def plot_probability_to_all(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
numTrainTour, numTrainRetour):
probabilities = [0.001, 0.005, 0.01, 0.05]
maxRank = [1.05, 1.1, 1.3, 1.7]
replaceNumber = [2, 5, 8, 10]
type = "proportional"
for i in replaceNumber:
plt.figure()
for j in probabilities:
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, j, numTrainTour,
numTrainRetour, maxRank[0], i, type)
plt.plot(x, y, label="probability=" + str(j))
plt.legend()
plt.title("Proportional, replaceNumber=" + str(i))
plt.xlabel("Computation time")
plt.ylabel("Fitness")
plt.savefig("Proportional.replaceNumber" + str(i) + ".pdf")
# plt.show()
type = "rank"
for i in maxRank:
for j in replaceNumber:
plt.figure()
for k in probabilities:
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations,
timeTotal, startTimeOneTour,
startTimeTwoTour, startTimeOneRetour,
startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains,
passByTour, k, numTrainTour,
numTrainRetour, i, j, type)
plt.plot(x, y, label="probability=" + str(k))
plt.legend()
plt.title("Rank, max=" + str(i) + ", replaceNumber=" + str(j))
plt.xlabel("Computation time")
plt.ylabel("Fitness")
plt.savefig("Rank.max" + str(i) + "replaceNumber" + str(j) +
".pdf")
def plot_individuals_analysis(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
numTrainTour, numTrainRetour):
probabilities = 0.01
maxRank = 1.1
replaceNumber = 2
individuals = [5, 10, 20, 50, 100]
type = "rank"
plt.figure()
for i in individuals:
xi = 0
for j in range(5):
meetingTrains, passByTour = ge.initialPopulation(
stations, i, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, i, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank,
int(i * 0.2), type)
xi += y
xi /= 5
plt.plot(x, y, label="n=" + str(i))
plt.legend()
plt.title("Rank, P=0.01, M=1.1, R=0.2")
plt.xlabel("Seconds")
plt.ylabel("Fitness")
plt.savefig("Individuals.pdf")
def plot_max_analysis(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour, startTimeOneRetour,
startTimeTwoRetour, frequencyOne, frequencyTwo,
trainTime, numTrainTour, numTrainRetour):
probabilities = 0.01
maxRank = [1.05, 1.1, 1.2, 1.4, 1.6, 1.8]
replaceNumber = 2
type = "rank"
plt.figure()
for i in maxRank:
xi = 0
for j in range(5):
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, i,
replaceNumber, type)
xi += y
xi /= 5
plt.plot(x, y, label="M=" + str(i))
plt.legend()
plt.title("Rank, P=0.01, R=2, n=10")
plt.xlabel("Seconds")
plt.ylabel("Fitness")
plt.savefig("Max.pdf")
def plot_Prob_analysis(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour, startTimeOneRetour,
startTimeTwoRetour, frequencyOne, frequencyTwo,
trainTime, numTrainTour, numTrainRetour):
probabilities = [0.001, 0.004, 0.007, 0.01, 0.03, 0.05]
maxRank = 1.1
replaceNumber = 2
type = "rank"
plt.figure()
for i in probabilities:
xi = 0
for j in range(5):
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, i, numTrainTour,
numTrainRetour, maxRank, replaceNumber, type)
xi += y
xi /= 5
plt.plot(x, y, label="P=" + str(i))
plt.legend()
plt.title("Rank, M=1.1, R=2, n=10")
plt.xlabel("Seconds")
plt.ylabel("Fitness")
plt.savefig("Probability.pdf")
def runNormal(stations, startTimeOneTour, startTimeOneRetour, startTimeTwoTour,
startTimeTwoRetour, frequencyOne, frequencyTwo, trainTime,
numTrainTour, numTrainRetour, trainTypesTour, trainTypesRetour,
runTime, individuals, probabilities, maxRank, replace, T, k,
timeTotal):
meetingTrains, passByTour = ge.initialPopulation(stations, individuals,
numTrainTour,
numTrainRetour)
x1, y1, meetingTrains1, passByTour1 = ge.convergence2(
runTime, individuals, stations, timeTotal, startTimeOneTour,
startTimeTwoTour, startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank, replace, "rank")
x2, y2, meetingTrains2, passByTour2 = en.run2(
runTime, stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne, frequencyTwo,
trainTime, T, numTrainTour, numTrainRetour, k, 1)
plt.figure()
plt.plot(x1, y1, label="Genetic algorithm")
plt.plot(x2, y2, label="Simulated annealing")
plt.xlabel("Computation time, seconds")
plt.ylabel("Fitness")
plt.title("Convergence plots for normal problem, train run time " +
str(timeTotal))
plt.legend()
plt.show()
tr.runTrains2(stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, passByTour1, numTrainTour,
numTrainRetour, trainTypesTour, meetingTrains1,
trainTypesRetour)
tr.runTrains2(stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, passByTour2, numTrainTour,
numTrainRetour, trainTypesTour, meetingTrains2,
trainTypesRetour)
def plot_prob2(runTime, individuals, stations, timeTotal, startTimeOneTour,
startTimeTwoTour, startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime, numTrainTour,
numTrainRetour):
probabilities = 0.01
maxRank = 1.1
replaceNumber = 2
type = "rank"
plt.figure()
meetingTrains, passByTour = ge.initialPopulation(stations, individuals,
numTrainTour,
numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains, passByTour,
probabilities, numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, label="M=1.1, P=0.01, R=2", c='r')
for i in range(4):
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, c='r')
maxRank = 1.7
type = "rank"
meetingTrains, passByTour = ge.initialPopulation(stations, individuals,
numTrainTour,
numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains, passByTour,
probabilities, numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, label="M=1.7, P=0.01, R=2", c='b')
for i in range(4):
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, c='b')
probabilities = 0.05
maxRank = 1.3
replaceNumber = 5
type = "rank"
meetingTrains, passByTour = ge.initialPopulation(stations, individuals,
numTrainTour,
numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains, passByTour,
probabilities, numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, label="M=1.3, P=0.05, R=5", c='g')
for i in range(4):
meetingTrains, passByTour = ge.initialPopulation(
stations, individuals, numTrainTour, numTrainRetour)
x, y = ge.convergence(runTime, individuals, stations, timeTotal,
startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour,
frequencyOne, frequencyTwo, trainTime,
meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank,
replaceNumber, type)
plt.plot(x, y, c='g')
plt.legend()
#plt.title("Rank, probability=0.1, replaceNumber=" + str(i))
plt.xlabel("Seconds")
plt.ylabel("Fitness")
plt.savefig("Different1")
def runToPerfectOne(runTime, individuals, probabilities, maxRank, replace, T,
k, timeTotal):
stations = 5
startTimeOneTour = 0
startTimeOneRetour = 1
startTimeTwoTour = 400
startTimeTwoRetour = 17
frequencyOne = 26
frequencyTwo = 26
trainTime = [[3, 8, 8, 3], [4, 7, 15, 4]]
numTrainTour, trainTypesTour = tr.calcNumTrain(startTimeOneTour,
startTimeTwoTour,
frequencyOne, frequencyTwo,
timeTotal)
numTrainRetour, trainTypesRetour = tr.calcNumTrain(startTimeOneRetour,
startTimeTwoRetour,
frequencyOne,
frequencyTwo, timeTotal)
#Perform genetic convergence plot
meetingTrains, passByTour = ge.initialPopulation(stations, individuals,
numTrainTour,
numTrainRetour)
x1, y1, meetingTrains1, passByTour1 = ge.convergence2(
runTime, individuals, stations, timeTotal, startTimeOneTour,
startTimeTwoTour, startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, meetingTrains, passByTour, probabilities,
numTrainTour, numTrainRetour, maxRank, replace, "rank")
#Perform simulated annealing convergence plot
x2, y2, meetingTrains2, passByTour2 = en.run2(
runTime, stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne, frequencyTwo,
trainTime, T, numTrainTour, numTrainRetour, k, 1)
#Assumed best solution
passByTour = [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]
meetingTrains = [[1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]
#Fitness of best solution
bestFitness = tr.fitness(stations, timeTotal, startTimeOneTour,
startTimeTwoTour, startTimeOneRetour,
startTimeTwoRetour, frequencyOne, frequencyTwo,
trainTime, passByTour, meetingTrains)
#Plotting
plt.figure()
plt.plot(x1, y1, label="Genetic algorithm")
plt.plot(x2, y2, label="Simulated annealing")
bestList = np.ones(len(x1)) * bestFitness
plt.plot(x1, bestList, label="Probably best solution")
plt.xlabel("Computation time, seconds")
plt.ylabel("Fitness")
plt.title("Convergence plots for constructed problem, train run time " +
str(timeTotal))
plt.legend()
plt.show()
#More plotting
tr.runTrains2(stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, passByTour1, numTrainTour,
numTrainRetour, trainTypesTour, meetingTrains1,
trainTypesRetour)
tr.runTrains2(stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, passByTour2, numTrainTour,
numTrainRetour, trainTypesTour, meetingTrains2,
trainTypesRetour)
tr.runTrains2(stations, timeTotal, startTimeOneTour, startTimeTwoTour,
startTimeOneRetour, startTimeTwoRetour, frequencyOne,
frequencyTwo, trainTime, passByTour, numTrainTour,
numTrainRetour, trainTypesTour, meetingTrains,
trainTypesRetour)