Data Structures and Algorithms
A.Baskar
BITS Pilani, K. K. Birla Goa Campus
16 February 2018
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 1 / 10
Recap
Non linear data structures: Trees
Tree ADT
Heap Sort
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 2 / 10
Tree Abstract Data Type
Tree ADT stores elements at nodes
element(v) returns the object stored at the node v O(1)
size(), root(), parent(v)O(1)
children(v) O(cv )
isInternal(v), isExternal(v), IsRoot(v) O(1)
elements(), positions() O(n)
swapElements(v,w), replaceElements(v,e) O(1)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 3 / 10
Outline
Dictionary ADT
Binary Search Tree
AVL Tree
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 4 / 10
Dictionary Abstract Data Type
Store items (k,e)
Search(k), insert (k,e), remove(k)
size(), isEmpty()
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 5 / 10
Dictionary Abstract Data Type
Store items (k,e)
Search(k), insert (k,e), remove(k)
size(), isEmpty()
Ordered Dictionary ADT: maximum(), minimum(), successor(k),
predecessor(k)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 5 / 10
Implementations
Logfile, Direct Address table, Hash table
Either they don’t support min, max, succ, pred operations
Or they are not efficient
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 6 / 10
Implementations
Logfile, Direct Address table, Hash table
Either they don’t support min, max, succ, pred operations
Or they are not efficient
Solution is to store data in a non-linear data structures
Store such a way that it can support all these operations
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 6 / 10
Binary Search Tree
Store items in the internal nodes of binary tree
No items in the external nodes
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 7 / 10
Binary Search Tree
Store items in the internal nodes of binary tree
No items in the external nodes
Key of a node is always less than every keys of the nodes in the
left subtree
Key of a node is always greater than every keys of the nodes in
the right subtree
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 7 / 10
Binary Search Tree
Store items in the internal nodes of binary tree
No items in the external nodes
Key of a node is always less than every keys of the nodes in the
left subtree
Key of a node is always greater than every keys of the nodes in
the right subtree
All keys are distinct
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 7 / 10
Performance
search, insert, delete will take O(h) time where h is height of the
tree
But in the worst case h can be O(n)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 8 / 10
Performance
search, insert, delete will take O(h) time where h is height of the
tree
But in the worst case h can be O(n)
The is because the tree might be unbalanced
We have seen one solution: AVL trees
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 8 / 10
AVL Trees
Balance the height between left subtree and right subtree
Height of the tree is O(log n)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 9 / 10
AVL Trees
Balance the height between left subtree and right subtree
Height of the tree is O(log n)
All the operations can be done in O(log n)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 9 / 10
Height Balance Property
For every internal node of a binary search tree, the heights of
the children of v can differ by at most 1
It is a property which characterizes the structure of BST
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 10 / 10
Height Balance Property
For every internal node of a binary search tree, the heights of
the children of v can differ by at most 1
It is a property which characterizes the structure of BST
Any BST which satisfies height balance property is an AVL Tree
(Adelson-Velskii and Landis)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 10 / 10
Height Balance Property
For every internal node of a binary search tree, the heights of
the children of v can differ by at most 1
It is a property which characterizes the structure of BST
Any BST which satisfies height balance property is an AVL Tree
(Adelson-Velskii and Landis)
Subtree of an AVL tree is an AVL tree
Height of AVL tree stroing is n keys is O(log n)
A.Baskar (BITS-Pilani) Data Structures and Algorithms 16 February 2018 10 / 10
