def tabu2genetic(TSnum, _map, _numberOfGuards, _maxTabuMemory, _maxIteration):
# TSnum is th number of TS solutions that will pass into the GA
tabuSolutions = []
for i in range(TSnum):
# initialize a solution
initialSolution = pddl.getSolution(_map, _numberOfGuards)
# pass solution in TS and get an optimal TS path
tabuSolutions.append(
tabuSearch.search(initialSolution, _maxIteration, _maxTabuMemory))
# pass TS solutions to GA
optSolutions = genetic.GA(tabuSolutions)
return optSolutions
def main():
# mapSize = 20
# numberOfGuards = 2
# numberOfObstacles = 2*numberOfGuards
# for mapSize in range(20,100,10):
# for numberOfGuards in range(1,10):
# numberOfObstacles = 2*numberOfGuards
# m = map.Map(numberOfObstacles, 0, -1, mapSize, mapSize, numberOfGuards)
#
# # for i in range(10):
# # m = map.Map(numberOfObstacles, 0, -1, mapSize, mapSize, numberOfGuards)
# # m.printMap()
#
# with open('Number_of_guards='+str(numberOfGuards)+',area_size='+str(mapSize*mapSize)+'.pkl', 'wb') as output:
# pickle.dump(m,output,pickle.HIGHEST_PROTOCOL)
# with open('map_test0.pkl', 'rb') as input:
# with open('map_test1.pkl', 'rb') as input:
# with open('map_test2.pkl', 'rb') as input:
# with open('map_test3.pkl', 'rb') as input:
# with open('map_test4.pkl', 'rb') as input:
# with open('map_test5.pkl', 'rb') as input:
# with open('map_test6.pkl', 'rb') as input:
# with open('map_test7.pkl', 'rb') as input:
# with open('map_test8.pkl', 'rb') as input:
# with open('map_test9.pkl', 'rb') as input:
# m=pickle.load(input)
# m.printMap()
for testcase in testcases:
with open(testcase, 'rb') as input:
m = pickle.load(input)
m.printMap()
help = testcase.split('Number_of_guards=')
numberOfGuards = int(help[1][0])
initialSolution = pddl.getSolution(m, numberOfGuards)
print("\nInitial solution\n")
initialSolution.printMapWithGuardsPath()
# print("\nFinal solution of Tabu Search\n")
bestSolution, performance = genetic.GA(m, numberOfGuards)
bestSolution.printMapWithGuardsPath()
# plt.plot(performance)
fig = plt.figure(performance)
fig.savefig(+str(testcase) + '_performance_plot.png')
def sa2genetic(SAnum, _t0, _tMin, _eMax):
# SAnum is th number of TS solutions that will pass into the GA
SAsolutions = []
for i in range(TSnum):
# initialize a solution
initialSolution = pddl.getSolution(_map, _numberOfGuards)
# pass solution in TS and get an optimal TS path
SAsolutions.append(
simulatedAnnealing.search(_t0,
initialSolution,
_tMin,
_eMax,
deltaT=0.1))
# pass SA solutions to GA
optSolutions = genetic.GA(SAsolutions)
return optSolutions
def main():
if len(sys.argv) != 2:
print('usage: python test_xor.py epochs')
epochs = int(sys.argv[1])
population = init()
ga = genetic.GA(POPULATION_SIZE, MUTATION_RATE, ELITISM)
for epoch in range(epochs):
fitness = [simulate(net) for net in population]
population = ga.step(population, fitness)
print('epoch: {} - best fitness: {}'.format(epoch, max(fitness)))
fitness = [simulate(net) for net in population]
best = np.argmax(fitness)
champion = population[best]
for test in [(True, False), (False, False), (True, True), (False, True)]:
for _ in range(10):
out = champion.step(test)[0]
if out == (test[0] != test[1]):
print("ok!")
else:
print("wrong: {} {} -> {}".format(test[0], test[1], out))
def main():
mapSize = 20
numberOfObstacles = 20
numberOfGuards = 5
maxIteration = 10
maxTabuMemory = 5
temperatureInit = 1000
temperatureMin = 900
eMin = 400
max_gen = 100
m = map.Map(numberOfObstacles, 0, -1, mapSize, mapSize, numberOfGuards)
m.printMap()
path1 = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/Number_of_guards=1,area_size=400.pkl'
path2 = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/Number_of_guards=1,area_size=900.pkl'
path3 = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/Number_of_guards=1,area_size=1600.pkl'
paths = [path1, path2, path3]
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N2'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N3'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N4'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N5'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N6'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N7'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N8'
# path = '/home/ferles/Desktop/Project and Essay/AI_project/ai17_Group20/src/testcases/N9'
for path in paths:
with open(path, 'rb') as input:
file = open(path + '_output.txt', 'w')
m = pickle.load(input)
file.write(m.printMap())
file.write('\n')
# numberOfGuards=int((testcases[0].split('Number_of_guards='))[1][0])
numberOfGuards = 1
initialSolution = pddl.getSolution(m, numberOfGuards)
for i in range(0, numberOfGuards):
file.write("\nInitial solution Guard " + str(i) + "\n")
# print("\nInitial solution Guard " + str(i) + "\n")
file.write(initialSolution.printMapWithOneGuardsPath(i))
file.write('\n')
file.write("\nBegin TabuSearch with memory " + str(maxTabuMemory) +
" and for " + str(maxIteration) + " iterations\n")
# print("\nBegin TabuSearch with memory " + str(maxTabuMemory) + " and for " + str(maxIteration) + " iterations\n")
bestSolution, performance = tabuSearch.search(
initialSolution, maxIteration, maxTabuMemory, file)
plt.plot(performance)
plt.xlabel('Iteration')
plt.ylabel('Fitness')
# plt.savefig(testcases[0]+'_TS_fitEvol.png', bbox_inches='tight')
plt.savefig(path + '_TS_fitEvol.png', bbox_inches='tight')
plt.clf()
# plt.show()
file.write("\nFinal solution of Tabu Search\n")
# print("\nFinal solution of Tabu Search\n")
for i in range(0, numberOfGuards):
# print("\nBest solution Guard " + str(i) + "\n")
file.write("\nBest solution Guard " + str(i) + "\n")
file.write(bestSolution.printMapWithOneGuardsPath(i))
file.write("\nFinal solution of Tabu Search\n")
# print("\nFinal solution of Tabu Search\n")
file.write(bestSolution.printMapWithGuardsPath())
# bestSolution.printMapWithGuardsPath()
file.write("\nEach path of guard\n")
# print("\nEach path of guad\n")
for g in bestSolution.guardsPath:
# print(g)
file.write(str(g) + "\n")
file.write(str(g) + "\n")
file.write(str(g) + "\n")
# print("\nBegin simulatedAnnealing with initial Energy " + str(temperatureInit) + " and for " + "\n")
file.write("\nBegin simulatedAnnealing with initial Energy " +
str(temperatureInit) + " and for " + "\n")
bestSolution, performance = simulatedAnnealing.search(
temperatureInit, initialSolution, temperatureMin, eMin, file)
plt.plot(performance)
plt.xlabel('Iteration')
plt.ylabel('Energy')
# plt.savefig(testcases[0]+'_SA_fitEvol.png', bbox_inches='tight')
plt.savefig(path + '_SA_fitEvol.png', bbox_inches='tight')
plt.clf()
# plt.show()
file.write("\nFinal solution of Simulated Annealing\n")
print("\nFinal solution of Simulated Annealing\n")
for i in range(0, numberOfGuards):
file.write("\nBest solution Guard " + str(i) + "\n")
# print("\nBest solution Guard " + str(i) + "\n")
# bestSolution.printMapWithOneGuardsPath(i)
file.write(bestSolution.printMapWithOneGuardsPath(i))
# print("\nFinal solution of Simulated Annealing\n")
file.write("\nFinal solution of Simulated Annealing\n")
file.write(bestSolution.printMapWithGuardsPath())
# bestSolution.printMapWithGuardsPath()
file.write("\nEach path of guard\n")
# print("\nEach path of guad\n")
for g in bestSolution.guardsPath:
# print(g)
file.write(str(g) + "\n")
file.write("\nBegin genetic with memory with max generations " +
str(max_gen) + "\n")
# print("\nBegin genetic with memory with max generations " + str(max_gen) + "\n")
bestSolution, performance = genetic.GA(m, numberOfGuards, file,
max_gen)
plt.plot(performance)
plt.xlabel('Iteration')
plt.ylabel('Fitness')
plt.savefig(path + '_GA_fitEvol.png', bbox_inches='tight')
plt.clf()
# plt.savefig(testcases[0]+'_GA_fitEvol.png', bbox_inches='tight')
# plt.show()
# print("\nFinal solution of Genetic Algorithm\n")
file.write("\nFinal solution of Genetic Algorithm\n")
for i in range(0, numberOfGuards):
file.write("\nBest solution Guard " + str(i) + "\n")
# print("\nBest solution Guard " + str(i) + "\n")
file.write(bestSolution.printMapWithOneGuardsPath(i))
# bestSolution.printMapWithOneGuardsPath(i)
file.write("\nFinal solution of Genetic Algorithm\n")
# print("\nFinal solution of Genetic Algorithm\n")
file.write(bestSolution.printMapWithGuardsPath())
# bestSolution.printMapWithGuardsPath()
file.write("\nEach path of guard\n")
# print("\nEach path of guard\n")
for g in bestSolution.guardsPath:
# print(g)
file.write(str(g) + "\n")
file.close()
print("ALL SET")