def create_data_model():
data = {}
data['depot'] = 0
data['num_vehicles'] = 0
data['demands'] = []
data['distance_matrix'] = []
data['vehicle_capacities'] = []
numVehicles = 4
maxCapacity = 400
capacity, graph, delivery_demand, cityref, datafile = dataset('data_2_23.txt')
#-------------------------
tempdata = pd.read_csv("../helpers/raw_data/" + datafile )
citydata = tempdata.set_index("x")
distanceMat = []
for pointA in graph.keys():
matrix = []
for pointB in graph.keys():
dist = int(get_city_distance( citydata, cityref[pointA], cityref[pointB] ))
matrix.append( dist )
distanceMat.append(matrix)
deliveryDemands = []
for demand in delivery_demand.keys():
deliveryDemands.append( delivery_demand[demand] )
deliveryCapacity= []
for i in range(0, numVehicles):
deliveryCapacity.append( maxCapacity )
data['num_vehicles'] = numVehicles
data['demands'] = get_normalized_demands(deliveryDemands)
data['distance_matrix'] = distanceMat
data['vehicle_capacities'] = deliveryCapacity
return data
from functions import dataset, plot_paths
from sweep import Sweep
import numpy as np
import pandas as pd
import random as rand
import time
if __name__ == '__main__':
"""
capacity : capacity
graph : Graph
delivery_demand : Delivery demands
"""
capacity, graph, delivery_demand, cityref, datafile = dataset(
'data_3_240.txt')
#-------------------------
tempdata = pd.read_csv("../helpers/raw_data/" + datafile)
citydata = tempdata.set_index("x")
for vehicle_capacity in [1000, 1500, 2500, 2000, 2745, 4000, 11000, 15000]:
sweepAlg = Sweep(vehicle_capacity, delivery_demand, cityref, citydata)
start_time = time.time()
sweepAlg.set_graph(graph)
bestSol = sweepAlg.process()
elasped_time = (time.time() - start_time)
if bestSol is not None:
plot_paths(graph, cityref, bestSol, elasped_time, False)
N,M = dataset_X.shape
print(N,M)
# network accuracy
t_idx = [int(line.strip()) for line in open("train_id.txt", 'r')]
te_idx = [int(line.strip()) for line in open("test_id.txt", 'r')]
x_tv = dataset_X[t_idx]
y_tv = dataset_Y[t_idx]
x_test = dataset_X[te_idx]
y_test = dataset_Y[te_idx]
x_train, y_train, x_val, y_val,tr_idx,val_idx = dataset(x_tv, y_tv, test_ratio=0.05)
# print x_test.shape
num_classes = y_test.shape[-1]
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
x_val = x_val.astype('float32')
x_train = x_train.astype('float32').reshape((len(x_train), M,1))
x_test = x_test.astype('float32').reshape((len(x_test),M,1))
x_val = x_val.astype('float32').reshape((len(x_val),M,1))
print(np.sum(y_test,axis=0))
print(np.sum(y_val,axis=0))
print(np.sum(y_train,axis=0))
plt.show()
if __name__ == '__main__':
"""
PARAMS
alpha: relative importance of pheromone
beta: relative importance of heuristic information
sigma:
rho: pheromone coefficient
theta:
num_ants: number of ants
MAX_NFC: max number of function calls
"""
vehicle_capacity, graph, delivery_demand, optimal = dataset('dataset.txt')
alpha = 2
beta = 5
sigma = 3
rho = 0.8
theta = 80
num_ants = 22
MAX_NFC = 1000
vrp = ACOVRP(alpha, beta, sigma, rho, theta, num_ants, vehicle_capacity,
delivery_demand, MAX_NFC)
vrp.set_graph(graph)
bestSol = vrp.process()
if bestSol is not None:
most_vector, max_count, count_vector = most_repeared_promoter(
dataset_X, dataset_Y)
_, p_count, n_count = count_vector
vart_pos = []
for i in range(len(most_vector)):
if most_vector[i] != '0':
vart_pos.append(i)
np.random.seed(42)
tf.set_random_seed(42)
random.seed(42)
# network accuracy
x_train, y_train, x_test, y_test = dataset(dataset_X,
dataset_Y,
test_ratio=0.1)
#print x_train.shape
#print x_test.shape
num_classes = 2
x_train = x_train.astype('float32') #.reshape((len(x_train),8,8,1))
x_test = x_test.astype('float32')
x_train = x_train.astype('float32').reshape((len(x_train), 64, 1))
x_test = x_test.astype('float32').reshape((len(x_test), 64, 1))
cnn = architecture(num_classes)
history = cnn.fit(x_train,
y_train,
batch_size=64,
epochs=50,
import json
import pandas as pd
from functions import dataset
if __name__ == '__main__':
dataset_name = "data_3_96"
capacity, graph, delivery_demand, cityref, indexes, datafile = dataset(
dataset_name + '.txt')
data_arr = {}
for i in graph:
data_arr[i] = {
"index": int(i),
"location": cityref[i],
"lat": float(graph[i][0]),
"lng": float(graph[i][1]),
"demand": float(delivery_demand[i]),
"cityref": int(indexes[i]),
}
jsonData = json.dumps(data_arr, indent=4)
with open("./dataset/" + dataset_name + ".json", "w") as f:
f.write(jsonData)
