Comment implémenter un arbre binaire?
Quelle est la meilleure structure de données pouvant être utilisée pour implémenter Binary Tree en Python?
Voici ma mise en œuvre récursive simple de l'arbre binaire.
#!/usr/bin/python
class Node:
def __init__(self, val):
self.l = None
self.r = None
self.v = val
class Tree:
def __init__(self):
self.root = None
def getRoot(self):
return self.root
def add(self, val):
if(self.root == None):
self.root = Node(val)
else:
self._add(val, self.root)
def _add(self, val, node):
if(val < node.v):
if(node.l != None):
self._add(val, node.l)
else:
node.l = Node(val)
else:
if(node.r != None):
self._add(val, node.r)
else:
node.r = Node(val)
def find(self, val):
if(self.root != None):
return self._find(val, self.root)
else:
return None
def _find(self, val, node):
if(val == node.v):
return node
Elif(val < node.v and node.l != None):
self._find(val, node.l)
Elif(val > node.v and node.r != None):
self._find(val, node.r)
def deleteTree(self):
# garbage collector will do this for us.
self.root = None
def printTree(self):
if(self.root != None):
self._printTree(self.root)
def _printTree(self, node):
if(node != None):
self._printTree(node.l)
print str(node.v) + ' '
self._printTree(node.r)
# 3
# 0 4
# 2 8
tree = Tree()
tree.add(3)
tree.add(4)
tree.add(0)
tree.add(8)
tree.add(2)
tree.printTree()
print (tree.find(3)).v
print tree.find(10)
tree.deleteTree()
tree.printTree()
# simple binary tree
# in this implementation, a node is inserted between an existing node and the root
class BinaryTree():
def __init__(self,rootid):
self.left = None
self.right = None
self.rootid = rootid
def getLeftChild(self):
return self.left
def getRightChild(self):
return self.right
def setNodeValue(self,value):
self.rootid = value
def getNodeValue(self):
return self.rootid
def insertRight(self,newNode):
if self.right == None:
self.right = BinaryTree(newNode)
else:
tree = BinaryTree(newNode)
tree.right = self.right
self.right = tree
def insertLeft(self,newNode):
if self.left == None:
self.left = BinaryTree(newNode)
else:
tree = BinaryTree(newNode)
tree.left = self.left
self.left = tree
def printTree(tree):
if tree != None:
printTree(tree.getLeftChild())
print(tree.getNodeValue())
printTree(tree.getRightChild())
# test tree
def testTree():
myTree = BinaryTree("Maud")
myTree.insertLeft("Bob")
myTree.insertRight("Tony")
myTree.insertRight("Steven")
printTree(myTree)
En savoir plus à ce sujet ici: -C'est une très simple implémentation d'un arbre binaire.
This est un tutoriel de Nice avec des questions entre les deux
Implémentation simple de BST en Python
class TreeNode:
def __init__(self, value):
self.left = None;
self.right = None;
self.data = value;
class Tree:
def __init__(self):
self.root = None;
def addNode(self, node, value):
if(node==None):
self.root = TreeNode(value);
else:
if(value<node.data):
if(node.left==None):
node.left = TreeNode(value)
else:
self.addNode(node.left, value);
else:
if(node.right==None):
node.right = TreeNode(value)
else:
self.addNode(node.right, value);
def printInorder(self, node):
if(node!=None):
self.printInorder(node.left)
print(node.data)
self.printInorder(node.right)
def main():
testTree = Tree()
testTree.addNode(testTree.root, 200)
testTree.addNode(testTree.root, 300)
testTree.addNode(testTree.root, 100)
testTree.addNode(testTree.root, 30)
testTree.printInorder(testTree.root)
Une façon très rapide d'implémenter une arborescence binaire en utilisant des listes . Pas la plus efficace, elle ne gère pas trop les valeurs nulles . Mais c'est très transparent (du moins pour moi):
def _add(node, v):
new = [v, [], []]
if node:
left, right = node[1:]
if not left:
left.extend(new)
Elif not right:
right.extend(new)
else:
_add(left, v)
else:
node.extend(new)
def binary_tree(s):
root = []
for e in s:
_add(root, e)
return root
def traverse(n, order):
if n:
v = n[0]
if order == 'pre':
yield v
for left in traverse(n[1], order):
yield left
if order == 'in':
yield v
for right in traverse(n[2], order):
yield right
if order == 'post':
yield v
Construire un arbre à partir d'un itérable:
>>> tree = binary_tree('A B C D E'.split())
>>> print tree
['A', ['B', ['D', [], []], ['E', [], []]], ['C', [], []]]
Traverser un arbre:
>>> list(traverse(tree, 'pre')), list(traverse(tree, 'in')), list(traverse(tree, 'post'))
(['A', 'B', 'D', 'E', 'C'],
['D', 'B', 'E', 'A', 'C'],
['D', 'E', 'B', 'C', 'A'])
Je ne peux pas m'empêcher de remarquer que la plupart des réponses ici implémentent un arbre de recherche binaire. Arbre de recherche binaire! = Arbre binaire.
Un arbre de recherche binaire a une propriété très spécifique: pour tout nœud X, la clé de X est plus grande que la clé de tout descendant de son enfant gauche et plus petite que celle de tout descendant de son enfant de droite.
Un arbre binaire n'impose aucune telle restriction. Un arbre binaire est simplement une structure de données avec un élément 'clé', et deux enfants, disent 'gauche' et 'droite'.
Un arbre est un cas encore plus général d'arbre binaire où chaque nœud peut avoir un nombre arbitraire d'enfants. En général, chaque nœud a un élément 'children' de type list/array.
Maintenant, pour répondre à la question du PO, j'inclus une implémentation complète d'un arbre binaire en Python. La structure de données sous-jacente stockant chaque BinaryTreeNode est un dictionnaire, étant donné qu'il offre des recherches optimales O(1). J'ai également implémenté des traversées en profondeur d'abord et en largeur d'abord. Ce sont des opérations très courantes effectuées sur des arbres.
from collections import deque
class BinaryTreeNode:
def __init__(self, key, left=None, right=None):
self.key = key
self.left = left
self.right = right
def __repr__(self):
return "%s l: (%s) r: (%s)" % (self.key, self.left, self.right)
def __eq__(self, other):
if self.key == other.key and \
self.right == other.right and \
self.left == other.left:
return True
else:
return False
class BinaryTree:
def __init__(self, root_key=None):
# maps from BinaryTreeNode key to BinaryTreeNode instance.
# Thus, BinaryTreeNode keys must be unique.
self.nodes = {}
if root_key is not None:
# create a root BinaryTreeNode
self.root = BinaryTreeNode(root_key)
self.nodes[root_key] = self.root
def add(self, key, left_key=None, right_key=None):
if key not in self.nodes:
# BinaryTreeNode with given key does not exist, create it
self.nodes[key] = BinaryTreeNode(key)
# invariant: self.nodes[key] exists
# handle left child
if left_key is None:
self.nodes[key].left = None
else:
if left_key not in self.nodes:
self.nodes[left_key] = BinaryTreeNode(left_key)
# invariant: self.nodes[left_key] exists
self.nodes[key].left = self.nodes[left_key]
# handle right child
if right_key == None:
self.nodes[key].right = None
else:
if right_key not in self.nodes:
self.nodes[right_key] = BinaryTreeNode(right_key)
# invariant: self.nodes[right_key] exists
self.nodes[key].right = self.nodes[right_key]
def remove(self, key):
if key not in self.nodes:
raise ValueError('%s not in tree' % key)
# remove key from the list of nodes
del self.nodes[key]
# if node removed is left/right child, update parent node
for k in self.nodes:
if self.nodes[k].left and self.nodes[k].left.key == key:
self.nodes[k].left = None
if self.nodes[k].right and self.nodes[k].right.key == key:
self.nodes[k].right = None
return True
def _height(self, node):
if node is None:
return 0
else:
return 1 + max(self._height(node.left), self._height(node.right))
def height(self):
return self._height(self.root)
def size(self):
return len(self.nodes)
def __repr__(self):
return str(self.traverse_inorder(self.root))
def bfs(self, node):
if not node or node not in self.nodes:
return
reachable = []
q = deque()
# add starting node to queue
q.append(node)
while len(q):
visit = q.popleft()
# add currently visited BinaryTreeNode to list
reachable.append(visit)
# add left/right children as needed
if visit.left:
q.append(visit.left)
if visit.right:
q.append(visit.right)
return reachable
# visit left child, root, then right child
def traverse_inorder(self, node, reachable=None):
if not node or node.key not in self.nodes:
return
if reachable is None:
reachable = []
self.traverse_inorder(node.left, reachable)
reachable.append(node.key)
self.traverse_inorder(node.right, reachable)
return reachable
# visit left and right children, then root
# root of tree is always last to be visited
def traverse_postorder(self, node, reachable=None):
if not node or node.key not in self.nodes:
return
if reachable is None:
reachable = []
self.traverse_postorder(node.left, reachable)
self.traverse_postorder(node.right, reachable)
reachable.append(node.key)
return reachable
# visit root, left, then right children
# root is always visited first
def traverse_preorder(self, node, reachable=None):
if not node or node.key not in self.nodes:
return
if reachable is None:
reachable = []
reachable.append(node.key)
self.traverse_preorder(node.left, reachable)
self.traverse_preorder(node.right, reachable)
return reachable
vous n'avez pas besoin d'avoir deux cours
class Tree:
val = None
left = None
right = None
def __init__(self, val):
self.val = val
def insert(self, val):
if self.val is not None:
if val < self.val:
if self.left is not None:
self.left.insert(val)
else:
self.left = Tree(val)
Elif val > self.val:
if self.right is not None:
self.right.insert(val)
else:
self.right = Tree(val)
else:
return
else:
self.val = val
print("new node added")
def showTree(self):
if self.left is not None:
self.left.showTree()
print(self.val, end = ' ')
if self.right is not None:
self.right.showTree()
Un peu plus "Pythonic"?
class Node:
def __init__(self, value):
self.value = value
self.left = None
self.right = None
def __repr__(self):
return str(self.value)
class BST:
def __init__(self):
self.root = None
def __repr__(self):
self.sorted = []
self.get_inorder(self.root)
return str(self.sorted)
def get_inorder(self, node):
if node:
self.get_inorder(node.left)
self.sorted.append(str(node.value))
self.get_inorder(node.right)
def add(self, value):
if not self.root:
self.root = Node(value)
else:
self._add(self.root, value)
def _add(self, node, value):
if value <= node.value:
if node.left:
self._add(node.left, value)
else:
node.left = Node(value)
else:
if node.right:
self._add(node.right, value)
else:
node.right = Node(value)
from random import randint
bst = BST()
for i in range(100):
bst.add(randint(1, 1000))
print (bst)
#!/usr/bin/python
class BinaryTree:
def __init__(self, left, right, data):
self.left = left
self.right = right
self.data = data
def pre_order_traversal(root):
print(root.data, end=' ')
if root.left != None:
pre_order_traversal(root.left)
if root.right != None:
pre_order_traversal(root.right)
def in_order_traversal(root):
if root.left != None:
in_order_traversal(root.left)
print(root.data, end=' ')
if root.right != None:
in_order_traversal(root.right)
def post_order_traversal(root):
if root.left != None:
post_order_traversal(root.left)
if root.right != None:
post_order_traversal(root.right)
print(root.data, end=' ')
Je sais que beaucoup de bonnes solutions ont déjà été publiées, mais j’ai généralement une approche différente pour les arbres binaires: aller avec une classe de noeuds et l’implémenter directement est plus lisible, mais quand vous avez beaucoup de noeuds, cela peut devenir très gourmand en mémoire, alors je suggère d'ajouter une couche de complexité et de stocker les nœuds dans une liste python, puis de simuler un comportement d'arborescence en utilisant uniquement la liste.
Vous pouvez toujours définir une classe de nœuds pour enfin représenter les nœuds de l'arborescence en cas de besoin, mais les conserver sous une forme simple [valeur, gauche, droite] dans une liste utilisera au moins la moitié de la mémoire!
Voici un exemple rapide d'une classe d'arborescence de recherche binaire stockant les nœuds dans un tableau. Il fournit des fonctions de base telles que ajouter, supprimer, trouver ...
"""
Basic Binary Search Tree class without recursion...
"""
__author__ = "@fbparis"
class Node(object):
__slots__ = "value", "parent", "left", "right"
def __init__(self, value, parent=None, left=None, right=None):
self.value = value
self.parent = parent
self.left = left
self.right = right
def __repr__(self):
return "<%s object at %s: parent=%s, left=%s, right=%s, value=%s>" % (self.__class__.__name__, hex(id(self)), self.parent, self.left, self.right, self.value)
class BinarySearchTree(object):
__slots__ = "_tree"
def __init__(self, *args):
self._tree = []
if args:
for x in args[0]:
self.add(x)
def __len__(self):
return len(self._tree)
def __repr__(self):
return "<%s object at %s with %d nodes>" % (self.__class__.__name__, hex(id(self)), len(self))
def __str__(self, nodes=None, level=0):
ret = ""
if nodes is None:
if len(self):
nodes = [0]
else:
nodes = []
for node in nodes:
if node is None:
continue
ret += "-" * level + " %s\n" % self._tree[node][0]
ret += self.__str__(self._tree[node][2:4], level + 1)
if level == 0:
ret = ret.strip()
return ret
def __contains__(self, value):
if len(self):
node_index = 0
while self._tree[node_index][0] != value:
if value < self._tree[node_index][0]:
node_index = self._tree[node_index][2]
else:
node_index = self._tree[node_index][3]
if node_index is None:
return False
return True
return False
def __eq__(self, other):
return self._tree == other._tree
def add(self, value):
if len(self):
node_index = 0
while self._tree[node_index][0] != value:
if value < self._tree[node_index][0]:
b = self._tree[node_index][2]
k = 2
else:
b = self._tree[node_index][3]
k = 3
if b is None:
self._tree[node_index][k] = len(self)
self._tree.append([value, node_index, None, None])
break
node_index = b
else:
self._tree.append([value, None, None, None])
def remove(self, value):
if len(self):
node_index = 0
while self._tree[node_index][0] != value:
if value < self._tree[node_index][0]:
node_index = self._tree[node_index][2]
else:
node_index = self._tree[node_index][3]
if node_index is None:
raise KeyError
if self._tree[node_index][2] is not None:
b, d = 2, 3
Elif self._tree[node_index][3] is not None:
b, d = 3, 2
else:
i = node_index
b = None
if b is not None:
i = self._tree[node_index][b]
while self._tree[i][d] is not None:
i = self._tree[i][d]
p = self._tree[i][1]
b = self._tree[i][b]
if p == node_index:
self._tree[p][5-d] = b
else:
self._tree[p][d] = b
if b is not None:
self._tree[b][1] = p
self._tree[node_index][0] = self._tree[i][0]
else:
p = self._tree[i][1]
if p is not None:
if self._tree[p][2] == i:
self._tree[p][2] = None
else:
self._tree[p][3] = None
last = self._tree.pop()
n = len(self)
if i < n:
self._tree[i] = last[:]
if last[2] is not None:
self._tree[last[2]][1] = i
if last[3] is not None:
self._tree[last[3]][1] = i
if self._tree[last[1]][2] == n:
self._tree[last[1]][2] = i
else:
self._tree[last[1]][3] = i
else:
raise KeyError
def find(self, value):
if len(self):
node_index = 0
while self._tree[node_index][0] != value:
if value < self._tree[node_index][0]:
node_index = self._tree[node_index][2]
else:
node_index = self._tree[node_index][3]
if node_index is None:
return None
return Node(*self._tree[node_index])
return None
J'ai ajouté un attribut parent afin que vous puissiez supprimer n'importe quel nœud et conserver la structure BST.
Désolé pour la lisibilité, en particulier pour la fonction "remove". En gros, lorsqu'un nœud est supprimé, nous ouvrons le tableau d'arbres et le remplaçons par le dernier élément (sauf si nous voulions supprimer le dernier nœud). Pour maintenir la structure BST, le nœud supprimé est remplacé par le maximum de ses enfants de gauche ou par ses enfants de droite et certaines opérations doivent être effectuées afin de conserver les index valides, mais c'est assez rapide.
J'ai utilisé cette technique pour créer des dictionnaires de mots volumineux avec un dictionnaire de base interne et j'ai pu diviser la consommation de mémoire par 7-8 (vous pouvez voir un exemple ici: https://Gist.github.com/fbparis/b3ddd5673b603b42c880974b23db7cda )
import random
class TreeNode:
def __init__(self, key):
self.key = key
self.left = None
self.right = None
self.p = None
class BinaryTree:
def __init__(self):
self.root = None
def length(self):
return self.size
def inorder(self, node):
if node == None:
return None
else:
self.inorder(node.left)
print node.key,
self.inorder(node.right)
def search(self, k):
node = self.root
while node != None:
if node.key == k:
return node
if node.key > k:
node = node.left
else:
node = node.right
return None
def minimum(self, node):
x = None
while node.left != None:
x = node.left
node = node.left
return x
def maximum(self, node):
x = None
while node.right != None:
x = node.right
node = node.right
return x
def successor(self, node):
parent = None
if node.right != None:
return self.minimum(node.right)
parent = node.p
while parent != None and node == parent.right:
node = parent
parent = parent.p
return parent
def predecessor(self, node):
parent = None
if node.left != None:
return self.maximum(node.left)
parent = node.p
while parent != None and node == parent.left:
node = parent
parent = parent.p
return parent
def insert(self, k):
t = TreeNode(k)
parent = None
node = self.root
while node != None:
parent = node
if node.key > t.key:
node = node.left
else:
node = node.right
t.p = parent
if parent == None:
self.root = t
Elif t.key < parent.key:
parent.left = t
else:
parent.right = t
return t
def delete(self, node):
if node.left == None:
self.transplant(node, node.right)
Elif node.right == None:
self.transplant(node, node.left)
else:
succ = self.minimum(node.right)
if succ.p != node:
self.transplant(succ, succ.right)
succ.right = node.right
succ.right.p = succ
self.transplant(node, succ)
succ.left = node.left
succ.left.p = succ
def transplant(self, node, newnode):
if node.p == None:
self.root = newnode
Elif node == node.p.left:
node.p.left = newnode
else:
node.p.right = newnode
if newnode != None:
newnode.p = node.p
Cette implémentation prend en charge les opérations d'insertion, de recherche et de suppression sans détruire la structure de l'arborescence. Ce n'est pas un arbre banlanced.
# Class for construct the nodes of the tree. (Subtrees)
class Node:
def __init__(self, key, parent_node = None):
self.left = None
self.right = None
self.key = key
if parent_node == None:
self.parent = self
else:
self.parent = parent_node
# Class with the structure of the tree.
# This Tree is not balanced.
class Tree:
def __init__(self):
self.root = None
# Insert a single element
def insert(self, x):
if(self.root == None):
self.root = Node(x)
else:
self._insert(x, self.root)
def _insert(self, x, node):
if(x < node.key):
if(node.left == None):
node.left = Node(x, node)
else:
self._insert(x, node.left)
else:
if(node.right == None):
node.right = Node(x, node)
else:
self._insert(x, node.right)
# Given a element, return a node in the tree with key x.
def find(self, x):
if(self.root == None):
return None
else:
return self._find(x, self.root)
def _find(self, x, node):
if(x == node.key):
return node
Elif(x < node.key):
if(node.left == None):
return None
else:
return self._find(x, node.left)
Elif(x > node.key):
if(node.right == None):
return None
else:
return self._find(x, node.right)
# Given a node, return the node in the tree with the next largest element.
def next(self, node):
if node.right != None:
return self._left_descendant(node.right)
else:
return self._right_ancestor(node)
def _left_descendant(self, node):
if node.left == None:
return node
else:
return self._left_descendant(node.left)
def _right_ancestor(self, node):
if node.key <= node.parent.key:
return node.parent
else:
return self._right_ancestor(node.parent)
# Delete an element of the tree
def delete(self, x):
node = self.find(x)
if node == None:
print(x, "isn't in the tree")
else:
if node.right == None:
if node.left == None:
if node.key < node.parent.key:
node.parent.left = None
del node # Clean garbage
else:
node.parent.right = None
del Node # Clean garbage
else:
node.key = node.left.key
node.left = None
else:
x = self.next(node)
node.key = x.key
x = None
# tests
t = Tree()
t.insert(5)
t.insert(8)
t.insert(3)
t.insert(4)
t.insert(6)
t.insert(2)
t.delete(8)
t.delete(5)
t.insert(9)
t.insert(1)
t.delete(2)
t.delete(100)
# Remember: Find method return the node object.
# To return a number use t.find(nº).key
# But it will cause an error if the number is not in the tree.
print(t.find(5))
print(t.find(8))
print(t.find(4))
print(t.find(6))
print(t.find(9))
Une bonne implémentation de binary search tree, extrait de ici :
'''
A binary search Tree
'''
from __future__ import print_function
class Node:
def __init__(self, label, parent):
self.label = label
self.left = None
self.right = None
#Added in order to delete a node easier
self.parent = parent
def getLabel(self):
return self.label
def setLabel(self, label):
self.label = label
def getLeft(self):
return self.left
def setLeft(self, left):
self.left = left
def getRight(self):
return self.right
def setRight(self, right):
self.right = right
def getParent(self):
return self.parent
def setParent(self, parent):
self.parent = parent
class BinarySearchTree:
def __init__(self):
self.root = None
def insert(self, label):
# Create a new Node
new_node = Node(label, None)
# If Tree is empty
if self.empty():
self.root = new_node
else:
#If Tree is not empty
curr_node = self.root
#While we don't get to a leaf
while curr_node is not None:
#We keep reference of the parent node
parent_node = curr_node
#If node label is less than current node
if new_node.getLabel() < curr_node.getLabel():
#We go left
curr_node = curr_node.getLeft()
else:
#Else we go right
curr_node = curr_node.getRight()
#We insert the new node in a leaf
if new_node.getLabel() < parent_node.getLabel():
parent_node.setLeft(new_node)
else:
parent_node.setRight(new_node)
#Set parent to the new node
new_node.setParent(parent_node)
def delete(self, label):
if (not self.empty()):
#Look for the node with that label
node = self.getNode(label)
#If the node exists
if(node is not None):
#If it has no children
if(node.getLeft() is None and node.getRight() is None):
self.__reassignNodes(node, None)
node = None
#Has only right children
Elif(node.getLeft() is None and node.getRight() is not None):
self.__reassignNodes(node, node.getRight())
#Has only left children
Elif(node.getLeft() is not None and node.getRight() is None):
self.__reassignNodes(node, node.getLeft())
#Has two children
else:
#Gets the max value of the left branch
tmpNode = self.getMax(node.getLeft())
#Deletes the tmpNode
self.delete(tmpNode.getLabel())
#Assigns the value to the node to delete and keesp tree structure
node.setLabel(tmpNode.getLabel())
def getNode(self, label):
curr_node = None
#If the tree is not empty
if(not self.empty()):
#Get tree root
curr_node = self.getRoot()
#While we don't find the node we look for
#I am using lazy evaluation here to avoid NoneType Attribute error
while curr_node is not None and curr_node.getLabel() is not label:
#If node label is less than current node
if label < curr_node.getLabel():
#We go left
curr_node = curr_node.getLeft()
else:
#Else we go right
curr_node = curr_node.getRight()
return curr_node
def getMax(self, root = None):
if(root is not None):
curr_node = root
else:
#We go deep on the right branch
curr_node = self.getRoot()
if(not self.empty()):
while(curr_node.getRight() is not None):
curr_node = curr_node.getRight()
return curr_node
def getMin(self, root = None):
if(root is not None):
curr_node = root
else:
#We go deep on the left branch
curr_node = self.getRoot()
if(not self.empty()):
curr_node = self.getRoot()
while(curr_node.getLeft() is not None):
curr_node = curr_node.getLeft()
return curr_node
def empty(self):
if self.root is None:
return True
return False
def __InOrderTraversal(self, curr_node):
nodeList = []
if curr_node is not None:
nodeList.insert(0, curr_node)
nodeList = nodeList + self.__InOrderTraversal(curr_node.getLeft())
nodeList = nodeList + self.__InOrderTraversal(curr_node.getRight())
return nodeList
def getRoot(self):
return self.root
def __isRightChildren(self, node):
if(node == node.getParent().getRight()):
return True
return False
def __reassignNodes(self, node, newChildren):
if(newChildren is not None):
newChildren.setParent(node.getParent())
if(node.getParent() is not None):
#If it is the Right Children
if(self.__isRightChildren(node)):
node.getParent().setRight(newChildren)
else:
#Else it is the left children
node.getParent().setLeft(newChildren)
#This function traversal the tree. By default it returns an
#In order traversal list. You can pass a function to traversal
#The tree as needed by client code
def traversalTree(self, traversalFunction = None, root = None):
if(traversalFunction is None):
#Returns a list of nodes in preOrder by default
return self.__InOrderTraversal(self.root)
else:
#Returns a list of nodes in the order that the users wants to
return traversalFunction(self.root)
#Returns an string of all the nodes labels in the list
#In Order Traversal
def __str__(self):
list = self.__InOrderTraversal(self.root)
str = ""
for x in list:
str = str + " " + x.getLabel().__str__()
return str
def InPreOrder(curr_node):
nodeList = []
if curr_node is not None:
nodeList = nodeList + InPreOrder(curr_node.getLeft())
nodeList.insert(0, curr_node.getLabel())
nodeList = nodeList + InPreOrder(curr_node.getRight())
return nodeList
def testBinarySearchTree():
r'''
Example
8
/ \
3 10
/ \ \
1 6 14
/ \ /
4 7 13
'''
r'''
Example After Deletion
7
/ \
1 4
'''
t = BinarySearchTree()
t.insert(8)
t.insert(3)
t.insert(6)
t.insert(1)
t.insert(10)
t.insert(14)
t.insert(13)
t.insert(4)
t.insert(7)
#Prints all the elements of the list in order traversal
print(t.__str__())
if(t.getNode(6) is not None):
print("The label 6 exists")
else:
print("The label 6 doesn't exist")
if(t.getNode(-1) is not None):
print("The label -1 exists")
else:
print("The label -1 doesn't exist")
if(not t.empty()):
print(("Max Value: ", t.getMax().getLabel()))
print(("Min Value: ", t.getMin().getLabel()))
t.delete(13)
t.delete(10)
t.delete(8)
t.delete(3)
t.delete(6)
t.delete(14)
#Gets all the elements of the tree In pre order
#And it prints them
list = t.traversalTree(InPreOrder, t.root)
for x in list:
print(x)
if __== "__main__":
testBinarySearchTree()
Une classe de nœud est le minimum nécessaire pour représenter un arbre binaire. Alors que d'autres réponses sont généralement correctes, elles ne sont pas nécessaires pour un arbre binaire (pas besoin d'étendre un objet, pas besoin de BST, pas besoin d'importer deque, etc.)
class Node:
def __init__(self, value = None):
self.left = None
self.right = None
self.value = value
Voici un exemple d'arbre:
n1 = Node(1)
n2 = Node(2)
n3 = Node(3)
n1.left = n2
n1.right = n3
Dans cet exemple, n1 est la racine de l'arbre ayant n2, n3 comme enfants.
Une classe de nœuds connectés basée sur Node est une approche standard. Ceux-ci peuvent être difficiles à visualiser.
Motivé par un essai sur Patterns Python - Graphiques d'implémentation, considérons un dictionnaire simple:
Donné
Un arbre binaire
a
/ \
b c
/ \ \
d e f
Code
Créez un dictionnaire de uniques _ nœuds:
tree = {
"a": ["b", "c"],
"b": ["d", "e"],
"c": [None, "f"],
"d": [None, None],
"e": [None, None],
"f": [None, None],
}
Détails
- Chaque paire clé-valeur est un nœud unique pointant vers ses enfants.
- Une liste (ou tuple) contient une paire ordonnée d’enfants gauche/droite.
- Avec un dict ayant ordonné l'insertion, supposons que la première entrée est la racine.
- Les méthodes courantes peuvent être des fonctions qui mutent ou traversent le dict (voir
find_all_paths()).
Les fonctions basées sur des arbres incluent souvent les opérations courantes suivantes:
- traverse: donne chaque noeud dans un ordre donné (généralement de gauche à droite)
- largeur-première recherche (BFS): niveaux transversaux
- recherche en profondeur d'abord (DFS): traverse les branches en premier (avant/après/après la commande)
- insert: ajoute un noeud à l'arborescence en fonction du nombre d'enfants
- remove: supprime un noeud en fonction du nombre d'enfants
- update: fusionne les noeuds manquants d'un arbre à l'autre
- visite: donne la valeur d'un noeud traversé
Nous démontrons ici une de ces fonctions - une traversée BFS:
Exemple
import collections as ct
def traverse(tree):
"""Yield nodes from a tree via BFS."""
q = ct.deque() # 1
root = next(iter(tree)) # 2
q.append(root)
while q:
node = q.popleft()
children = filter(None, tree.get(node))
for n in children: # 3
q.append(n)
yield node
list(traverse(tree))
# ['a', 'b', 'c', 'd', 'e', 'f']
Il s'agit d'un algorithme de recherche width-first (level-order) appliqué à un dict de noeuds et d'enfants.
- Initialiser une file d'attente FIFO _ . Nous utilisons une
deque, mais unequeueou unelist(cette dernière est inefficace). - Récupère et met en file d'attente le nœud racine (en supposant que la racine est la première entrée du dictée, Python 3.6+).
- Annulez temporairement un noeud, mettez ses enfants en file d'attente et donnez la valeur du noeud.
Voir aussi ce tutoriel en profondeur tutorial sur les arbres.