Summary

 Topics
• general trees, definitions and properties
• interface and implementation
• tree traversal algorithms
• depth and height
• pre-order traversal
• post-order traversal
• binary trees
• properties
• interface
• implementation

• binary search trees

• definition
• h-n relationship
• search, insert, delete
• performance

 READING:
• GT textbook chapter 7 and 10.1
2
 Trees
 So far we have seen linear structures
• linear: before and after relationship
• lists, vectors, arrays, stacks, queues, etc
 Non-linear structure: trees
• probably the most fundamental structure in computing
• hierarchical structure
• Terminology: from family trees (genealogy)

3
 Trees
root
 store elements hierarchically

 the top element: root

 except the root, each element has a parent

 each element has 0 or more children

4
 Trees
 Definition
• A tree T is a set of nodes storing elements such that the nodes have a parent-child
relationship that satisfies the following
• if T is not empty, T has a special tree called the root that has no parent
• each node v of T different than the root has a unique parent node w; each node with parent w is a child of w

 Recursive definition
• T is either empty
• or consists of a node r (the root) and a possibly empty set of trees whose roots are the
children of r

 Terminology
• siblings: two nodes that have the same parent are called siblings
• internal nodes
• nodes that have children
• external nodes or leaves
• nodes that don’t have children
• ancestors
• descendants
5
 Trees
root

internal nodes

leaves

6
 Trees

ancestors of u

7
 Trees

descendants of u

8
 Application of trees

 Applications of trees
• class hierarchy in Java
• file system
• storing hierarchies in organizations

9
 Tree ADT

 Whatever the implementation of a tree is, its interface is the following

• root()

• size()

• isEmpty()

• parent(v)

• children(v)

• isInternal(v)

• isExternal(v)

• isRoot()

10
 Tree Implementation

class Tree {
TreeNode root;

//tree ADT methods..

}

class TreeNode<Type> {
Type data;
int size;
TreeNode parent;
TreeNode firstChild;
TreeNode nextSibling;

getParent();
getChild();
getNextSibling();
}
11
 Algorithms on trees: Depth
 Depth:
• depth(T, v) is the number of ancestors of v, excluding v itself

 Recursive formulation
• if v == root, then depth(v) = 0
• else, depth(v) is 1 + depth (parent(v))

 Compute the depth of a node v in tree T: int depth(T, v)

 Algorithm:
int depth(T,v) {
if T.isRoot(v) return 0
return 1 + depth(T, T.parent(v))
}
 Analysis:
• O(number of ancestors) = O(depth_v)
• in the worst case the path is a linked-list and v is the leaf
• ==> O(n), where n is the number of nodes in the tree
12
 Algorithms on trees: Height
 Height:
• height of a node v in T is the length of the longest path from v to any leaf

 Recursive formulation:
• if v is leaf, then its height is 0
• else height(v) = 1 + maximum height of a child of v

 Definition: the height of a tree is the height of its root

 Compute the height of tree T: int height(T,v)

 Height and depth are “symmetrical”

 Proposition: the height of a tree T is the maximum depth of one of its leaves.

13
 Height
 Algorithm:
int height(T,v) {
if T.isExternal(v) return 0;
int h = 0;
for each child w of v in T do
h = max(h, height(T, w))
return h+1;
}
 Analysis:
• total time: the sum of times spent at all nodes in all recursive calls
• the recursion:
• v calls height(w) recursively on all children w of v
• height() will eventually be called on every descendant of v
• overall: height() is called on each node precisely once, because each node has one parent
• aside from recursion
• for each node v: go through all children of v
– O(1 + c_v) where c_v is the number of children of v
• over all nodes: O(n) + SUM (c_v)
– each node is child of only one node, so its processed once as a child
– SUM(c_v) = n - 1

14
• total: O(n), where n is the number of nodes in the tree
 Tree traversals
 A traversal is a systematic way to visit all nodes of T.

 pre-order: root, children

• parent comes before children; overall root first
 post-order: children, root
• parent comes after children; overall root last

void preorder(T, v)
visit v
for each child w of v in T do
preorder(w)

void postorder(T, v)
for each child w of v in T do
postorder(w)
visit v

Analysis: O(n) [same arguments as before]

15
 Examples
 Tree associated with a document

Pape
r

Title Abstract Ch1 Ch2 Ch3 Refs

1.1 1.2 3.1 3.2

 In what order do you read the document?

16
 Example
 Tree associated with an arithmetical expression
+

3 *

- +

12 5 1 7

 Write method that evaluates the expression. In what order do you traverse the tree?

17
Binary trees

18
 Binary trees
 Definition: A binary tree is a tree such that
• every node has at most 2 children
• each node is labeled as being either a left chilld or a right child

 Recursive definition:
• a binary tree is empty;
• or it consists of
• a node (the root) that stores an element
• a binary tree, called the left subtree of T
• a binary tree, called the right subtree of T
 Binary tree interface
• left(v)
• right(v)
• hasLeft(v)
• hasRight(v)
• + isInternal(v), is External(v), isRoot(v), size(), isEmpty()

19
 Properties of binary trees
 In a binary tree
• level 0 has <= 1 node
d=0
• level 1 has <= 2 nodes
• level 2 has <= 4 nodes
• ...
d=1
• level i has <= 2^i nodes
d=2

d=3

 Proposition: Let T be a binary tree with n nodes and height h. Then

• h+1 <= n <= 2 h+1 -1

• lg(n+1) - 1 <= h <= n-1

20
 Binary tree implementation
 use a linked-list structure; each node points to its left and right children ; the tree
class stores the root node and the size of the tree

 implement the following functions:

BTreeNode: parent
• left(v)
• right(v)
data
• hasLeft(v)
• hasRight(v) left right
• isInternal(v)
• is External(v)
• isRoot(v)
• size()
• isEmpty()
• also
• insertLeft(v,e)
• insertRight(v,e)
• remove(e)
• addRoot(e) 21
 Binary tree operations
 insertLeft(v,e):
• create and return a new node w storing element e, add w as the left child of v
• an error occurs if v already has a left child

 insertRight(v,e)

 remove(v):
• remove node v, replace it with its child, if any, and return the element stored at v
• an error occurs if v has 2 children

 addRoot(e):
• create and return a new node r storing element e and make r the root of the tree;
• an error occurs if the tree is not empty

 attach(v,T1, T2):
• attach T1 and T2 respectively as the left and right subtrees of the external node v
• an error occurs if v is not external
22
 Performance
 all O(1)
• left(v)
• right(v)
• hasLeft(v)
• hasRight(v)
• isInternal(v)
• is External(v)
• isRoot(v)
• size()
• isEmpty()
• addRoot(e)
• insertLeft(v,e)
• insertRight(v,e)
• remove(e)

23
 Binary tree traversals

 Binary tree computations often involve traversals

• pre-order: root left right

• post-order: left right root
 additional traversal for binary trees
• in-order: left root right
• visit the nodes from left to right

 Exercise:
• write methods to implement each traversal on binary trees

24
 Application: Tree drawing
 Come up with a solution to “draw” a binary tree in the following way. Essentially, we
need to assign coordinate x and y to each node.
• node v in the tree
• x(v) = ?
• y(v) = ?

0
1
2
3
4
0 1 2 3 4 5 6 7

25
 Application: Tree drawing
 We can use an in-order traversal and assign coordinate x and y of each node in the
following way:

• x(v) is the number of nodes visited before v in the in-order traversal of v

• y(v) is the depth of v

0
1
2
3
4
0 1 2 3 4 5 6 7

26
 Binary tree searching
 write search(v, k)
• search for element k in the subtree rooted at v
• return the node that contains k
• return null if not found

 performance
• ?

27
 Binary Search Trees (BST)
 Motivation:
• want a structure that can search fast
• arrays: search fast, updates slow
• linked lists: search slow, updates fast
 Intuition:
• tree combines the advantages of arrays and linked lists

 Definition:
• a BST is a binary tree with the following “search” property

– for any node v v allows to search efficiently

k

T1 T2

all nodes in T1<= k all node in T2 > k 28

 v
BST k
T1 T2
 Example

<= k >k

29
 Sorting a BST
 Print the elements in the BST in sorted order

30
 Sorting a BST
 Print the elements in the BST in sorted order.

//print the elements in tree of v in order

sort(BSTNode v)
if (v == null) return;
sort(v.left());
 in-order traversal: left -node-right print v.getData();
sort(v.right());
 Analysis: O(n)
31
Searching in a BST

32
 Searching in a BST
//return the node w such that w.getData() == k or null if such a node
//does not exist
BSTNode search (v, k) {
if (v == null) return null;
if (v.getData() == k) return v;
if (k < v.getData()) return search(v.left(), k);
else return search(v.right(), k)
}

 Analysis:
• search traverses (only) a path down from the root
• does NOT traverse the entire tree
• O(depth of result node) = O(h), where h is the height of the tree
33
 Inserting in a BST
 insert 25

34
 Inserting in a BST
 insert 25
• There is only one place where 25 can go
25

 //create and insert node with key k in the right place

 void insert (v, k) {
//this can only happen if inserting in an empty tree
if (v == null) return new BSTNode(k);
if (k <= v.getData()) {
if (v.left() == null) {
//insert node as left child of v
u = new BSTNode(k);
v.setLeft(u);
} else {
return insert(v.left(), k);
}
} else //if (v.getData() > k) {
...
}
} 35
 Inserting in a BST
 Analysis:
• similar with searching
• traverses a path from the root to the inserted node
• O(depth of inserted node)
• this is O(h), where h is the height of the tree

36
 Deleting in a BST
 delete 87
 delete 21
 delete 90

 case 1: delete a leaf x

• if x is left of its parent, set parent(x).left = null
• else set parent(x).right = null

 case 2: delete a node with one child

• link parent(x) to the child of x

 case 2: delete a node with 2 children

• ?? 37
 Deleting in a BST
 delete 90

 copy in u 94 and delete 94

• the left-most child of right(x) node has <=1 child
 or
 copy in u 87 and delete 87
• the right-most child of left(x)
node has <=1 child

38
 Deleting in a BST

 Analysis:
• traverses a path from the root to the deleted node
• and sometimes from the deleted node to its left-most child
• this is O(h), where h is the height of the tree

39
 BST performance
 Because of search property, all operations follow one root-leaf path
• insert: O(h)
• delete: O(h)
• search: O(h)

 We know that in a tree of n nodes

• h >= lg (n+1) - 1
• h <= n-1

 So in the worst case h is O(n)

• BST insert, search, delete: O(n)
• just like linked lists/arrays

40
 BST performance
 worst-case scenario
• start with an empty tree
• insert 1
• insert 2
• insert 3
• insert 4
• ...
• insert n

 it is possible to maintain that the height of the tree is Theta(lg n) at all times
• by adding additional constraints
• perform rotations during insert and delete to maintain these constraints

 Balanced BSTs: h is Theta(lg n)

• Red-Black trees
• AVL trees
• 2-3-4 trees
• B-trees
 41
to find out more.... take csci231 ( Algorithms)