Scribd
Upload
1 views

Lecture 1.2_correctness_binary_search

The document provides an overview of the binary search algorithm, including its implementation, input/output definitions, and proof of correctness through loop invariants. It discusses the time complexity of the algorithm, detailing both worst-case and best-case scenarios, concluding that the worst-case time complexity is O(log n) and the best-case is O(1). The author, Neelima Gupta, presents this material as part of a course on the design and analysis of algorithms.

Uploaded by

rdxsingh01
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
1 views

Lecture 1.2_correctness_binary_search

The document provides an overview of the binary search algorithm, including its implementation, input/output definitions, and proof of correctness through loop invariants. It discusses the time complexity of the algorithm, detailing both worst-case and best-case scenarios, concluding that the worst-case time complexity is O(log n) and the best-case is O(1). The author, Neelima Gupta, presents this material as part of a course on the design and analysis of algorithms.

Uploaded by

rdxsingh01
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

MCSC 101: Design and Analysis of Algorithms

Neelima Gupta

[email protected]

January 4, 2021

1/12
Binary Search
Binary-Search(A[], n, key )
left 1, right n
while left  right do
mid (left+right)/2
if (A[mid]=key) then
return mid
end
if (A[mid] < key) then
left mid+1
end
else
right mid-1
end
end
return 0

2/12
Defining the I/O of Binary Search

I Input : Given an array L containing n items (n 0) ordered


such that A[1]  A[2]  A[3]  . . .  A[n] and given x
I Output : The binary search algorithm terminates with
index mid = an occurrence of x in A, if found and,
index mid= 0 otherwise.

Binary search can only be applied if the array to be searched is


already sorted

3/12
4/12
5/12
Proof of correctness: The Loop Invariant

Let r be the maximum number of times the loop starting from line
3 runs. For 1  k  r , let fk and lk be the values of first and last
respectively when the control reaches the while statement for the
k th time.
I Loop Invariant: For 1  k  r , H(k): when the control
reaches the while statement for the k th time A[i] 6= x for
every i < fk and for every i > lk .
I We will prove the loop invariant by induction.
I Base Case: k = 1. fk = 1 and lk = n. Hence the claim is
vacuously true.

6/12
Inductive Step: Suppose H(k) is true. We will prove that H(k + 1)
is true. In the k th iteration, we have three possibilities:

I Case 1: x = A[mk ] . . . .not possible else we wouldn’t enter the


loop (k + 1)th time.

7/12
I Case 2. x < A[mk ].

Since the array is sorted, we have


x < A[mk ]  A[i] 8i = mk + 1 . . . lk .
i.e. x 6= A[i] 8i = mk . . . lk .

8/12
I Case 3. x > A[mk ] .

Since the array is sorted, we have


x > A[mk ] A[i] 8i = fk . . . mk 1.
i.e. x 6= A[i] 8i = fk . . . mk .

9/12
Correctness of the algorithm assuming the loop invariant

Suppose the test condition is executed t times.


I Case 1: ft  lt . We exit from the loop because A[mt ] = x.
Thus, index mid is a position of occurrence of x as index mid
= mt
I Case 2: If ft > lt , we exit the while statement and return 0.
We will next argue that the element is not present in the array
in this case:
By loop invariant hypothesis, A[i] 6= x for i < ft and for
i > lt . the claim follows as ft > lt .
Hence assuming that the statement H(k) is correct the algorithm
works correctly.

10/12
Correctness of the algorithm assuming the loop invariant

By loop invariant hypothesis, A[i] 6= x for i < ft and for i > lt . the
claim follows as ft > lt .

11/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?
W (n) = W (n/2) + 2

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?
W (n) = W (n/2) + 2
W (n) =?

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?
W (n) = W (n/2) + 2
W (n) =? W (n) = log n
Let B(n) be the number of comparisons performed by the binary
search algorithm in best case. Then,

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?
W (n) = W (n/2) + 2
W (n) =? W (n) = log n
Let B(n) be the number of comparisons performed by the binary
search algorithm in best case. Then,
B(n) =?

12/12
Time Complexity

Let W (n) be the number of comparisons performed by the binary


search algorithm in worst case. Then,
W (n) =?
W (n) = W (n/2)+?
W (n) = W (n/2) + 2
W (n) =? W (n) = log n
Let B(n) be the number of comparisons performed by the binary
search algorithm in best case. Then,
B(n) =?
B(n) = 1

12/12

You might also like