def mainPane(self):
locations = ["Tavern", "Shop", "Forest", "Inn", "Castle"]
character_approval = False
while character_approval != True:
hero = Hero(Weapon().generate_weapon())
print("Your character is " + hero.char.name +
"\nEquipped with a " + hero.get_weapon().name +
" that does " + str(hero.get_weapon().low_damage) + "-" +
str(hero.get_weapon().high_damage) + " damage.")
char_generation = input("Do you approve of this character: ")
if "Y" in char_generation.upper():
character_approval = True
while True:
print("\nLocations: ")
counter = 1
for i in locations:
print(str(counter) + ". " + i)
counter += 1
location_choice = input("Where would you like to go: ")
if int(location_choice) is 1:
Tavern(hero)
if int(location_choice) is 2:
Shop(hero)
if int(location_choice) is 3:
forest = Forest()
forest.entry(hero)
if int(location_choice) is 4:
hotel = Inn(hero)
if int(location_choice) is 5:
Castle(hero)
input()
def test_getNeighborEdgeCase(self):
location = (0, 0)
forest = Forest(self.size)
neighbor_locations = forest.getNeighborLocations(location)
self.assertEqual(len(neighbor_locations), 2)
def getFather(self, tree):
if (self.isEmpty()):
return
else:
f = Forest([self])
return f.getFatherOfTree(tree)
def getDegree(self):
if self.isEmpty():
return 0
else:
# transform the rooted tree in forest
f = Forest([self])
# call the coresponding method on the forest
return f.getDegree()
def getAscending(self, root, tree):
if self.isEmpty():
return []
else:
# transform the rooted tree in forest
f = Forest([self])
# call the coresponding method on the forest
return f.getAscending(root, tree)
def test_isInBounds(self):
forest = Forest(self.size)
location_inbounds = (1, 3)
self.assertTrue(forest.isInBounds(location_inbounds))
location_outofbounds = (self.size, self.size + 1)
self.assertFalse(forest.isInBounds(location_outofbounds))
def displayWidth(self):
# do nothing if the rooted tree is empty
if self.isEmpty():
return
else:
# transform the rooted tree in forest
f = Forest([self])
# call the coresponding method on the forest
return f.displayWidth()
def newForest(self):
forest = Forest()
for i in xrange(0, self.numTrees):
position = Point(trand(self.xrange), trand(self.yrange))
tree = Tree(position=position, value = trand(self.valuerange),
length = trand(self.lengthrange))
tree.id = i
forest.addTree(tree)
return forest
def test_elementManipulation(self):
location = (1, 3)
value = 5
forest = Forest(self.size)
forest.setValue(location, value)
get_value = forest.getValue(location)
self.assertEqual(get_value, value)
def __init__(self, player):
self.array = []
self.player = player
self.c = 0
self.towns = [Town("Farnfos", [0, 0]), Town("Lunari", [1, 1]),
Town("Cewmann", [2, 2]),
Town("Caister", [3, 3]), Town("Aramore", [4, 4]),
Town("Kilerth", [5, 5]),
Town("Eldhams", [6, 6]), Town("Cirrane", [7, 7]),
Town("Warcest", [8, 8])]
self.forest = Forest()
def __init__(self, facilities, customers, params):
self.facilities = dict(
zip([facility.index for facility in facilities],
[facility for facility in facilities]))
self.customers = customers
self.subproblemSolutionForest = Forest()
self.currentIteration = 0
self.mip = MIP(facilities, customers,
"Instance_%s_%s" % (len(facilities), len(customers)),
params["mipSolver"])
self.facilitiesCount = len(facilities)
self.quantiles = []
self.params = params
self.totalDemand = sum([customer.demand for customer in customers])
self.currentObjectiveFunction = 0
self.currentSolutionAssignment = []
def init_UI(self):
self.FAQ.clicked.connect(self.show_faq)
self.Generate_button.clicked.connect(self.generate_image_main_logic)
self.Save_button.setEnabled(False)
self.Save_button.clicked.connect(self.save_image)
# base_settings_update connections
self.base_settings = {'size': 200, 'zoom': 50}
self.Size_ComboBox.valueChanged.connect(self.base_settings_update)
self.Zoom_ComboBox.valueChanged.connect(self.base_settings_update)
# Biomes
self.Forest = Forest(self.ForestLayout)
self.Desert = Desert(self.DesertLayout)
self.Frost = Frost(self.FrostLayout)
def forestTest(trainSet,
trainLab,
testSet,
testLab,
ntree,
nbayes,
pruneTreeSplit=0.0):
forest = Forest(ntree, nbayes, pruneSplit=pruneTreeSplit)
forest.fit(trainSet, trainLab)
predLab = forest.predict(testSet)
print("forest (trees: {}, bayes:{}) errors:".format(ntree, nbayes))
print(sum(testLab != predLab))
print(1 - sum(testLab != predLab) / len(testLab))
print("forest (trees: {}, bayes:{}) confusion matrix:".format(
ntree, nbayes))
print(confusion_matrix(testLab, predLab))
print()
def main():
# Init the forest
forest = Forest()
win = forest.getWindow()
# Draw an input box
e = Entry(Point(150, 75), 10)
e.setFill("white")
e.draw(win)
# Draw the text above the box
text1 = Text(Point(150, 86), "Probability of Spreading")
text = Text(Point(150, 81), "Enter 0-1 (Ex. 0.25)")
text1.draw(win)
text.draw(win)
# Draw two buttons
createButton(0, win)
createButton(1, win)
# Finally, run the simulation!
closeProgram = False
while not closeProgram:
# Get the selected tree and burn probability
tree, burnPercent = getTreeAndPercent(win, forest, e)
# Begin the fire!
forest.beginBurn(tree, burnPercent)
forest.displayPercentBurned()
# Get a click and see if they want to close the program
# or if they want to reset and go again
clickpt = win.getMouse()
while not(clickpt.getX() <= 170 and clickpt.getX() >= 130 and \
((clickpt.getY() <= 55 and clickpt.getY() >= 45) or \
(clickpt.getY() <= 35 and clickpt.getY() >= 25))):
# Get another click
clickpt = win.getMouse()
if clickpt.getX() <= 170 and clickpt.getX() >= 130 and \
clickpt.getY() <= 55 and clickpt.getY() >= 45:
# Reset the forest
forest.resetForest()
elif clickpt.getX() <= 170 and clickpt.getX() >= 130 and \
clickpt.getY() <= 35 and clickpt.getY() >= 25:
# Close the window
closeProgram = True
def test_addTree(self):
forest = Forest(self.size)
location1 = (1, 3)
location2 = (2, 3)
location3 = (4, 4)
forest.addTree(location1)
forest.addTree(location2)
forest.addTree(location3)
self.assertEqual(forest.getValue(location1), 2)
self.assertEqual(forest.getValue(location2), 2)
self.assertEqual(forest.getValue(location3), 1)
def test_getForestLocations(self):
forest = Forest(self.size)
location1 = (1, 2)
location2 = (1, 3)
location3 = (2, 3)
forest.addTree(location1)
forest.addTree(location2)
forest.addTree(location3)
tree_locations = forest.getForestLocations(location1)
self.assertEqual(len(tree_locations), 3)
def test_getForests(self):
forest = Forest(self.size)
location1 = (1, 3)
location2 = (2, 3)
location3 = (4, 4)
location4 = (1, 1)
location5 = (1, 2)
location6 = (4, 2)
forest.addTree(location1)
forest.addTree(location2)
forest.addTree(location3)
forest.addTree(location4)
forest.addTree(location5)
forest.addTree(location6)
forests = forest.getForests()
num_forests = 3
self.assertEqual(len(forests), num_forests)
def test_getUnoccupiedLocations(self):
forest = Forest(self.size)
location1 = (1, 3)
location2 = (2, 3)
location3 = (4, 4)
forest.addTree(location1)
forest.addTree(location2)
forest.addTree(location3)
unoccupied_locations = forest.getUnoccupiedLocations()
expected_ammount = self.size * self.size - 3
self.assertEqual(len(unoccupied_locations), expected_ammount)
def nextState(self, action):
response = self.scenario.nextState(action)
if self.game_state == "mountain":
if "Forest" == response:
self.game_state = "Forest"
self.scenario = Forest()
else:
self.game_state = "Castle"
#self.scenario = Castle()
elif self.game_state == "Forest":
if "Town" == response:
self.game_state = "Town"
self.scenario = Town()
elif self.game_state == "Castle":
pass
else:
#game_state is "town"
pass
def __evaluate(self, newObj, assignments, candidateForest, cluster):
if (self.DEBUG_MESSAGES):
print("=============================")
print("Evaluate Method Started...")
print(
"Current Partial Objective: %s || Candidate Partial Objective %s"
% (newObj, candidateForest.getTotalCost()))
if (newObj - candidateForest.getTotalCost() <= self.EPS):
newSolution = Forest()
newSolution.buildForestFromArray(
self.subproblemSolutionForest.getAssignmentsArray(),
self.facilities, self.customers)
partialSolution = Forest()
partialSolution.buildForestFromDict(
Util.getDictSolutionFromMIP(assignments), self.facilities,
self.customers)
currentForestObj = self.subproblemSolutionForest.getTotalCost()
newFacilities = set()
previousFacilities = set()
for tree in candidateForest.getTrees().values():
newSolution.removeTree(tree.getRoot().index)
previousFacilities.add(tree.getRoot().index)
for tree in partialSolution.getTrees().values():
newSolution.addTree(Tree(tree.getRoot()))
newFacilities.add(tree.getRoot().index)
for node in tree.getNodes().values():
newSolution.getTrees().get(
tree.getRoot().index).addNode(node)
#Facilities that were in the solution, but was not even selected as candidates
clusterIntersection = set([
tree.getRoot().index
for tree in self.subproblemSolutionForest.getTrees().values()
]).intersection(cluster)
notInterestingFacilities = clusterIntersection.difference(
previousFacilities)
for facilityIndex in notInterestingFacilities:
newSolution.removeTree(facilityIndex)
newSolution.updateStatistics()
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
print("Current Objective: %s || Candidate Objective: %s" %
(currentForestObj, newSolution.getTotalCost()))
if (self.params["improvementType"] == ImprovementType.Best):
if (Util.truncate(
Util.truncate(newSolution.getTotalCost(), 10) -
Util.truncate(
self.subproblemSolutionForest.getTotalCost(), 10),
10) <= self.EPS):
if (self.DEBUG_MESSAGES):
print("NEW SOLUTION FOUND!")
self.subproblemSolutionForest.buildForestFromArray(
newSolution.getAssignmentsArray(), self.facilities,
self.customers)
newForestObj = self.subproblemSolutionForest.getTotalCost()
self.__updateFrequency(list(newFacilities),
self.ASSIGNMENT_REWARD)
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
reward = Util.truncate(
float(len(newFacilities)) /
float(len(previousCandidates)), 3)
self.__updateFrequency(previousCandidates, reward)
if (self.DEBUG_MESSAGES):
print("Previous Objective: %s || New Objective: %s" %
(currentForestObj, newForestObj))
print("Partial Solution")
partial = ""
partial = '%.2f' % self.subproblemSolutionForest.getTotalCost(
) + ' ' + str(0) + '\n'
partial += ' '.join(
map(
str,
self.subproblemSolutionForest.
getAssignmentsArray()))
print(partial)
else:
candidates = [
tree.getRoot().index
for tree in candidateForest.getTrees().values()
]
#reward = (Util.truncate(float(candidateForest.getTreesCount()/self.facilitiesCount),3))*self.ASSIGNMENT_REWARD
reward = self.NO_ASSIGNMENT_REWARD
self.__updateFrequency(candidates, reward)
elif self.params["improvementType"] == ImprovementType.First:
self.subproblemSolutionForest.buildForestFromArray(
newSolution.getAssignmentsArray(), self.facilities,
self.customers)
newForestObj = self.subproblemSolutionForest.getTotalCost()
self.__updateFrequency(list(newFacilities),
self.ASSIGNMENT_REWARD)
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
reward = Util.truncate(
float(len(newFacilities)) / float(len(previousCandidates)),
3)
self.__updateFrequency(previousCandidates, reward)
if (self.DEBUG_MESSAGES):
print("Previous Objective: %s || New Objective: %s" %
(currentForestObj, newForestObj))
print("Partial Solution")
partial = ""
partial = '%.2f' % self.subproblemSolutionForest.getTotalCost(
) + ' ' + str(0) + '\n'
partial += ' '.join(
map(
str,
self.subproblemSolutionForest.getAssignmentsArray(
)))
print(partial)
candidates = [
tree.getRoot().index
for tree in candidateForest.getTrees().values()
]
#reward = (Util.truncate(float(candidateForest.getTreesCount()/self.facilitiesCount),3))*self.ASSIGNMENT_REWARD
reward = self.NO_ASSIGNMENT_REWARD
self.__updateFrequency(candidates, reward)
if (self.DEBUG_MESSAGES):
print("Evaluate Method Finished...")
print("=============================")
class LNS:
EPS = 1.e-10
DEBUG_MESSAGES = False
ASSIGNMENT_REWARD = 1
INITIAL_REWARD = 0.5
NO_ASSIGNMENT_REWARD = 0.25
MAX_PROBLEM_SIZE = 25000
def __init__(self, facilities, customers, params):
self.facilities = dict(
zip([facility.index for facility in facilities],
[facility for facility in facilities]))
self.customers = customers
self.subproblemSolutionForest = Forest()
self.currentIteration = 0
self.mip = MIP(facilities, customers,
"Instance_%s_%s" % (len(facilities), len(customers)),
params["mipSolver"])
self.facilitiesCount = len(facilities)
self.quantiles = []
self.params = params
self.totalDemand = sum([customer.demand for customer in customers])
self.currentObjectiveFunction = 0
self.currentSolutionAssignment = []
def __InitializeProblem(self, facilities, customers):
self.currentIteration = 0
self.clusterAreas = Preprocessing.getClusters(
facilities.values(), self.params["quantile_intervals"])
if (self.params["initialSolutionFunction"] ==
InitialSolutionFunction.Radius):
Preprocessing.getDistanceQuantiles(
facilities, self.params["quantile_intervals"])
self.subproblemSolutionForest.buildForestFromArray(
Util.formatSolutionFromMIP(
Preprocessing.getRadiusDistanceInitialSolution(
facilities, customers, self.clusterAreas.get(0))),
self.facilities, self.customers)
else:
self.subproblemSolutionForest.buildForestFromArray(
Util.formatSolutionFromMIP(
Preprocessing.getNearestNeighbourInitialSolution(
facilities, customers,
self.params["initialSolutionFunction"])),
self.facilities, self.customers)
for facility in facilities.keys():
self.facilities[facility] = self.facilities[facility]._replace(
frequency=self.INITIAL_REWARD)
def __getQuantiles(self):
firstQuantile = Util.truncate(1.0 / float(len(self.clusterAreas)), 10)
quantileIntervals = len(self.clusterAreas)
quantile = firstQuantile
for interval in range(0, quantileIntervals - 1):
self.quantiles.append(quantile)
quantile = Util.truncate(quantile + firstQuantile, 10)
self.quantiles.append(Util.truncate(quantile - firstQuantile, 10))
def __getCandidateFacilities(self,
cluster,
demand,
threshold,
fulfillDemand=True):
freqs = [self.facilities[i].frequency for i in cluster]
candidateIndexes = Util.filterbyThreshold(freqs, threshold,
self.currentIteration + 1)
result = [cluster[i] for i in candidateIndexes]
candidatesCapacity = 0
for index in result:
candidatesCapacity = candidatesCapacity + self.facilities[
index].capacity
print("Demand: %s || Capacity: %s" % (demand, candidatesCapacity))
if (candidatesCapacity < demand):
if fulfillDemand:
remainingFacilitiesIndex = set(cluster).difference(
set(cluster).intersection(set(result)))
remainingFacilities = [
self.facilities[i] for i in remainingFacilitiesIndex
]
remainingFacilities.sort(key=lambda x: x.cost_per_capacity,
reverse=True)
print("Demand above Capacity")
for facility in remainingFacilities:
result.append(facility.index)
candidatesCapacity = candidatesCapacity + facility.capacity
if (candidatesCapacity >= demand):
print("Demand: %s || New Capacity: %s" %
(demand, candidatesCapacity))
break
else:
return None
return result
def __updateFrequency(self, facilities, reward):
for index in facilities:
if (index not in self.facilities.keys()):
input("key does not exists!")
freq = self.facilities[index].frequency + reward
self.facilities[index] = self.facilities[index]._replace(
frequency=freq)
def __destroy(self, cluster):
if (self.DEBUG_MESSAGES):
print("=============================")
print("Destroy Method Started...")
clusterDemand = 0
for facilityIndex in cluster:
if (facilityIndex
in self.subproblemSolutionForest.getTrees().keys()):
for node in self.subproblemSolutionForest.getTrees().get(
facilityIndex).getNodes().values():
clusterDemand = clusterDemand + node.demand
#candidateFacilities = self.__getCandidateFacilities(cluster,clusterDemand,Util.truncate(self.quantiles[self.currentIteration],5))
candidateFacilities = copy.deepcopy(cluster)
print("Current Forest: %s/%s - Candidate Facilities: %s/%s" %
(self.subproblemSolutionForest.getTreesCount(),
self.subproblemSolutionForest.getTotalNodes(),
len(candidateFacilities), len(cluster)))
reassignmentCandidates = Forest()
for facilityIndex in candidateFacilities:
if (facilityIndex not in reassignmentCandidates.getTrees().keys()):
reassignmentCandidates.addTree(
Tree(self.facilities[facilityIndex]))
for facilityIndex in cluster:
if (facilityIndex
in self.subproblemSolutionForest.getTrees().keys()):
for node in self.subproblemSolutionForest.getTrees().get(
facilityIndex).getNodes().values():
reassignmentCandidates.getTrees().get(
candidateFacilities[0]).addNode(node)
reassignmentCandidates.updateStatistics()
if (self.DEBUG_MESSAGES):
print("Facilities: %s - Customers %s" %
(reassignmentCandidates.getTreesCount(),
reassignmentCandidates.getTotalNodes()))
print("Destroy Method Finished...")
print("=============================")
return reassignmentCandidates
def __repair(self, candidatesFacility, candidatesCustomer):
if (self.DEBUG_MESSAGES):
print("=============================")
print("Repair Method Started...")
self.mip.clear()
self.mip.initialize(
candidatesFacility, candidatesCustomer, "Instance_%s_%s" %
(len(candidatesFacility), len(candidatesCustomer)),
self.params["mipSolver"])
obj, assignments, status = self.mip.optimize(
self.params["mipTimeLimit"])
if (self.DEBUG_MESSAGES):
print("Repair Method Finished...")
print("=============================")
return obj, assignments, status
def __evaluate(self, newObj, assignments, candidateForest, cluster):
if (self.DEBUG_MESSAGES):
print("=============================")
print("Evaluate Method Started...")
print(
"Current Partial Objective: %s || Candidate Partial Objective %s"
% (newObj, candidateForest.getTotalCost()))
if (newObj - candidateForest.getTotalCost() <= self.EPS):
newSolution = Forest()
newSolution.buildForestFromArray(
self.subproblemSolutionForest.getAssignmentsArray(),
self.facilities, self.customers)
partialSolution = Forest()
partialSolution.buildForestFromDict(
Util.getDictSolutionFromMIP(assignments), self.facilities,
self.customers)
currentForestObj = self.subproblemSolutionForest.getTotalCost()
newFacilities = set()
previousFacilities = set()
for tree in candidateForest.getTrees().values():
newSolution.removeTree(tree.getRoot().index)
previousFacilities.add(tree.getRoot().index)
for tree in partialSolution.getTrees().values():
newSolution.addTree(Tree(tree.getRoot()))
newFacilities.add(tree.getRoot().index)
for node in tree.getNodes().values():
newSolution.getTrees().get(
tree.getRoot().index).addNode(node)
#Facilities that were in the solution, but was not even selected as candidates
clusterIntersection = set([
tree.getRoot().index
for tree in self.subproblemSolutionForest.getTrees().values()
]).intersection(cluster)
notInterestingFacilities = clusterIntersection.difference(
previousFacilities)
for facilityIndex in notInterestingFacilities:
newSolution.removeTree(facilityIndex)
newSolution.updateStatistics()
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
print("Current Objective: %s || Candidate Objective: %s" %
(currentForestObj, newSolution.getTotalCost()))
if (self.params["improvementType"] == ImprovementType.Best):
if (Util.truncate(
Util.truncate(newSolution.getTotalCost(), 10) -
Util.truncate(
self.subproblemSolutionForest.getTotalCost(), 10),
10) <= self.EPS):
if (self.DEBUG_MESSAGES):
print("NEW SOLUTION FOUND!")
self.subproblemSolutionForest.buildForestFromArray(
newSolution.getAssignmentsArray(), self.facilities,
self.customers)
newForestObj = self.subproblemSolutionForest.getTotalCost()
self.__updateFrequency(list(newFacilities),
self.ASSIGNMENT_REWARD)
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
reward = Util.truncate(
float(len(newFacilities)) /
float(len(previousCandidates)), 3)
self.__updateFrequency(previousCandidates, reward)
if (self.DEBUG_MESSAGES):
print("Previous Objective: %s || New Objective: %s" %
(currentForestObj, newForestObj))
print("Partial Solution")
partial = ""
partial = '%.2f' % self.subproblemSolutionForest.getTotalCost(
) + ' ' + str(0) + '\n'
partial += ' '.join(
map(
str,
self.subproblemSolutionForest.
getAssignmentsArray()))
print(partial)
else:
candidates = [
tree.getRoot().index
for tree in candidateForest.getTrees().values()
]
#reward = (Util.truncate(float(candidateForest.getTreesCount()/self.facilitiesCount),3))*self.ASSIGNMENT_REWARD
reward = self.NO_ASSIGNMENT_REWARD
self.__updateFrequency(candidates, reward)
elif self.params["improvementType"] == ImprovementType.First:
self.subproblemSolutionForest.buildForestFromArray(
newSolution.getAssignmentsArray(), self.facilities,
self.customers)
newForestObj = self.subproblemSolutionForest.getTotalCost()
self.__updateFrequency(list(newFacilities),
self.ASSIGNMENT_REWARD)
previousCandidates = list(
previousFacilities.difference(
newFacilities.intersection(previousFacilities)))
if (len(previousCandidates) == 0):
previousCandidates = list(newFacilities)
previousCandidates.extend(list(notInterestingFacilities))
reward = Util.truncate(
float(len(newFacilities)) / float(len(previousCandidates)),
3)
self.__updateFrequency(previousCandidates, reward)
if (self.DEBUG_MESSAGES):
print("Previous Objective: %s || New Objective: %s" %
(currentForestObj, newForestObj))
print("Partial Solution")
partial = ""
partial = '%.2f' % self.subproblemSolutionForest.getTotalCost(
) + ' ' + str(0) + '\n'
partial += ' '.join(
map(
str,
self.subproblemSolutionForest.getAssignmentsArray(
)))
print(partial)
candidates = [
tree.getRoot().index
for tree in candidateForest.getTrees().values()
]
#reward = (Util.truncate(float(candidateForest.getTreesCount()/self.facilitiesCount),3))*self.ASSIGNMENT_REWARD
reward = self.NO_ASSIGNMENT_REWARD
self.__updateFrequency(candidates, reward)
if (self.DEBUG_MESSAGES):
print("Evaluate Method Finished...")
print("=============================")
def optimize(self):
start = time.time()
clock = Clock()
clock.setStart(start)
if (self.DEBUG_MESSAGES):
print("=============================")
print("LNS Optimize Method Started...")
customerSubset = copy.deepcopy(self.customers)
facilitySubet = copy.deepcopy(self.facilities)
self.__InitializeProblem(facilitySubet, customerSubset)
self.__getQuantiles()
self.currentObjectiveFunction = self.subproblemSolutionForest.getTotalCost(
)
self.currentSolutionAssignment = self.subproblemSolutionForest.getAssignmentsArray(
)
initialQuantiles = copy.deepcopy(self.quantiles)
quantileSize = len(initialQuantiles)
quantilesCount = 0
customerCount = len(self.customers)
noImprovementIterations = 0
while True:
if (clock.isTimeOver(time.time(),
self.params["executionTimeLimit"])):
break
iterationsCount = len(self.clusterAreas)
for iteration in range(0, iterationsCount):
self.currentIteration = iteration
clustersCount = 0
clustersSize = len(self.clusterAreas.get(iteration))
for cluster in self.clusterAreas.get(iteration).values():
clustersCount = clustersCount + 1
print("Iteration: %s/%s || Instance: %s_%s" %
(quantilesCount + 1, quantileSize,
self.facilitiesCount, customerCount))
print("Subproblem: %s/%s" %
(self.currentIteration + 1, iterationsCount))
candidateForest = self.__destroy(cluster)
if (candidateForest.getTreesCount() *
candidateForest.getTotalNodes() >
self.MAX_PROBLEM_SIZE):
print(
"Problem instance is larger than limit. Skipping..."
)
candidateFacilities = [
tree.getRoot()
for tree in candidateForest.getTrees().values()
]
self.__updateFrequency(
dict([(facility.index, facility)
for facility in candidateFacilities]),
self.NO_ASSIGNMENT_REWARD)
continue
cFacilities, cCustomers = candidateForest.getData()
print(
"Current Cluster: %s/%s || Facilities: %s || Customers Assigned: %s"
% (clustersCount, clustersSize,
candidateForest.getTreesCount(),
candidateForest.getTotalNodes()))
if (candidateForest.getTotalNodes() == 0):
if (self.DEBUG_MESSAGES):
print("No Customers Assigned... Continue")
continue
obj, assignment, status = self.__repair(
cFacilities, cCustomers)
if (status == 'optimal'):
self.__evaluate(obj, assignment, candidateForest,
cluster)
else:
print("No Optimal Solution Found for this instance")
candidateFacilities = [
tree.getRoot()
for tree in candidateForest.getTrees().values()
]
self.__updateFrequency(
dict([(facility.index, facility)
for facility in candidateFacilities]),
self.ASSIGNMENT_REWARD)
print("Subproblem Forest: %s/%s" %
(self.subproblemSolutionForest.getTreesCount(),
self.subproblemSolutionForest.getTotalNodes()))
print("Subproblem Objective Funciton: %s" %
self.subproblemSolutionForest.getTotalCost())
print("Current Objective Function: %s" %
self.currentObjectiveFunction)
if (self.DEBUG_MESSAGES):
print("Partial Solution")
partial = ""
partial = '%.2f' % self.subproblemSolutionForest.getTotalCost(
) + ' ' + str(0) + '\n'
partial += ' '.join(
map(
str,
self.subproblemSolutionForest.getAssignmentsArray(
)))
print(partial)
if (self.currentObjectiveFunction >=
self.subproblemSolutionForest.getTotalCost()):
self.currentObjectiveFunction = self.subproblemSolutionForest.getTotalCost(
)
self.currentSolutionAssignment = self.subproblemSolutionForest.getAssignmentsArray(
)
else:
noImprovementIterations = noImprovementIterations + 1
print("====================================================")
print("CURRENT OBJECTIVE FUNCTION: %s" %
self.currentObjectiveFunction)
print("====================================================")
if (quantilesCount >= quantileSize):
print("Maximum Iteration Count Reached! Stopping...")
break
if (noImprovementIterations >
self.params["noImprovementIterationLimit"]):
print("No improvement limit reached! Stopping the search...")
break
##filtrar as facilities mais interessantes e jogar no facility subset
candidates = [
facility.index for facility in facilitySubet.values()
]
lastCandidateCount = len(candidates)
while quantilesCount < quantileSize and len(
candidates) == lastCandidateCount:
candidates = self.__getCandidateFacilities(
candidates, self.totalDemand,
Util.truncate(initialQuantiles[quantilesCount], 5), False)
quantilesCount = quantilesCount + 1
if (candidates is None or len(candidates) == 0
or len(candidates) == lastCandidateCount):
break
facilitySubet = dict(
zip([index for index in candidates],
[facilitySubet[index] for index in candidates]))
self.facilities = facilitySubet
self.__InitializeProblem(facilitySubet, customerSubset)
self.__getQuantiles()
return self.currentObjectiveFunction, self.currentSolutionAssignment
def __destroy(self, cluster):
if (self.DEBUG_MESSAGES):
print("=============================")
print("Destroy Method Started...")
clusterDemand = 0
for facilityIndex in cluster:
if (facilityIndex
in self.subproblemSolutionForest.getTrees().keys()):
for node in self.subproblemSolutionForest.getTrees().get(
facilityIndex).getNodes().values():
clusterDemand = clusterDemand + node.demand
#candidateFacilities = self.__getCandidateFacilities(cluster,clusterDemand,Util.truncate(self.quantiles[self.currentIteration],5))
candidateFacilities = copy.deepcopy(cluster)
print("Current Forest: %s/%s - Candidate Facilities: %s/%s" %
(self.subproblemSolutionForest.getTreesCount(),
self.subproblemSolutionForest.getTotalNodes(),
len(candidateFacilities), len(cluster)))
reassignmentCandidates = Forest()
for facilityIndex in candidateFacilities:
if (facilityIndex not in reassignmentCandidates.getTrees().keys()):
reassignmentCandidates.addTree(
Tree(self.facilities[facilityIndex]))
for facilityIndex in cluster:
if (facilityIndex
in self.subproblemSolutionForest.getTrees().keys()):
for node in self.subproblemSolutionForest.getTrees().get(
facilityIndex).getNodes().values():
reassignmentCandidates.getTrees().get(
candidateFacilities[0]).addNode(node)
reassignmentCandidates.updateStatistics()
if (self.DEBUG_MESSAGES):
print("Facilities: %s - Customers %s" %
(reassignmentCandidates.getTreesCount(),
reassignmentCandidates.getTotalNodes()))
print("Destroy Method Finished...")
print("=============================")
return reassignmentCandidates
# nos movemos hacia el frame que se desea evaluar.
imagen_original.seek(frame)
imagen_procesada.seek(frame)
# imprimimos la imagen original y procesada
# imagen_original.show()
# imagen_procesada.show()
imagen_original.save("original.png")
# imprimimos las imagenes por los clasificadores
imagen_caracteristicas = myImage(512, 512)
imagen_caracteristicas.readDataFromCSV("Datos/" + str(frame) + ".csv")
# creamos bosque para clasificar
forest = Forest()
# imprimimos imagen por las caracteristicas
forest.generarImagenArbolIntensidad(
imagen_caracteristicas).obtenerImagenPorClase().show()
forest.generarImagenArbolGradiente(
imagen_caracteristicas).obtenerImagenPorClase().show()
forest.generarImagenArbolLaplaciano(
imagen_caracteristicas).obtenerImagenPorClase().show()
# imprimimos imagen utilizando el bosque
forest.generarImagenUsandoBosque(
imagen_caracteristicas).obtenerImagenPorClase().show()
if EvOtrosClasificadores:
def __init__(self,components,nclasses=2,class_dist=np.array([]),p=0):
'''
Grow the Forest and store the leaves to compute a function to label the points
'''
self.components = map(np.array,components)
self.p = p
self.num_components = len(components)
self.num_nodes = np.array(map(np.prod,components)) + 1
self.num_vars = map(len,components)
self.n = sum(self.num_vars)
self.forest = Forest(components)
self.leaves = self.forest.get_leaves()
self.nleaves = len(self.leaves)
self.cardinalities = np.concatenate([self.components])[0]
if len(class_dist) == 0:
self.nclasses = nclasses
self.classes = ['C%d'%(x) for x in np.arange(1,self.nclasses+1)]
self.class_dist = class_dist
else:
self.class_dist=np.array(class_dist)
self.nclasses = len(self.class_dist)
self.classes = ['C%d'%(x) for x in np.arange(1,self.nclasses+1)]
self.class_limit = np.floor(self.class_dist * self.nleaves)
'''
Assign leaves to classes
Currently:
(1) deterministically switching the classes of the leaves
(2) Any number of classes (provided the number is less than the number of leaves)
'''
self.label_function = {}
if len(self.class_dist) == 0:
counter = 1
for leaf in self.leaves:
self.label_function[leaf] = self.classes[np.mod(counter,self.nclasses)]
counter = counter + 1
else:
self.class_counts = np.zeros(self.nclasses)
counter = 0
for leaf in self.leaves:
if self.class_counts[ np.mod(counter,self.nclasses) ] < self.class_limit[np.mod(counter,self.nclasses)]:
self.label_function[leaf] = self.classes[np.mod(counter,self.nclasses)]
self.class_counts[np.mod(counter,self.nclasses) ] += 1
counter = counter + 1
else:
possible_labels, = np.where(self.class_counts < self.class_limit)
possible_labels = possible_labels+1
if(len(possible_labels)>0):
alternative_label = np.random.choice(possible_labels)
self.label_function[leaf] = 'C%d'%alternative_label
self.class_counts[alternative_label-1] += 1
else:
self.label_function[leaf]='C1'
self.class_counts[0] += 1
counter = counter + 1
hyper_parameters_options = [
{
'number_of_trees': 5,
'number_of_attr_samples': 5
}
]
print(f'Dataset será dividido em {divisions} partes, das quais uma para teste e uma para validação')
dataset = DatasetFile("path").read
divisor = StratifiedDivisor(dataset, divisions)
forests = []
for hyper_parameter in hyper_parameters_options:
forests.append(Forest(hyper_parameter['number_of_trees'],
hyper_parameter['number_of_attr_samples']
))
performances = []
for forest in forests:
performances.append(ModelPerformance())
# para cada possibilidade de divisão do dataset original (Ver Cross Validation)
for i in range(divisions):
training_set = divisor.get_training_set(i)
test_set = divisor.get_test_set(i)
validation_set = divisor.get_validation_set(i)
# calcular a performance de cada floresta (média e desvio padrão)
for forest_index in range(len(forests)):
forest = forests[forest_index]
class RFSampler:
def __init__(self,components,nclasses=2,class_dist=np.array([]),p=0):
'''
Grow the Forest and store the leaves to compute a function to label the points
'''
self.components = map(np.array,components)
self.p = p
self.num_components = len(components)
self.num_nodes = np.array(map(np.prod,components)) + 1
self.num_vars = map(len,components)
self.n = sum(self.num_vars)
self.forest = Forest(components)
self.leaves = self.forest.get_leaves()
self.nleaves = len(self.leaves)
self.cardinalities = np.concatenate([self.components])[0]
if len(class_dist) == 0:
self.nclasses = nclasses
self.classes = ['C%d'%(x) for x in np.arange(1,self.nclasses+1)]
self.class_dist = class_dist
else:
self.class_dist=np.array(class_dist)
self.nclasses = len(self.class_dist)
self.classes = ['C%d'%(x) for x in np.arange(1,self.nclasses+1)]
self.class_limit = np.floor(self.class_dist * self.nleaves)
'''
Assign leaves to classes
Currently:
(1) deterministically switching the classes of the leaves
(2) Any number of classes (provided the number is less than the number of leaves)
'''
self.label_function = {}
if len(self.class_dist) == 0:
counter = 1
for leaf in self.leaves:
self.label_function[leaf] = self.classes[np.mod(counter,self.nclasses)]
counter = counter + 1
else:
self.class_counts = np.zeros(self.nclasses)
counter = 0
for leaf in self.leaves:
if self.class_counts[ np.mod(counter,self.nclasses) ] < self.class_limit[np.mod(counter,self.nclasses)]:
self.label_function[leaf] = self.classes[np.mod(counter,self.nclasses)]
self.class_counts[np.mod(counter,self.nclasses) ] += 1
counter = counter + 1
else:
possible_labels, = np.where(self.class_counts < self.class_limit)
possible_labels = possible_labels+1
if(len(possible_labels)>0):
alternative_label = np.random.choice(possible_labels)
self.label_function[leaf] = 'C%d'%alternative_label
self.class_counts[alternative_label-1] += 1
else:
self.label_function[leaf]='C1'
self.class_counts[0] += 1
counter = counter + 1
'''
Method to generate a sample based on the tree
'''
def get_sample(self,N,filename):
col_names = ['A%d'%(x) for x in np.arange(1,self.n+2)]
sim_result = pd.DataFrame(index = np.arange(N),columns=col_names)
for sample_num in np.arange(N):
var_index = 0
old_value = 0
for comp_index in np.arange(self.num_components):
curr_component = self.forest.get_tree(comp_index)
new_value = np.random.choice(curr_component[self.forest.roots[comp_index]])
var_index = var_index + 1 if var_index > 0 else 0
sim_result.ix[sample_num, var_index ] = new_value
old_value = new_value
for i in np.arange(1,self.num_vars[comp_index]):
var_index = var_index + 1
new_value = np.random.choice(curr_component[old_value])
sim_result.ix[sample_num,var_index] = (new_value - (np.sum([np.prod(self.cardinalities[:j]) for j in np.arange(var_index-1)]) +1)) % self.cardinalities[var_index] + 1
old_value = new_value
if(np.random.random() < self.p ):
other_classes = list(set(self.classes) - set(self.label_function[new_value]))
sim_result.ix[sample_num,self.n] = np.random.choice( other_classes )
else:
sim_result.ix[sample_num,self.n] = self.label_function[new_value]
sim_result.to_csv(filename,index=False)
''' Created on Jan 17, 2010 @author: sahar ''' from Bin import Bin from Node import Node from Tube import Tube from Forest import Forest import math import numpy as np import xml.etree.ElementTree as ET forest = Forest() forest.initializeGrid() forest.generateBaseNodes() forest.addBaseSegments() def step(dT): pass running = False while running: if segmentNum > MAX_SEGMENTS: running = False
def __init__(self, grid_size, design_parameter):
self.grid_size = grid_size
self.design_parameter = design_parameter
self.prob_scale = self.getCharacteristicScale()
self.norm_const = self.getNormalizationConstant()
self.forest = Forest(self.grid_size)
class HotModel:
def __init__(self, grid_size, design_parameter):
self.grid_size = grid_size
self.design_parameter = design_parameter
self.prob_scale = self.getCharacteristicScale()
self.norm_const = self.getNormalizationConstant()
self.forest = Forest(self.grid_size)
def getNormalizationConstant(self):
return ((1 - exp(1 / self.prob_scale)) /
(1 - exp(-self.grid_size / self.prob_scale)))**2
def getCharacteristicScale(self):
return self.grid_size / 10
def probability(self, location):
# characteristic scale for the distribution
l = self.prob_scale
x = location[0] + 1
y = location[1] + 1
return self.norm_const * exp(-x / l) * exp(-y / l)
def isFullyPopulated(self):
return self.getNumTrees() >= self.grid_size**2
def getNumTrees(self):
return self.forest.getNumTrees()
def getNormalizedNumTrees(self):
return self.getNumTrees() / self.grid_size**2
def getForestYield(self):
return self.normalizedAverageYieldFrom(self.forest)
def getForestMatrix(self):
matrix = np.copy(self.forest.getForestMatrix())
return matrix
def getForests(self):
return self.forest.getForests()
def normalizedAverageYieldFrom(self, forest):
return self.averageYieldFrom(forest) / self.grid_size**2
def averageYieldFrom(self, forest):
return forest.getNumTrees() - self.averageLightningDamageFrom(forest)
def averageLightningDamageFrom(self, forest):
forest_matrix = forest.getForestMatrix()
damage = 0
for index, forest_size in np.ndenumerate(forest_matrix):
loc_prob = self.probability(index)
damage += forest_size * loc_prob
return damage
def addTree(self):
location = self.getNextTreeLocation()
self.forest.addTree(location)
def getNextTreeLocation(self):
best_location = None
best_yield = -1
for location in self.getFutureSearchLocations():
simulated_forest = self.forest.copy()
simulated_forest.addTree(location)
avg_yield = self.averageYieldFrom(simulated_forest)
if avg_yield > best_yield:
best_yield = avg_yield
best_location = location
if best_location is None:
raise RuntimeError('Next Tree Location Not Found')
return best_location
def getFutureSearchLocations(self):
possible_locations = self.forest.getUnoccupiedLocations()
shuffle(possible_locations)
max_index = min(self.design_parameter, len(possible_locations))
search_locations = possible_locations[0:max_index]
return search_locations
class World(object):
def __init__(self, player):
self.array = []
self.player = player
self.c = 0
self.towns = [Town("Farnfos", [0, 0]), Town("Lunari", [1, 1]),
Town("Cewmann", [2, 2]),
Town("Caister", [3, 3]), Town("Aramore", [4, 4]),
Town("Kilerth", [5, 5]),
Town("Eldhams", [6, 6]), Town("Cirrane", [7, 7]),
Town("Warcest", [8, 8])]
self.forest = Forest()
def menu(self):
if self.c == 1:
self.root2.destroy()
self.c = 0
self.root = tk.Tk()
self.array = []
start = self.player.location
for i in range(len(self.towns)):
st = self.towns[i].location
if math.sqrt((st[0] - start[0]) * (st[0] - start[0]) + (st[1] - start[1]) * (st[1] - start[1])) < 2:
self.array.append(self.towns[i].name)
frame = tk.Frame(self.root, bg='#80c1ff', bd=5)
frame.place(relx=0, rely=0, relwidth=1, relheight=1)
button = tk.Button(frame, text=self.array[0], command=lambda: self.town(self.array[0]))
button.place(relx=0, rely=.2 * 0, relwidth=.5, relheight=.1)
button = tk.Button(frame, text=self.array[1], command=lambda: self.town(self.array[1]))
button.place(relx=0, rely=.2 * 1, relwidth=.5, relheight=.1)
button = tk.Button(frame, text=self.array[2], command=lambda: self.town(self.array[2]))
button.place(relx=0, rely=.2 * 2, relwidth=.5, relheight=.1)
button = tk.Button(frame, text="forest", command=lambda: self.forrest())
button.place(relx=0, rely=.2 * 4, relwidth=.5, relheight=.1)
button = tk.Button(frame, text="Quests", command=lambda: self.quest())
button.place(relx=0, rely=.2 * 3, relwidth=.5, relheight=.1)
self.root.mainloop()
def town(self, name):
self.root.destroy()
for i in range(len(self.towns)):
if name == self.towns[i].name:
self.player.location = self.towns[int(i)].roads[0].start
self.towns[int(i)].menu(self.player.location, self.player)
self.player.location = self.towns[int(i)].location
self.menu()
def forrest(self):
self.root.destroy()
self.forest.move(self.player, 0)
self.menu()
def quest(self):
self.root.destroy()
self.c = 1
self.root2 = tk.Tk()
frame = tk.Frame(self.root2, bg='#80c1ff', bd=5)
frame.place(relx=0, rely=0, relwidth=1, relheight=1)
if len(self.player.quest) > 0:
for i in range(len(self.player.quest)):
self.label = tk.Label(frame, text=self.player.quest[i], bg='#80c1ff')
self.label.place(x=0, y=10 + 20 * i)
button = tk.Button(frame, text="leave", command=lambda: self.menu())
button.place(relx=.5, y=10, relwidth=.5, relheight=.1)
else:
self.label = tk.Label(frame, text="No Quests", bg='#80c1ff')
self.label.place(x=0, y=0)
button = tk.Button(frame, text="leave", command=lambda: self.menu())
button.place(relx=.5, rely=0, relwidth=.5, relheight=.1)
self.root2.mainloop()
def analyze(file, folds=10):
"""
Executes the main analysis of a Dataset using the algorithm.
Reading the dataset from 'file' csv, executes a k-folds Accuracy analyisis,
with k=folds.
:param file: File where reading the dataset
:param folds: K of the k-folds analysis. If -1 then a Leave One Out
analysis will be performed.
:return:
Prints the mean validation accuracy.
"""
folder = os.path.join('Results', 'dataset-' + file[:-len('.csv')])
if not os.path.isdir(folder):
os.makedirs(folder)
for forest_type, params in experiments.items():
forest_folder = os.path.join(folder, forest_type)
if not os.path.isdir(forest_folder):
os.makedirs(forest_folder)
acc_matrix = []
for NT in params['NT']:
acc_row = [NT]
for F in params['F']:
# For each fold
accuracy = []
attributes_relevance = []
for i, (X_train, Y_train, X_val, Y_val, names) in enumerate(
get_dataset_k_folds(file=file, k=folds)):
M = X_train.shape[-1]
# Fit the forest
forest = Forest(F=F(M), NT=NT,
forest_type=forest_type).fit(
X=X_train,
Y=Y_train,
X_names=names[:-1])
# Predict values for the validation fold
predicted = forest.predict(X=X_val, X_names=names[:-1])
# Extract the attribute relevance on this forest
attributes_relevance.append(forest.attributes_relevance())
# Saves the accuracy
accuracy.append(100 * np.sum(Y_val == predicted) /
len(Y_val))
# Verbose each 20 folds
if i % 20 == 0 and i > 0:
print(str(i) + " Folds Done")
accuracy = np.round(np.mean(accuracy), 1)
print(str(i) + " Folds Executed")
# Print Results
print("-" * 50)
print("Accuracy: " + str(accuracy) + "%")
acc_row.append(accuracy)
f_description = str(
F(M)) if forest_type == 'random_forest' else str(F(M)())
forest.save_attributes_relevance_plot(
file=os.path.join(forest_folder,
'F-' + f_description + '-NT-' + str(NT)),
title=forest_type.replace('_', ' ').title() +
' - Feature Relevance')
acc_matrix.append(acc_row)
print(forest_type)
with open(
os.path.join(forest_folder,
'Acc Matrix ' + str(folds) + '-folds.txt'),
'w') as f:
f.write(str(np.array(acc_matrix)))
def build(key):
if key[0] == "G":
if len(key) > 1:
return Grass(key[1])
else:
return Grass(None)
if key[0] == "M":
if len(key) > 1:
return Mountain(key[1])
else:
return Mountain(None)
if key[0] == "F":
if len(key) > 1:
return Forest(key[1])
else:
return Forest(None)
if key[0] == "W":
if len(key) > 1:
return Wall(key[1])
else:
return Wall(None)
if key[0] == "~":
if len(key) > 1:
return Water(key[1])
else:
return Water(None)
if key[0] == "S":
if len(key) > 1:
return Sand(key[1])
else:
return Sand(None)
if key[0] == "E":
if len(key) > 1:
return MetalWall(key[1])
else:
return MetalWall(None)
if key[0] == "T":
if len(key) > 1:
return MetalFloor(key[1])
else:
return MetalFloor(None)
if key[0] == "P":
if len(key) > 1:
return Plating(key[1])
else:
return Plating(None)
if key[0] == ".":
if len(key) > 1:
return Void(key[1])
else:
return Void(None)
if key[0] == "O":
if len(key) > 1:
return WoodFloor(key[1])
else:
return WoodFloor(None)
if key[0] == "L":
if len(key) > 1:
return WoodFloorDam(key[1])
else:
return WoodFloorDam(None)
if key[0] == "B":
if len(key) > 1:
return Bridge(key[1])
else:
return Bridge(None)
if key[0] == "-":
if len(key) > 1:
return BridgeTop(key[1])
else:
return BridgeTop(None)
if key[0] == "+":
if len(key) > 1:
return BridgeBot(key[1])
else:
return BridgeBot(None)
class MainWindow(QMainWindow, Ui_MainWindow):
def __init__(self):
super().__init__()
self.setupUi(self)
self.init_UI()
# Setting elements, connections and logic
def init_UI(self):
self.FAQ.clicked.connect(self.show_faq)
self.Generate_button.clicked.connect(self.generate_image_main_logic)
self.Save_button.setEnabled(False)
self.Save_button.clicked.connect(self.save_image)
# base_settings_update connections
self.base_settings = {'size': 200, 'zoom': 50}
self.Size_ComboBox.valueChanged.connect(self.base_settings_update)
self.Zoom_ComboBox.valueChanged.connect(self.base_settings_update)
# Biomes
self.Forest = Forest(self.ForestLayout)
self.Desert = Desert(self.DesertLayout)
self.Frost = Frost(self.FrostLayout)
def generate_image_main_logic(self):
QApplication.setOverrideCursor(Qt.WaitCursor)
# Generate default noise
if self.Generation_ComboBox.currentText() == 'Default':
print('Generated default map')
self.image = generate_perlin_noise(self.base_settings)
# Generate forest map
elif self.Generation_ComboBox.currentText() == 'Forest':
print('Generated forest map')
self.image = add_color(generate_perlin_noise(self.base_settings),
self.Forest.get_colors())
# Generate desert map
elif self.Generation_ComboBox.currentText() == 'Desert':
print('Generated desert map')
self.image = add_color(generate_perlin_noise(self.base_settings),
self.Desert.get_colors())
# Generate frost map
elif self.Generation_ComboBox.currentText() == 'Frost':
print('Generated frost map')
self.image = add_color(generate_perlin_noise(self.base_settings),
self.Frost.get_colors())
# Pixmap set
tmp_img = ImageQt(self.image)
pixmap = QPixmap.fromImage(tmp_img)
self.Result.setPixmap(pixmap)
self.Save_button.setEnabled(True)
QApplication.restoreOverrideCursor()
# Save image
def save_image(self):
file_name = QFileDialog.getSaveFileName(self, 'Save', '', '*.png')[0]
if file_name:
save_file = self.image
save_file.save(file_name)
# Updating base_settings dict
def base_settings_update(self):
self.base_settings['size'] = self.Size_ComboBox.value()
self.base_settings['zoom'] = self.Zoom_ComboBox.value()
print(self.base_settings)
def show_faq(self):
self.faq = FAQ()
self.faq.show()