dist_home + int(
distance_home[best_points[i % num_points]])
new_route.append(max_vol)
new_route.append(best_points[i % num_points])
trips = 0
dist = dist + distance_matrix[
best_points[i % num_points],
best_points[(i + 1) % num_points]] + 2 * dist_home
return new_route
final_best = best_points_route(best_points, max_vol)
print(type(final_best))
#Total volumes (m3) are the daily volumes of all the farms per day
total_volume = total_vol(volume)
total_solids_perc = vol_breakdown(volume, solids)
total_cattle_perc = vol_breakdown(volume, cattle)
total_pig_perc = vol_breakdown(volume, pigs)
total_chicken_perc = vol_breakdown(volume, chicken)
manure_comp = [total_cattle_perc, total_pig_perc, total_chicken_perc]
from matplotlib.ticker import FormatStrFormatter
final_best.append(max_vol)
best_points_ = np.array(final_best)
best_points_coordinate = points_coordinate[best_points_, :]
print("The best route is: " + str(final_best) +
" and the distance on this route is " + str(best_distance))
print("Optimal location is area # " + str(max_vol) +
" in radians for DIGESTOR is latitude: " + str(digestor_loc[0][0]) +
" and longitude: " + str(digestor_loc[0][1]))
print("Total daily distance from farms to digestor travelled is " +
str(best_distance) + " km")
print("Total VOLUME manure supplied per day is " + str(total_volume) +
" m3")
print("Weighted average solids percentage of the manure supplied is " +
str(total_solids_perc * 100) + " %")
print("Manure composition is CATTLE-PIGS-CHICKS is " + str(manure_comp))
fig, ax = plt.subplots(1, 2)
ax[0].plot(sa_tsp.best_y_history)
ax[0].set_xlabel("Iteration")
ax[0].set_ylabel("Distance")
ax[1].plot(best_points_coordinate[:, 0], best_points_coordinate[:, 1],\
marker='o', markerfacecolor='b', color='c', linestyle='-')
ax[1].xaxis.set_major_formatter(FormatStrFormatter('%.3f'))
ax[1].yaxis.set_major_formatter(FormatStrFormatter('%.3f'))
ax[1].set_xlabel("Longitude")
ax[1].set_ylabel("Latitude")
plt.show()
return [best_distance, total_volume, total_solids_perc, manure_comp]
def load_data(f1=1,f2=1,f3=1,f4=1,f5=1,f6=1,f7=1,dict_total=dict_total, printt=False):
#file_name = 'location_data.csv'
#data = np.loadtxt(file_name, delimiter=',')
Farm_data = dict_total["Farm_data"]
transport_data = []
#Farm_name = [Longitude, Latitudem Volume_per_day, Solid_percentage, cattle_percentage, pig_percentage, poultry_percentrage]
if f1==1:
transport_data.append(np.array(Farm_data["Farm_1"]))
if f2==1:
transport_data.append(np.array(Farm_data["Farm_2"]))
if f3==1:
transport_data.append(np.array(Farm_data["Farm_3"]))
if f4==1:
transport_data.append(np.array(Farm_data["Farm_4"]))
if f5==1:
transport_data.append(np.array(Farm_data["Farm_5"]))
if f6==1:
transport_data.append(np.array(Farm_data["Farm_6"]))
if f7==1:
transport_data.append(np.array(Farm_data["Farm_7"]))
points_coordinate = np.zeros((len(transport_data),2))
volume = np.zeros((len(transport_data)))
solids = np.zeros((len(transport_data)))
cattle = np.zeros((len(transport_data)))
pigs = np.zeros((len(transport_data)))
chicken = np.zeros((len(transport_data)))
truck_vol = dict_total['V_per_truck'] #truck has capacity of 18m3
for n in range(0,len(transport_data)):
points_coordinate[n][0:2]=transport_data[n][0:2]
volume[n] = transport_data[n][2]
solids[n] = transport_data[n][3]
cattle[n] = transport_data[n][4]
pigs[n] = transport_data[n][5]
chicken[n] = transport_data[n][6]
max_vol = np.argmax(volume, axis=0)
digestor_loc, farm_digestor = optimal_loc(points_coordinate,volume, max_vol)
num_points = points_coordinate.shape[0]
distance_matrix = spatial.distance.cdist(points_coordinate, points_coordinate, metric='euclidean')
distance_matrix = distance_matrix * 111 # 1 degree of lat/lon ~ = 111km
#print(distance_matrix)
distance_home = spatial.distance.cdist(points_coordinate, digestor_loc, metric='euclidean')*111
def cal_total_distance(routine):
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = routine.shape
trip_vol = 0
dist = 0
for i in range(num_points):
trip_vol = trip_vol + volume[routine[i % num_points]]
trips = 0
dist_home = 0
if trip_vol>truck_vol:
trips = trip_vol // truck_vol
trip_vol = trip_vol % truck_vol
dist_home = dist_home + int(distance_home[routine[i % num_points]])
dist += distance_matrix[routine[i % num_points], routine[(i + 1) % num_points]] + 2*dist_home
return dist
if f1+f2+f3+f4+f5+f6+f7 >2:
sa_tsp = SA_TSP(func=cal_total_distance, x0=range(num_points), T_max=100, T_min=1, L=10 * num_points)
best_points, best_distance = sa_tsp.run()
elif f1+f2+f3+f4+f5+f6+f7 ==2:
if volume[1]>volume[0]:
best_points = np.array([1,0])
else:
best_points = np.array([0,1])
best_distance = cal_total_distance(best_points)
elif f1+f2+f3+f4+f5+f6+f7 ==1:
best_points = np.array([0])
best_distance = 0
def best_points_route(best_points, start):
'''The objective function. input routine, return total distance.
cal_total_distance(np.arange(num_points))
'''
num_points, = best_points.shape
new_route = [start]
trip_vol = 0
dist = 0
for i in range(num_points):
trip_vol = trip_vol + volume[best_points[i % num_points]]
trips = 0
dist_home = 0
new_route.append(best_points[i])
if best_points[i % num_points]==farm_digestor:
trips = 0
trip_vol = 0
if (trip_vol>truck_vol) & (best_points[i % num_points]!=farm_digestor):
trips = trip_vol // truck_vol
trip_vol = trip_vol % truck_vol
dist_home = dist_home + int(distance_home[best_points[i % num_points]])
new_route.append(farm_digestor)
new_route.append(best_points[i % num_points])
trips = 0
dist = dist + distance_matrix[best_points[i % num_points], best_points[(i + 1) % num_points]] + 2*dist_home
return new_route
#print("BEST POINTS")
best_points = np.delete(best_points,np.where(best_points==farm_digestor))
#print(best_points)
#print(type(best_points))
final_best = best_points_route(best_points, farm_digestor)
#Total volumes (m3) are the daily volumes of all the farms per day
total_volume = total_vol(volume)
total_solids_perc = vol_breakdown(volume, solids)
total_cattle_perc = vol_breakdown(volume, cattle)
total_pig_perc = vol_breakdown(volume, pigs)
total_chicken_perc = vol_breakdown(volume, chicken)
manure_comp = [total_cattle_perc, total_pig_perc, total_chicken_perc]
from matplotlib.ticker import FormatStrFormatter
if (final_best[-1])!=farm_digestor:
final_best.append(farm_digestor)
best_points_ = np.array(final_best)
best_points_coordinate = points_coordinate[best_points_, :]
#print(volume)
if printt:
print("The best route is: "+str(final_best)+" and the distance on this route is "+str(best_distance))
print("The route is as follow:")
print(best_points_coordinate)
print("Optimal location is area # "+str(farm_digestor)+" in radians for DIGESTOR is latitude: "+str(digestor_loc[0][0])+" and longitude: "+str(digestor_loc[0][1]))
print("Total daily distance from farms to digestor travelled is "+str(best_distance)+" km")
print("Total VOLUME manure supplied per day is "+str(total_volume)+" m3")
print("Weighted average solids percentage of the manure supplied is "+str(total_solids_perc*100)+" %")
print("Manure composition is CATTLE-PIGS-CHICKS is "+str(manure_comp))
#fig, ax = plt.subplots(1, 2)
#ax[0].plot(sa_tsp.best_y_history)
#ax[0].set_xlabel("Iteration")
#ax[0].set_ylabel("Distance")
#ax[1].plot(best_points_coordinate[:, 0], best_points_coordinate[:, 1],
# marker='o', markerfacecolor='b', color='c', linestyle='-')
#ax[1].xaxis.set_major_formatter(FormatStrFormatter('%.3f'))
#ax[1].yaxis.set_major_formatter(FormatStrFormatter('%.3f'))
#ax[1].set_xlabel("Longitude")
#ax[1].set_ylabel("Latitude")
#plt.show()
return [best_distance, total_volume, total_solids_perc, manure_comp,final_best]