def bilancedPartion(self,subTree, threshold, G):
monoWeight = True
if (subTree.data.number_of_nodes() <= 1): # if subtree is a leaf, return it
return subTree
else:
balanceCut = int(round(((self.epsilon / 3) / (
1 + self.epsilon / 3)) / 2 * subTree.data.number_of_nodes())) # Evaluation of epsilon according to formula provided in paper
listOfNodes = subTree.data.nodes() # Selection of two randon nodes in order to divide the graph in two separate parts
print("Avvio del taglio")
partition = kernighan_lin_bisection(subTree.data)
X = G.subgraph(partition[0]) # spotting the two sub-sets of division inside the original Graph G
Y = G.subgraph(partition[1])
print(str(len(subTree.data)) + " " + str(len(X)) + " " + str(len(Y)))
leftChild = Tree() # declaration of offsprings
leftChild.data = X # Right Child data assignment
leftChild.left = None
leftChild.right = None
rightChild = Tree()
rightChild.data = Y # Left Child data
rightChild.left = None
rightChild.right = None
subTree.left = leftChild
subTree.right = rightChild
if (len(partition[0]) > threshold):
self.bilancedPartion(leftChild, threshold,G)
if (len(partition[1]) > threshold):
self.bilancedPartion(rightChild, threshold,G)
def bilancedPartition(self, subTree, threshold, G):
if (subTree.data.number_of_nodes() <=
1): # if subtree is a leaf, return it
return subTree
else:
partition = kernighan_lin_bisection(subTree.data)
X = G.subgraph(
partition[0]
) # spotting the two sub-sets of division inside the original Graph G
Y = G.subgraph(partition[1])
leftChild = Tree() # declaration of offsprings
leftChild.data = X # Right Child data assignment
leftChild.left = None
leftChild.right = None
rightChild = Tree()
rightChild.data = Y # Left Child data
rightChild.left = None
rightChild.right = None
subTree.left = leftChild
subTree.right = rightChild
if (
self.getWeightsNodes(leftChild.data.node) > threshold
): #comments are the same of createSubTree function (before seen)
self.bilancedPartition(leftChild, threshold, G)
if (self.getWeightsNodes(rightChild.data.node) > threshold):
self.bilancedPartition(rightChild, threshold, G)
def createSubTree(self, partition, subTree, threshold, G): #
X = G.subgraph(str(x) for x in partition[0]) # spotting the two sub-sets divided inside the original Graph G
Y = G.subgraph(str(x) for x in partition[1])
leftChild = Tree() # declaration offsprings
leftChild.data = X # Right Child data assignment
leftChild.left = None
leftChild.right = None
rightChild = Tree()
rightChild.data = Y # Left Child data assignment
rightChild.left = None
rightChild.right = None
subTree.left = leftChild # childreen relation assignment
subTree.right = rightChild
if (len(partition[0]) > threshold): # progressive recursive approach from root to leaves
self.division(leftChild, threshold, G)
if (len(partition[
1]) > threshold): # treshold is evaluated in order to avoid the creation of parts that need to be removed (epsilon*n)/ 3*k
self.division(rightChild, threshold, G)
def buildTree(self,input):
i=0
lhp,rhp=self.parser.getProdAtIndex(input[i])
root=Node(lhp)
for symbol in rhp:
root.addChild(Node(symbol))
i+=1
self.Util(root,i,input)
tree=Tree(root)
return tree
def tree_structure(file):
line = file.readline()
tree_temp = Tree(line)
tree_temp.get_root().set_parent(None)
node = tree_temp.get_root()
branch = 0
for line in file:
# print(line)
count = 0
while line[0] == '=':
line = line[1:]
count += 1
if count == branch:
branch -= 1
temp = node
node = node.get_parent()
node = node.add_child(Node(line))
node.set_parent(temp)
elif count > branch:
branch += 1
temp = node
node = node.add_child(Node(line))
node.set_parent(temp)
elif count < branch:
levels = branch - count
branch -= levels
for i in range(levels+1):
node = node.get_parent()
temp = node
node = node.add_child(Node(line))
node.set_parent(temp)
return tree_temp
node.set_parent(temp)
elif count < branch:
levels = branch - count
branch -= levels
for i in range(levels+1):
node = node.get_parent()
temp = node
node = node.add_child(Node(line))
node.set_parent(temp)
return tree_temp
if __name__ == "__main__":
# f = open(sys.argv[1], "r")
filename = "testdata2.ir"
f = open(filename, "r")
try:
tree = tree_structure(f)
representation = ""
Tree.depth_tree(tree.get_root())
f.close()
except IOError as erno:
print("Could not find the file")
print("Error number is:", erno)
finally:
print("Done")