ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Binary Trees
Atul Gupta
Binary Tree Traversals
Depth-First Traversals
Pre-order
Post-order
In-order
Breadth First Traversals
Note that
1. these traversals use stacks/queues as auxiliary data structures
2. It is not simple to determine predecessor/successor for an order
traversal
3. Here we demonstrate how to perform traversal without
stacks/queues and to get the predecessor/successor for an order
traversal
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Binary Trees
An observation: Majority of pointers in a
binary tree are NULL
e.g. a binary tree with n nodes has n+1 NULL
pointers, and they are wasted
Treaded Binary Tree
"A binary tree is threaded by making all right
child pointers that would normally be null point
to the inorder successor of the node, and all left
child pointers that would normally be null point
to the inorder predecessor of the node
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Binary Tree
makes it possible to traverse the values in the
binary tree via a linear traversal that is more
rapid than a recursive in-order traversal
to discover the parent of a node from a
threaded binary tree
Threaded Binary Trees: Classification
Based on left/right
threaded
Left-threaded
Right-threaded
Fully Threaded
Based on traversal
order
Preorder threaded
In-order threaded
Post-order threaded
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Normal Binary Tree vs. Threaded
Binary Tree
Normal Binary Tree
Left
data
Right
struct BinaryTree {
struct BinaryTree *left;
int data;
struct BinaryTree *right;
}
Threaded Binary Tree
Left
LTag
data
RTag
Right
struct ThreadedBinaryTree {
struct ThreadedBinaryTree *left;
int Ltag;
int data;
int Rtag;
struct ThreadedBinaryTree *right;
}
Meaning
LTag == 0
Left points to in-order predecessor
LTag == 1
Left point to left child
RTag == 0
Left points to in-order successor
RTag == 1
Right points to right child
Non recursive Inorder traversal for a
Threaded Binary Tree
1. curr-node node leftmost (root)
2. While (curr_node != Null)
a. print (curr_node)
b. If (curr_node.RTag == 0) then
curr_node curr_node.right
go to step 2.
c. else
curr_node leftmost(curr_node.right )
go to step 2.
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversal
Output
1
6
8
3
1
11
9
13
Start at leftmost node, print it
9
Threaded Tree Traversal
Output
1
3
6
8
3
1
11
9
13
Follow thread to right, print node
10
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversal
Output
1
3
5
6
8
3
1
11
9
13
Follow link to right, go to leftmost
node and print
11
Threaded Tree Traversal
Output
1
3
5
6
6
8
3
1
11
9
13
Follow thread to right, print node
Amir Kamil
8/8/02
12
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversal
Output
1
3
5
6
7
6
8
3
1
11
9
13
Follow link to right, go to
leftmost node and print
Amir Kamil
8/8/02
13
Threaded Tree Traversal
Output
1
3
5
6
7
8
6
8
3
1
11
9
13
Follow thread to right, print node
14
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversal
Output
1
3
5
6
7
8
9
6
8
3
1
11
9
13
Follow link to right, go to
leftmost node and print
15
Threaded Tree Traversal
6
8
3
1
11
9
13
Output
1
3
5
6
7
8
9
11
Follow thread to right, print node
16
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversal
6
8
3
1
11
9
13
Output
1
3
5
6
7
8
9
11
13
Follow link to right, go to
leftmost node and print
17
Threaded Tree Traversal Algorithm
(In-Order)
struct Node* leftMost(struct Node** n) {
struct Node* ans = n;
if (ans == null) {
return null;
}
while (ansleft != null) {
ans = ans left;
}
return ans;
}
void inOrder(struct Node* n) {
struct Node* cur = leftmost(n);
while (cur != null) {
print(cur);
if not (cur RTag) {
cur = cur right;
} else {
cur = leftmost(cur right);
}
}
}
18
ES 103: Data Structures and Algorithms 2012
Instructor Dr Atul Gupta
Threaded Tree Traversals
Similarly can also be employed to pre-order
and post-order traversals
Summary: Threaded Binary Tree
A useful idea to utilize the otherwise wasted links to store
meaningful information
Can be threaded using any order traversal
Useful in applications which require
Information about predecessor and successor nodes
For efficient order traversals
10
