Scribd
Upload
76 views

Week5 PDF

This document discusses Red-Black trees, which are self-balancing binary search trees where each node is colored red or black. Red-Black trees maintain the property that the path from any node to any leaf is no more than twice as long as any other, allowing operations like search, insertion, and deletion to occur in O(log n) time through rotations and recoloring nodes. The document covers the properties, rotations, insertion, and deletion algorithms for Red-Black trees through examples and analysis.

Uploaded by

Sidharth Malik
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
76 views

Week5 PDF

This document discusses Red-Black trees, which are self-balancing binary search trees where each node is colored red or black. Red-Black trees maintain the property that the path from any node to any leaf is no more than twice as long as any other, allowing operations like search, insertion, and deletion to occur in O(log n) time through rotations and recoloring nodes. The document covers the properties, rotations, insertion, and deletion algorithms for Red-Black trees through examples and analysis.

Uploaded by

Sidharth Malik
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 38

CSE - 5311 Advanced Algorithms

Instructor : Nomaan Mufti

Submitted by
Raja Rajeshwari Anugula & Srujana Tiruveedhi
Contents
 RED – BLACK Trees
 Properties
 Rotations
 Insertion
 Deletion
RED BLACK TREES
 A special class of binary trees that avoids the worst-case
behavior of O(n) that we can see in “plain” binary search
trees
 Balanced: height is O(lg n), where n is the number of
nodes
 Operations will take O(lg n) time in the worst case.
 Consist of one extra bit of storage per node; its color,
which can be either RED or BLACK.
 Each node of the tree contains fields: color, key, left,
right, parent.
 Red-Black Trees ensure that longest path is no more
than twice as long as any other path
RB Tree Properties
RB Trees
Points to remember:
 All leaves are empty (nil) and colored black.
 We may use a single sentinel, T.nil, for all the leaves of
red-black tree T
 T.nil.color is black
 Root’s parent is also nil
 All other attributes of binary search trees are inherited
by red-black trees (key, left, right, and p)
 key of T.nil is not utilized

Sentinel - A soldier or guard whose job is to stand and keep watch


A node with a special value
RB Trees
More points to remember:
 A RB tree with n internal nodes has a height of almost
2lg(n+1)
 Maximum path length is O(lg n)
 Finding an element is real quick in RB trees, i.e., it
takes O(lg n) time
 Insertion and Deletion take O(lg n) time
RB Height
 Height of a node is the number of edges in a longest
path to a leaf
 Black-height of a node x, bh(x), is the number of black
nodes (including T.nil) on the path from x to leaf, not
counting x
(see also Property 5 above)
RB Height Claim 1
Any node with height h has black-height ≥ ℎ/2

Proof: By Property 4, from the node to a leaf ≤ ℎ/2 nodes


are red
Hence more than ℎ/2 are black nodes to the leaf
RB Height Claim 2
A subtree rooted at any node 𝑥 contains
≥ 2bh 𝑥 −1 internal nodes
Proof: By induction on the height of 𝑥
 Height of 𝑥 = 0 is a leaf ⟹ bh 𝑥 = 0. The
subtree rooted at 𝑥 has zero internal nodes
⇒ 20 − 1 = 0
 Let the height of 𝑥 be ℎ and bh 𝑥 = 𝑏. Any
child of 𝑥 has height ℎ − 1 and black height is
𝑏 (if the child is red) or 𝑏 − 1 (if the child is
black)
RB Height Claim 2
A subtree rooted at any node 𝑥 contains
≥ 2bh 𝑥 −1 internal nodes
Proof: By induction on the height of 𝑥 (continues)
 By the inductive hypothesis, each child has
≥ 2𝑏−1 − 1 internal nodes
 Thus, the subtree rooted at 𝑥 contains
≥ 2 2𝑏−1 − 1 + 1 = 2𝑏 − 1 internal nodes.
(two times for children and +1 for 𝑥 itself)
 Thus, 𝑥 contains ≥ 2bh 𝑥 −1
RB Height, Lemma
A RB tree with 𝑛 internal nodes has height
ℎ ≤ 2lg(𝑛 + 1)
Proof: Let ℎ and 𝑏 be the height and black-height
of the root, respectively. By the above two claims
𝑛 ≥ 2𝑏 − 1 ≥ 2ℎ/2 − 1
Adding 1 to RHS and LHS and then taking logs we
get lg(𝑛 + 1) ≥ ℎ/2, which provides the proof
Operations on RB Trees
Rotations
 Insertion and Deletion modify the tree, the
result may violate the properties of red
black trees. To restore these properties
rotations are conducted
 We can have either LEFT rotation or
RIGHT rotation by which we must change
colors of some of the nodes in the tree
and also change the pointer structure.
Rotations
Rotations, Example
Rotations
 When we do a LEFT
rotation on a node x
we assume that its
right child y is not nil,
i.e x may be any node in
the tree whose right child
is not Nil

 It makes y the new root of


the sub tree, with x as
y’s left child and y’s
left child as x’s right
child
Insertion

 RB-INSERT ends by coloring


the new node z red.
 Then it calls RB-INSERT-
FIXUP because we could
have violated a red-black
property
Which Properties are Violated?
C

A D

B
a b c
Removing the Violations
Insertion, Discussion
Insertion, Case 1: y is red
Insertion, Case 2 & 3: y is black
 z is a right child

 z is a left child
RED BLACK TREES
Example:

11 Apply 11
Case 1b
2 14 2 14

7 15 7 15
1 1

5 8 8
Insert 5
Node
4
4 4
RED BLACK TREES
Example:

11 Apply 11
Case 2
2 14 7 14

7 15 8 15
1 2

5 8
1 5

4
4
RED BLACK TREES
Example:

11 Apply 7
Case 3
7 14 2 11

8 15
2 5 8 14
1

4 15
5
1

4
Insertion, Analysis
Deletion
 RB-DELETE calls a RB-
TRANSPLANT procedure.
 Then it calls RB-DELTE-
FIXUP because we could
have violated a red-black
property
 TREE-MINIMUM returns the
minimum key (the minimum
key of a binary search tree is
located at the leftmost node)
 y or x is the successor of z
RB-DELETE: Discussion
RB-Transplant RB-TRANSPLANT replaces the
subtree rooted at u by the
subtree rooted at v:
 Changes u’s parent to v’s
parent (unless u is the root, in
which case it makes v the
root)
 u’s parent gets v as either its
left or right child, depending
on whether u was a left or
right child
 Doesn’t update v.left or v.right
RB-DELETE: Violations
RB-DELETE-FIXUP
RB-DELETE-FIXUP, The Idea

Note:
When a black node is deleted and replaced by a black child, the
child is marked as doubly black
When a black node is deleted and replaced by a red child (or
vice verse), the child is marked as red & black node
RB-DELETE-FIXUP, Case 1
RB-DELETE-FIXUP, Case 2
RB-DELETE-FIXUP, Case 3
RB-DELETE-FIXUP, Case 4
RB-DELETE, Analysis
HW
Exercises
 13.1-3, 13.1-4
 13.2-4
 13.3-3, 13.3-4
 13.4-6

You might also like