6 1 Algorithms (323)

1 Algorithms (323) top

1.1 Algorithm Design(8) top

1.1.1 Algorithm Design: CMI2010 - 6 top [Link]

You are given a list of positive integers along with a sequence of operations from the set {∗, +}
.You construct expressions from these two lists so that:

The numbers in the expression are drawn from the first list, without repetition and without
altering their order.
All the operators in the second list are used in the expression, in the same order.

For example, if the two lists are [1, 3, 2, 1, 4] and [' ∗ ', ' + '] the set of possible expressions you
can form are
1 ∗ 3 + 2, 1 ∗ 3 + 1, 1 ∗ 3 + 4, 1 ∗ 2 + 1, 1 ∗ 2 + 4, … , 2 ∗ 1 + 4, 1 ∗ 3 + 2, 1 ∗ 3 + 1, 1 ∗ 3 + 4, 1 ∗
For each expression, the value is computed by bracketing operators from the right. That is, the
expressions above are evaluated as

1 ∗ (3 + 2), 1 ∗ (3 + 1), 1 ∗ (3 + 4), 1 ∗ (2 + 1), 1 ∗ (2 + 4), … , 2 ∗ (1 + 4), 1 ∗ (3 + 2), 1 ∗ (3 +

The aim is to determine maximum value among these expressions. In this example, the maximum
value is 18, from the expression 3 ∗ 2 + 4, which is evaluated as 3 ∗ (2 + 4) = 3 ∗ 6 = 18.
You may assume that the length of the first list is more than the length of the second list.
Describe an algorithm to solve this problem.

algorithms descriptive cmi2010 algorithm-design

Answer

1.1.2 Algorithm Design: CMI2012-B-03a top [Link]

Let A be an array of n integers, sorted, so that A[1] ≤ A[2] ≤ … A[n]. Suppose you are given a
number x and you wish to find out if there are indices k and l such that A[k] + A[l] = x .

a. Design an O(n log n) time algorithm for this problem.

cmi2012 descriptive algorithms algorithm-design

Answer

1.1.3 Algorithm Design: CMI2012-B-03b top [Link]

Let A be an array of n integers, sorted so that A[1] ≤ A[2] ≤ … A[n]. Suppose you are given a
number x and you wish to find out if there exist indices k and l such that A[k] + A[l] = x .

a. Design an O(n) algorithm for this problem.

descriptive cmi2012 algorithms algorithm-design

Answer

1.1.4 Algorithm Design: GATE1992-8 top [Link]

Let T be a Depth First Tree of a undirected graph G. An array P indexed by the vertices of G is
given. P[V ] is the parent of vertex V , in T . Parent of the root is the root itself.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 7

Give a method for finding and printing the cycle formed if the edge (u, v) of G not in T (i.e.,
e ∈ G − T ) is now added to T .
Time taken by your method must be proportional to the length of the cycle.

Describe the algorithm in a PASCAL (C) – like language. Assume that the variables have been
suitably declared.
gate1992 algorithms descriptive algorithm-design

Answer

1.1.5 Algorithm Design: GATE2006-17 top [Link]

An element in an array X is called a leader if it is greater than all elements to the right of it in X.
The best algorithm to find all leaders in an array

A. solves it in linear time using a left to right pass of the array

B. solves it in linear time using a right to left pass of the array
C. solves it using divide and conquer in time Θ(n log n)
D. solves it in time Θ(n2 )

gate2006 algorithms normal algorithm-design

Answer

1.1.6 Algorithm Design: GATE2006-54 top [Link]

Given two arrays of numbers a1 , . . . , an and b1 , . . . , bn where each number is 0 or 1, the fastest
algorithm to find the largest span (i, j) such that ai + ai+1 + ⋯ + aj = bi + bi+1 + ⋯ + bj or
report that there is not such span,

A. Takes O(3n ) and Ω(2n ) time if hashing is permitted

B. Takes O(n3 ) and Ω(n2.5 ) time in the key comparison mode
C. Takes Θ(n) time and space
D. Takes O(√−n ) time only if the sum of the 2n elements is an even number

gate2006 algorithms normal algorithm-design time-complexity

Answer

1.1.7 Algorithm Design: ISI2015-CS-2a top [Link]

You are given two strings S and T , each of length α, consisting only of lower case English letters
(a, b, … , z). Propose an O(α)-time algorithm to decide whether S can be obtained by permuting
the symbols of T . For example, if S = algorithm, T = logarithm, your algorithm should
return Y ES ; but if S = trainee, T = retinaa, your algorithm should return NO.
descriptive isi2015 algorithms algorithm-design

Answer

1.1.8 Algorithm Design: TIFR2011-B-29 top [Link]

You are given ten rings numbered from 1 to 10, and three pegs labeled A, B, and C . Initially all the
rings are on peg A, arranged from top to bottom in ascending order of their numbers. The goal is to
move all the rings to peg B in the minimum number of moves obeying the following constraints:

© Copyright GATE Overflow. All rights reserved.

8 1 Algorithms (323)

i. In one move, only one ring can be moved.

ii. A ring can only be moved from the top of its peg to the top of a new peg.
iii. At no point can a ring be placed on top of another ring with a lower number.

How many moves are required?

A. 501
B. 1023
C. 2011
D. 10079
E. None of the above.

tifr2011 algorithms algorithm-design

Answer

Answers: Algorithm Design

1.1.1 Algorithm Design: CMI2010 - 6 top [Link]


Selected Answer

Suppose the first list has m numbers and the second list has n operators, here (m > n).
So, at each expression we will have (n + 1) numbers.
Note that maximum value of the expression occurs when (n + 1) maximum numbers out of m
numbers in the first list will be present in the expression.

So, our goal is to find (n + 1) maximum numbers out of m numbers in the first list.
Now, finding the maximum number out of m numbers in the first list can be done in O(n) time.
Build a max heap and extract the root node. Building a max heap takes O(n) time.
Now, the remaining n maximum numbers can be determined in O(logn) time. (Call max-heapify
on the max-heap and delete the root node).

Note that here n is a constant number of operators in the second list.

(n + 1) maximum numbers out of m numbers in the first list can be done in
So, to find
n + O(logn) time.
So, to determine the maximum value among all the expressions can be done in n + O(logn)
time..

 1 votes -- Kushagra Chatterjee (8.2k points)

1.1.2 Algorithm Design: CMI2012-B-03a top [Link]


Selected Answer

a. Pseudo code for time complexity O(n log n) :

for(i=1; i<=n; i++)

{

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 9

if (binary_search( x- A[i] ) // O(logn)

return true;
}
return false;

b. Pseudo code for time complexity O(n) :

for( k=1, l=n; k<l ; )
{
temp =A[k] + A[l];
if (temp== x)
return true;
else if(temp > x)
l--;
else
k++;
}
return false;

 2 votes -- Dhananjay Kumar Sharma (25.7k points)

1.1.3 Algorithm Design: CMI2012-B-03b top [Link]


Selected Answer

Here is the algorithm, which returns True if there is a number x present such that (
A[k] + A[l] == x ) else returns False.
// Consider this algorithm is called as
// AlgorithmCheck(A,1,size,x);
// Where A is the sorted array

Bool AlgorithmCheck(A, L, K, x){

while(L<K){
if(A[L]+A[K] == x)
return true;
else if(A[L]+A[K] < x)
L++;
else
K--;
}
return false;
}

This will take only O(n) time to do its work.

 15 votes -- Muktinath Vishwakarma (35.4k points)

1.1.4 Algorithm Design: GATE1992-8 top [Link]


Selected Answer

Union-Find Algorithm can be used to find the cycle.

© Copyright GATE Overflow. All rights reserved.

10 1 Algorithms (323)

Ref: [Link]

 10 votes -- Rajarshi Sarkar (34.1k points)

1.1.5 Algorithm Design: GATE2006-17 top [Link]


Selected Answer

Option B) We can move from right to left, while keeping a note of the maximum element so far
(let's call it current_max).

Starting from the rightmost element, we initialize our current_max with it, since the rightmost
element will always be a leader.
Moving from right to left, if an element x is greater than our current_max, then x is also a
leader. Add this element to list of leaders (or simply print it). Set current_max to x and carry-
on leftward.

Time Complexity would be Θ(n).

 33 votes -- Madhur Rawat (2.3k points)

1.1.6 Algorithm Design: GATE2006-54 top [Link]


Selected Answer

Answer is (C). Following algorithm would do.

Since array is binary, the max sum will go until n and so the sum difference of the two arrays can
vary between −n and n. We use array start to keep the starting index of each possible sum
(hence of size 2n + 1) and array end to keep the ending index (these two arrays work like hash
tables and since we have only 2n + 1 possible keys, we can do a perfect hashing). So, our
required solution will be max(end[i] − start[i]) provided both are assigned values.

The algorithm works as follows:

1. Initialize diff array to contain the difference of sum of elements of array a and b. i.e.,
diff[i] = ∑ni=0 a[i] − b[i] .
2. Now diff[i] can have values from −n to n which gives 2n + 1 possible values and the first
occurrence of a diff value marks the beginning of a span and the last occurrence marks the
end. We use start and end array for storing these two positions for the 2n + 1 possible
values.
3. Now, the largest value of end[i] − start[i] for any i, will be the largest span and the start of it
will be start[i] + 1, and end will be end[i]. If the span is starting from first position itself
(arrays a and b have same first elements), then it will start from start[i] itself.

#include <stdio.h>
#define size 100 //assume n is less than 100
int main()
{
int n, a[size], b[size];
int start[2*size+1], end[2*size+1];
int sum1 = 0, sum2 = 0, i;
int diff[size];
printf("Enter n: ");
scanf("%d", &n);
for(i = 0; i < n; i++)
{
printf("Enter a[%d]: ", i);
scanf("%d", &a[i]);
}

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 11

for(i = 0; i < n; i++)

{
printf("Enter b[%d]: ", i);
scanf("%d", &b[i]);
}
for(i = 0; i < n; i++)
{
if(a[i]) sum1++;
if(b[i]) sum2++;
diff[i] = sum1 - sum2;
}
for(i = 0; i < 2*n; i++)
start[i] = -1,end[i] = -1;
start[n] = end[n] = 0;
//initially sum is 0 at the beginning of array and
//the first n-1 elements of start and end are used
//if sum of A till ith element is less than sum of B till ith element
for(i=0; i < n; i++)
{
if(start[diff[i] + n] == -1)//interested only in the first occurrence of diff[i]
start[diff[i] + n] = i;
end[diff[i] + n] = i;//interested in the last occurrence of diff[i]
}
int max = -1;
int savei = -1; //savei is for storing the sum having the largest span
for(i = 0; i < 2*n; i++)
{
if(start[i] > -1 && (end[i] - start[i] > max))
{
max = end[i] - start[i];
savei = i;
}
}
if(savei >= 0)
{
printf("The largest span is from %d to %d\n", start[savei]+(savei != n), end[savei
//when sum zero is having the largest span, span starts from first element itself.
//Else, the span starts from the next element from which the span does not change
}
else
{
printf("No span\n");
}
}

 20 votes -- Arjun Suresh (350k points)

1.1.7 Algorithm Design: ISI2015-CS-2a top [Link]


Selected Answer

1) Take an array of 26 elements where 'a' corresponds to index 0,....,z corresponds to index 25,
initially initialized to all 0s

2) For the first string, increment the corresponding positions of the array by 1 corresponding to
every character as the string is traversed.

3) For the second string, decrement the corresponding positions of the array by 1 corresponding to
every character as the string is traversed.

4) If all the elements of the array are 0 finally, then output YES, else NO.
 0 votes -- Prasita Mukherjee (429 points)

1.1.8 Algorithm Design: TIFR2011-B-29 top [Link]


Selected Answer

I think its Tower of Hanoi problem.

Therefore, total number of function call 2n − 1 = 1023.
∴ Option B.

© Copyright GATE Overflow. All rights reserved.

12 1 Algorithms (323)

 8 votes -- Umang Raman (15.7k points)

1.2 Algorithm Design Techniques(5) top

1.2.1 Algorithm Design Techniques: GATE1990-2-vii top

[Link]

Match the pairs in the following questions:

(a) Strassen's matrix multiplication algorithm (p) Greedy method

(b) Kruskal's minimum spanning tree algorithm (q) Dynamic programming
(c) Biconnected components algorithm (r) Divide and Conquer
(d) Floyd's shortest path algorithm (s) Depth first search
gate1990 match-the-following algorithms algorithm-design-techniques

Answer

1.2.2 Algorithm Design Techniques: GATE1997-1.5 top [Link]

The correct matching for the following pairs is

A. All pairs shortest path 1. Greedy

B. Quick Sort 2. Depth-First Search

C. Minimum weight spanning tree 3. Dynamic Programming

D. Connected Components 4. Divide and Conquer

A. A-2 B-4 C-1 D-3

B. A-3 B-4 C-1 D-2
C. A-3 B-4 C-2 D-1
D. A-4 B-1 C-2 D-3
gate1997 algorithms normal algorithm-design-techniques

Answer

1.2.3 Algorithm Design Techniques: GATE2015-1-6 top[Link]

Match the following:

P. Prim's algorithm for minimum spanning tree i. Backtracking

Q. Floyd-Warshall algorithm for all pairs shortest path ii. Greedy method
R. Merge sort iii. Dynamic programming
S. Hamiltonian circuit iv. Divide and conquer

A. P-iii, Q-ii, R-iv, S-i

B. P-i, Q-ii, R-iv, S-iii
C. P-ii, Q-iii, R-iv, S-i
D. P-ii, Q-i, R-iii, S-iv

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 13

gate2015-1 algorithms normal algorithm-design-techniques

Answer

1.2.4 Algorithm Design Techniques: GATE2015-2-36 top

[Link]

Given below are some algorithms, and some algorithm design paradigms.

i. Divide and
1. Dijkstra's Shortest Path
Conquer
ii. Dynamic
2. Floyd-Warshall algorithm to compute all pairs shortest path
Programming
3. Binary search on a sorted array iii. Greedy design
iv. Depth-first
4. Backtracking search on a graph
search
v. Breadth-first
search

Match the above algorithms on the left to the corresponding design paradigm they follow.

A. 1-i, 2-iii, 3-i, 4-v

B. 1-iii, 2-iii, 3-i, 4-v
C. 1-iii, 2-ii, 3-i, 4-iv
D. 1-iii, 2-ii, 3-i, 4-v
gate2015-2 algorithms easy algorithm-design-techniques

Answer

1.2.5 Algorithm Design Techniques: GATE2017-1-05 top

[Link]

Consider the following table:

Algorithms Design Paradigms

P. Kruskal i. Divide and Conquer

Q. Quicksort ii. Greedy

R. Floyd-Warshall iii. Dynamic Programming

Match the algorithms to the design paradigms they are based on.

A. (P) ↔ (ii), (Q) ↔ (iii), (R) ↔ (i)

B. (P) ↔ (iii), (Q) ↔ (i), (R) ↔ (ii)
C. (P) ↔ (ii), (Q) ↔ (i), (R) ↔ (iii)
D. (P) ↔ (i), (Q) ↔ (ii), (R) ↔ (iii)

gate2017-1 algorithms algorithm-design-techniques

Answer

Answers: Algorithm Design Techniques

© Copyright GATE Overflow. All rights reserved.

14 1 Algorithms (323)

1.2.1 Algorithm Design Techniques: GATE1990-2-vii top

[Link]

Selected Answer

(a) Strassen's matrix multiplication algorithm - (r) Divide and Conquer

(b) Kruskal's minimum spanning tree algorithm - (p) Greedy method
(c) Biconnected components algorithm - (s) Depth first search
(d) Floyd's shortest path algorithm - (q) Dynamic programming
 6 votes -- vishwa ratna (2.5k points)

1.2.2 Algorithm Design Techniques: GATE1997-1.5 top [Link]


Selected Answer

Answer : B) A-3 B-4 C-1 D-2

(A) All pairs shortest (3) Dynamic
path Programming
(4) Divide
(B) Quick Sort
and Conquer
(C) Minimum weight
(1) Greedy
spanning tree
(D) Connected (2) Depth-First
Components Search

Reference : Read the Intro/Algo Sub-Heading.

[Link]
[Link]
[Link]
[Link]

 7 votes -- Siddharth Mahapatra (1.3k points)

1.2.3 Algorithm Design Techniques: GATE2015-1-6 top[Link]


Selected Answer

P. Prim - ii. Greedy

[Link]
Q. Floyd Warshall - iii. Dynamic
[Link]
R. Merge Sort - iv. Divide & Conquer
[Link]
S. Hamiltonian Circuit - i. Backtracking
[Link]

Option is C.

 9 votes -- Pronomita Dey (2.1k points)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 15

1.2.4 Algorithm Design Techniques: GATE2015-2-36 top

[Link]

Selected Answer

Answer: C

 17 votes -- Rajarshi Sarkar (34.1k points)

1.2.5 Algorithm Design Techniques: GATE2017-1-05 top

[Link]

Selected Answer

In Kruskal, in every iteration, an edge of the most minimum weight (greediest) possible is
selected and added to MST construction. Hence, greedy.

In Quick Sort, we partition the problem into subproblems, solve them and then combine. Hence, it
is Divide & Conquer.

Floyd-Warshall uses Dynamic programming.

Hence correct answer is : OPTION

(C).

 17 votes -- sriv_shubham (3.3k points)

1.3 Asymptotic Notations(22) top

1.3.1 Asymptotic Notations: GATE1994-1.23 top [Link]

Consider the following two functions:

n3 for 0 ≤ n ≤ 10, 000

g1 (n) = {
n2 for n ≥ 10, 000

g2 (n) = {
n for 0 ≤ n ≤ 100
n3 for n > 100
Which of the following is true?

A. g1 (n) is O(g2 (n))

B. g1 (n) is O(n3 )
C. g2 (n) is O(g1 (n))
D. g2 (n) is O(n)

gate1994 algorithms asymptotic-notations normal

Answer

1.3.2 Asymptotic Notations: GATE1996-1.11 top [Link]

Which of the following is false?

100n log n = O( n log n )

© Copyright GATE Overflow. All rights reserved.
16 1 Algorithms (323)

log n
A. 100n log n = O( n 100 )
−−−−
B. √log n = O(log log n)
C. If 0 < x < y then nx = O (ny )
D. 2n ≠ O (nk)

gate1996 algorithms asymptotic-notations normal

Answer

1.3.3 Asymptotic Notations: GATE1999-2.21 top [Link]

If T1 = O(1), give the correct matching for the following pairs:

(M) Tn = Tn−1
(U) Tn = O(n)
+n
(V)
(N) Tn = O(n log
Tn = Tn/2 + n
n)
(O) Tn = Tn/2
(W) T = O(n2 )
+ n log n
(P) Tn = Tn−1 (X) Tn = O(log2
+ log n n)
A. M-W, N-V, O-U, P-X
B. M-W, N-U, O-X, P-V
C. M-V, N-W, O-X, P-U
D. M-W, N-U, O-V, P-X
gate1999 algorithms recurrence asymptotic-notations normal

Answer

1.3.4 Asymptotic Notations: GATE2000-2.17 top [Link]

Consider the following functions

f(n) = 3n√n
g(n) = 2√nlog2 n
h(n) = n!
Which of the following is true?

A. h(n) is O(f(n))
B. h(n) is O(g(n))
C. g(n) is not O(f(n))
D. f(n) is O(g(n))

gate2000 algorithms asymptotic-notations normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 17

1.3.5 Asymptotic Notations: GATE2001-1.16 top [Link]

Let f(n) = n2 log n and g(n) = n(log n)10 be two positive functions of n. Which of the following
statements is correct?

A. f(n) = O(g(n)) and g(n) ≠ O(f(n))

B. g(n) = O(f(n)) and f(n) ≠ O(g(n))
C. f(n) ≠ O(g(n)) and g(n) ≠ O(f(n))
D. f(n) = O(g(n)) and g(n) = O(f(n))

gate2001 algorithms asymptotic-notations time-complexity normal

Answer

1.3.6 Asymptotic Notations: GATE2003-20 top [Link]

Consider the following three claims:

I. (n + k)m = Θ(nm ) where k and m are constants

II. 2n+1 = O(2n )
III. 22n+1 = O(2n )
Which of the following claims are correct?

A. I and II
B. I and III
C. II and III
D. I, II, and III

gate2003 algorithms asymptotic-notations normal

Answer

1.3.7 Asymptotic Notations: GATE2004-IT-55 top [Link]

Letf(n), g(n) and h(n) be functions defined for positive integers such that
f(n) = O(g(n)), g(n) ≠ O(f(n)), g(n) = O(h(n)), and h(n) = O(g(n)).
Which one of the following statements is FALSE?

A. f(n) + g(n) = O(h(n) + h(n))

B. f(n) = O(h(n))
C. h(n) ≠ O(f(n))
D. f(n)h(n) ≠ O(g(n)h(n))

gate2004-it algorithms asymptotic-notations normal

Answer

1.3.8 Asymptotic Notations: GATE2005-37 top [Link]

Suppose T (n) = 2T ( n2 ) + n, T (0) = T (1) = 1

Which one of the following is FALSE?

A. T (n) = O(n2 )
T (n) = Θ(n log n)
© Copyright GATE Overflow. All rights reserved.
18 1 Algorithms (323)

B. T (n) = Θ(n log n)

C. T (n) = Ω(n2 )
D. T (n) = O(n log n)

gate2005 algorithms asymptotic-notations recurrence normal

Answer

1.3.9 Asymptotic Notations: GATE2008-39 top [Link]

Consider the following functions:

f(n) = 2n
g(n) = n!
h(n) = nlog n
Which of the following statements about the asymptotic behavior of f(n), g(n) and h(n) is true?
A. f (n) = O (g (n)) ; g (n) = O (h (n))
B. f (n) = Ω (g (n)) ; g(n) = O (h (n))
C. g (n) = O (f (n)) ; h (n) = O (f (n))
D. h (n) = O (f (n)) ; g (n) = Ω (f (n))

gate2008 algorithms asymptotic-notations normal

Answer

1.3.10 Asymptotic Notations: GATE2008-IT-10 top [Link]

Arrange the following functions in increasing asymptotic order:

A. n1/3
B. en
C. n7/4
D. n log9 n
E. 1.0000001n
A. a, d, c, e, b
B. d, a, c, e, b
C. a, c, d, e, b
D. a, c, d, b, e

gate2008-it algorithms asymptotic-notations normal

Answer

1.3.11 Asymptotic Notations: GATE2011-37 top [Link]

Which of the given options provides the increasing order of asymptotic complexity of functions
f1 , f2 , f3
and f4 ?

f1 (n) = 2n
f2 (n) = n3/2
f3 (n) = n log2 n
f4 (n) = nlog2 n
f3 , f2 , f4 , f1
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 19

A. f3 , f2 , f4 , f1
B. f3 , f2 , f1 , f4
C. f2 , f3 , f1 , f4
D. f2 , f3 , f4 , f1

gate2011 algorithms asymptotic-notations normal

Answer

1.3.12 Asymptotic Notations: GATE2012-18 top [Link]

L e t W(n) and A(n) denote respectively, the worst case and average case running time of an
algorithm executed on an input of size n. Which of the following is ALWAYS TRUE?

A. A(n) = Ω(W(n))
B. A(n) = Θ(W(n))
C. A(n) = O(W(n))
D. A(n) = o(W(n))

gate2012 algorithms easy asymptotic-notations

Answer

1.3.13 Asymptotic Notations: GATE2015-3-4 top [Link]

n
Consider the equality ∑i3 = X and the following choices for X:
i=0

I. Θ(n4 )
II. Θ(n5 )
III. O(n5 )
IV. Ω(n3 )
The equality above remains correct if X is replaced by
A. Only I
B. Only II
C. I or III or IV but not II
D. II or III or IV but not I

gate2015-3 algorithms asymptotic-notations normal

Answer

1.3.14 Asymptotic Notations: GATE2015-3-42 top [Link]

L e t f(n) = n and g(n) = n(1+sin n) , where n is a positive integer. Which of the following
statements is/are correct?

I. f(n) = O(g(n))
II. f(n) = Ω(g(n))
A. Only I
B. Only II
C. Both I and II
D. Neither I nor II

© Copyright GATE Overflow. All rights reserved.

20 1 Algorithms (323)

gate2015-3 algorithms asymptotic-notations normal

Answer

1.3.15 Asymptotic Notations: GATE2017-1-04 top [Link]

Consider the following functions from positive integers to real numbers:

10, √−
n , n, log2 n, 100
n .
The CORRECT arrangement of the above functions in increasing order of asymptotic complexity is:

log2 n, 100 −
n , 10, √n , n
A.
100 −
n , 10, log2 n , √n , n
B.
10, 100 −
n , √n , log2 n , n
C.
100 −
n , log2 n , 10, √n , n
D.

gate2017-1 algorithms asymptotic-notations normal

Answer

1.3.16 Asymptotic Notations: TIFR2011-B-27 top [Link]

Let n be a large integer. Which of the following statements is TRUE?

−−−−−
A. n1/√log2 n < √log2 n < n1/100
−−−−−
B. n1/100 < n1/√log2 n < √log2 n
−−−−−
C. n1/√log2 n < n1/100 < √log2 n
−−−−−
D. √log2 n < n1/√log2 n < n1/100
−−−−−
E. √log2 n < n1/100 < n1/√log2 n

tifr2011 asymptotic-notations

Answer

1.3.17 Asymptotic Notations: TIFR2012-B-6 top [Link]

Let n be a large integer. Which of the following statements is TRUE?

A. 2√2 log n < n
log n
< n1/3
B. logn n < n1/3 < 2√2 log n
−−−−
C. 2√2 log n < n1/3 < logn n
−−−−
D. n1/3 < 2√2 log n < logn n
−−−−
E. n
log n
< 2√2 log n < n1/3

tifr2012 algorithms asymptotic-notations

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 21

1.3.18 Asymptotic Notations: TIFR2014-B-8 top [Link]

Which of these functions grows fastest with n?

A. en /n.
B. en−0.9 log n .
C. 2n .
D. (log n)n−1 .
E. None of the above.

tifr2014 algorithms asymptotic-notations

Answer

1.3.19 Asymptotic Notations: TIFR2016-B-7 top [Link]

Let n = m!. Which of the following is TRUE?

A. m = Θ(log n/ log log n)
B. m = Ω(log n/ log log n) but not m = O(log n/ log log n)
C. m = Θ(log2 n)
D. m = Ω(log2 n) but not m = O((log2 n)
E. m = Θ(log1.5 n)

tifr2016 asymptotic-notations

Answer

1.3.20 Asymptotic Notations: TIFR2017-A-4 top [Link]

Which of the following functions asymptotically grows the fastest as n goes to infinity?

A. (log log n)!

B. (log log n)log n
C. (log log n)log log log n

D. (log n)log log n

E. 2√log log n
tifr2017 algorithms asymptotic-notations

Answer

1.3.21 Asymptotic Notations: TIFR2018-A-3 top [Link]

Which of the following statements is TRUE for all sufficiently large integers n ?
√loglogn
A. 22 < 2√logn < n
√loglogn
B. 2√logn < n < 22
√loglogn
C. n < 2√logn < 22
√loglogn
D. n < 22 < 2√logn
√loglogn
E. 2√logn < 22 <n
tifr2018 asymptotic-notations

© Copyright GATE Overflow. All rights reserved.

22 1 Algorithms (323)

Answer

1.3.22 Asymptotic Notations: TIFR2018-B-5 top [Link]

Which of the following functions, given by there recurrence, grows the fastest asymptotically ?

A. T(n) = 4T( n
2 ) + 10n
B. T(n) = 8T( 3 ) + 24n2
n
2
C. T(n) = 16T( n
4 ) + 10n
1.99
5 ) + 20(nlogn)
D. T(n) = 25T( n
E. They all are asymptotically the same

tifr2018 asymptotic-notations recurrence

Answer

Answers: Asymptotic Notations

1.3.1 Asymptotic Notations: GATE1994-1.23 top [Link]


Selected Answer

For asymptotic complexity, we assume sufficiently large n. So, g1 (n) = n2 and g2 (n) = n3 .
Growth rate of g1 is less than that of g2 , i.e., g1 (n) = O(g2 (n)).

Options A and B are TRUE here.

 26 votes -- Arjun Suresh (350k points)

1.3.2 Asymptotic Notations: GATE1996-1.11 top [Link]


Selected Answer

log n
A. 100n log n = O( n 100 ) : Big-O denotes the growth rate of functions and multiplication or
division by a constant does not change the growth rate. So, this is TRUE and here O can even
be replaced by Θ or Ω.
−−−−
B. √log n = O(log log n) : FALSE. Take any long value like 256. LHS results in 16 but RHS
results in 4 only . Generally we take log on the left side but that is wrong.
C. 0 < x < y then nx = O (ny ) : TRUE since y is always greater than x. So, RHS is always
greater than LHS.
D. 2n ≠ O (nk) : TRUE since k is constant. So, for large values of n, LHS is much higher than
RHS (exponential function always greater than linear).

Only
B is FALSE.

 22 votes -- Prashant Singh (59.8k points)

1.3.3 Asymptotic Notations: GATE1999-2.21 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 23

n(n+1)
(M) T (n) = Sum of first n natural numbers = 2 = O(n2 )

(N) T (n) = Θ(n) = O(n), third case of Master theorem

(f(n) = n = Ω (nlogb a+ϵ ) = Ω (nlog2 1+ϵ ) = Ω (n0+ϵ ) , satisfied for any positive ϵ ≤ 1. Also,
af ( nb ) < cf(n) ⟹ f ( n2 ) < cf(n) ⟹ n2 < cn , satisfied for any c between 0 and 0.5)

(O) T (n) = Θ(n log n) = O(n log n), third case of Master theorem

(f(n) = n log n = Ω (nlogb a+ϵ ) = Ω (nlog2 1+ϵ ) = Ω (n0.5+ϵ ) , satisfied for positive ϵ = 0.5.
Also, af ( nb ) < cf(n) ⟹ f ( n 2 ) < cf(n) ⟹ 2 log 2 < cn log n , satisfied for c = 0.5)
n n

(P) Like in (M), here we are adding the log of the first n natural numbers. So,

Tn = log 1 + log 2 + log 3 + ⋯ + log n

= log(1 × 2 × ⋯ × n)
= log(n!)
= Θ(n log n) (Stirling's Approximation)

 28 votes -- Arjun Suresh (350k points)

1.3.4 Asymptotic Notations: GATE2000-2.17 top [Link]


Selected Answer

n = 256 n = 65536
3 × 65536256
16
3 × 256
f(n) = 3n√n = 3 × 216×256
128
=3×2
= 3 × 24096

216×8 2256×16
g(n) = 2√nlog2 n
= 2128 = 24096

256! 65536!
256 65536
h(n) = n! = O ((28 ) ) = O ((216 ) )

= O (22048 ) = O (21M )

Case of h(n) is given only by an upper bound but factorial has higher growth rate than
exponential.

[Link]
exponential-functions

f(n) and g(n) are having same order of growth as f(n) is simply 3 × g(n) (we can prove this by

© Copyright GATE Overflow. All rights reserved.

24 1 Algorithms (323)

taking log also). So, (d) is correct and all other choices are false.

 36 votes -- Arjun Suresh (350k points)

1.3.5 Asymptotic Notations: GATE2001-1.16 top [Link]


Selected Answer

A more formal approach:

f(n) = n2 log n
g(n) = n(log n)10
We can use the limit definition of O-notation
[Link]

f(n)
lim = 0, ⟹ f(n) = o(g(n))
n→∞ g(n)
small o implying f is strictly asymptotically lower than g. Also by definition,
o ⟹ O but O / ⟹ o.
f(n)
lim = c, c > 0 ⟹ f(n) = Θ(g(n))
n→∞ g(n)

f(n)
lim = ∞, ⟹ f(n) = ω(g(n))
n→∞ g(n)
small ω implying f is strictly asymptotically higher than g. Also by definition,
ω ⟹ Ω, but Ω / ⟹ ω.
We can use this to prove the above question

For any k>0

(log n)k
lim
n→∞ n
Applying L'Hôpital's rule,

k ∗ (log n)k−1
= lim
n→∞ n
k!
= lim =0
n→∞ n

So, (log n)k = o(n)

Now for large n, n > (log n)9
i.e n2 log n > n(log n)10
So n(log n)10 = o(n2 log n)
Or

10 2
n(log n = O( log n)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 25

n(log n)10 = O(n2 log n)

and

O(n2 log n) ≠ n(log n)10

Option B.

 14 votes -- janakyMurthy (1k points)

f(n) g(n)
10 10 10 10 10
n=2 10 ∗ 2 ∗ 2 2 ∗ 10
n = 2256 256 ∗ 2256 ∗ 2256 2256 ∗ 25610
So, as n is going larger, f(n) is overtaking g(n) and the growth rate of f is faster than that of g.
So, g(n) = O(f(n)) and f(n) ≠ O(g(n)).

B choice.

 39 votes -- Arjun Suresh (350k points)

1.3.6 Asymptotic Notations: GATE2003-20 top [Link]


Selected Answer

I. Rate of growth of (n + k)m is same as that of (nm ) as k and m are constants. (If either k or m
is a variable then the equality does not hold), i.e., for sufficiently large values of n,

(n + k)m ≤ anm and

nm ≤ b(n + k)m
where a and b are positive constants. Here, a can be km and b can be 1.
So, TRUE.

II. 2n+1 = 2 × (2n ) = Θ (2n ) as 2 is a constant here.

As 2n+1 is both upper and lower bounded by 2n we can say 2n+1 = O (2n ) . (Θ implies both O
as well as Ω)

So, TRUE.

III. 22n+1 has same rate of growth as 22n .

22n = 2n 2 = 2n × 2n

2n is upper bounded by (2n )2 , not the other way round as 22n is increasing by a factor of 2n
which is not a constant.

So, FALSE.

 27 votes -- Danish (3.8k points)

© Copyright GATE Overflow. All rights reserved.

26 1 Algorithms (323)

1.3.7 Asymptotic Notations: GATE2004-IT-55 top [Link]

Answer is
(D).
We can verify as : f⩽g but g  f . Therefore,
f < g.
Also,
g = h, as g = O(h) and h = O(g).

 16 votes -- Sandeep_Uniyal (7.5k points)

1.3.8 Asymptotic Notations: GATE2005-37 top [Link]


Selected Answer

Applying Masters theorem,

T (n) = Θ(n log n)

So, it cannot be Ω(n2 ).
Hence, answer is
(C).

 25 votes -- shreya ghosh (3.5k points)

1.3.9 Asymptotic Notations: GATE2008-39 top [Link]


Selected Answer

g(n) = n!.
On expanding the factorial we get g(n) = O(nn ) :
nn > nlog n
nn > 2n

This condition is violated by options A, B and C by first statements of each. Hence, they cannot
be said to be TRUE.

Second statement of option D says that g(n) is asymptotically biggest of all.

Answer is option
(D).

 19 votes -- Amar Vashishth (30.5k points)

1.3.10 Asymptotic Notations: GATE2008-IT-10 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 27


Selected Answer

A<C and A<D

E<B
and

C, D < E as E is exponential function.

Now, we just need to see if C or D is larger.

In C we have a term n3/4 and correspondingly in D we have log9 n (after taking n out).

n3/4 is asymptotically larger than log9 n as when n = 10100 , log9 n gives 1009 , while n3/4 gives
1075 > 10037 a much higher value and this is true for all higher values of n. So, D < C.
Thus, A is correct.
 28 votes -- Arjun Suresh (350k points)

1.3.11 Asymptotic Notations: GATE2011-37 top [Link]


Selected Answer

Answer is (A).

n log2 n < n3/2 is quite straightforward as n3/2 = n × n1/2 and log n < n1/2 as logarithmic
growth is smaller than exponential growth however small be the exponentiation factor.

Also, n3/2 < nlog2 n and n3/2 < 2n .

Now only nlog2 n and 2n need to be compared.

Taking log of both (log2 n)2 and n,

n > (log2 n)2

Hence, 2n > nlog2 n .

NOTE: We cannot compare two functions for asymptotic growth by taking log if they are giving
constants after log operation.

 28 votes -- Ankit Rokde (9k points)

1.3.12 Asymptotic Notations: GATE2012-18 top [Link]


Selected Answer

Worst case complexity can never be lower than the average case complexity, but it can be higher.
So, (C) is the answer.
A(n) = O(W(n)).
 31 votes -- Arjun Suresh (350k points)

© Copyright GATE Overflow. All rights reserved.

28 1 Algorithms (323)

1.3.13 Asymptotic Notations: GATE2015-3-4 top [Link]


Selected Answer

Sum of the cubes of the first n natural numbers is given by (n(n + 1)/2)2 which is Θ(n4 ). So, I,
III and IV are correct. II is wrong.

∴ (C) is correct.

 38 votes -- Arjun Suresh (350k points)

1.3.14 Asymptotic Notations: GATE2015-3-42 top [Link]


Selected Answer

The answer is option

(D).
Since the value of sin(n) will always range from −1 to +1, hence g(n) can take values 1, n, n2 .
Hence, if g(n) = 1, then statement I is incorrect.
And, if g(n) = n2 , then statement II is incorrect.

 39 votes -- saurabhrk (1.3k points)

1.3.15 Asymptotic Notations: GATE2017-1-04 top [Link]


Selected Answer

10 is constant. ∴ Growth rate is 0.

− log. (Consider √n2
= √−
log(n2 )
= 2)
√n grows slower than linear but faster than
√n
n , whereas log n

n : Growth rate is linear.

log2 n : Growth rate is logarithmic. For asymptotic growth, the base does not matter.

100
n : Growth rate decreases with n.

So, correct answer is

(B).
NOTE: Please never substitute large values of n in such questions. If ever you do, at least do for 2
such values and take ratio to get the growth rate or plot a graph. Remember 1.01n ≠ O (n100 ) .

 24 votes -- Arjun Suresh (350k points)

10−
,Constant
√ ,Square root
n

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 29

n, polynomial
log2 n, Logorithmic
100
n . Constant division by polynomial (clearly less than constant for every value of n
>100)

Now we know Constant division by polynomial < Constant < Logorithmic <Square
root < polynomial

So correct order is 100

n , 10, log2 n, √−
n, n

 21 votes -- Prashant Singh (59.8k points)

1.3.16 Asymptotic Notations: TIFR2011-B-27 top [Link]


Selected Answer

Let n = 2x . Then, log2 n = x.

f(n) = n1/√log2 n = (2x )1/√x = 2x/√x = 2√x

−−−−− −−−−−−− −
g(n) = √log2 n = √log2 (2x ) = √x

h(n) = n1/100 = (2x )1/100 = 2x/100

Since exponentials grow faster than polynomials, h(n) > g(n) for large n.

x
Since linear functions grow faster than square roots, 100 > √−
x for large x. Thus, h(n) > f(n)
for large n.

Since exponentials grow faster than polynomials, 2√x > √−

x for large √−
x . Thus, f(n) > g(n) for
large n.

Hence, the relation is,

g(n) < f(n) < h(n)

Thus, option

(D)
© Copyright GATE Overflow. All rights reserved.
30 1 Algorithms (323)

(D) is correct.

 23 votes -- Pragy Agarwal (20.6k points)

1.3.17 Asymptotic Notations: TIFR2012-B-6 top [Link]


Selected Answer

Answer will be (C).

Take n = 21024

Now, 2√(2 log n) ≈ 245

1
n 3 ≈ 2341

n/ log n = 21024 /1024 ≈ 21014

Just one value is not enough to confirm growth rate. So, take n = 1024.

Now, 2√(2 log n) ≈ 24

1
n 3 ≈ 23

n/ log n = 210 /10 ≈ 27

So, as n increases, the gap between second and third function increases and also the second
function overtakes the first. So, f1 < f2 < f3.

 13 votes -- srestha (87k points)

1.3.18 Asymptotic Notations: TIFR2014-B-8 top [Link]


Selected Answer

Assuming that the base of the log in the question is e.

Let us try to rewrite each of these functions in the form e something , to make the comparison easier.
en
a. en / n = = e (n − ln n)
e (ln n)

b. en−0.9 ln n = e (n − 0.9 ln n)

n
c. 2n = (e ln 2 ) = e (n ln 2)

n−1
d. (ln n)n−1 = (e ln ln n ) = e (n ln ln n − ln ln n)

Now, if we just compare the exponents of all, we can clearly see that (n ln ln n − ln ln n) grows

(C)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 31

faster than the rest. Note that in option (C), the multiplicative ln 2 is a constant, and hence grows
slower than the multiplicative ln ln n from option (D).

This implies that e (n ln ln n − ln ln n) grows the fastest, and hence, (ln n)n−1 grows the fastest.

Thus, option
(D) is the correct answer.

 25 votes -- Pragy Agarwal (20.6k points)

1.3.19 Asymptotic Notations: TIFR2016-B-7 top [Link]

log n/ log
m n = m!
log n
4 24 4/2 = 2
6 720 9/3 = 3
8 720 ∗ 56 15/3 = 5
720 ∗ 56 21/4 = 5
10
∗ 90
If we see, m is growing at same rate as log n/ log log n.

 9 votes -- Arjun Suresh (350k points)

1.3.20 Asymptotic Notations: TIFR2017-A-4 top [Link]


Selected Answer

Let N = 2256
A. (log log N)! = 8!
B. (log log N )log N = (8)256 = 2768
C. (log log N )log log log N = 83 = 512
D. (log N )log log N = (256)8 = 264
E. 2√log log N = 22√2
Let N = 216
A. (log log N)! = 4!
B. (log log N )log N = (4)16 = 232
C. (log log N )log log log N = 42 = 16
D. (log N )log log N = (16)4 = 216
E. 2√log log N = 22 = 4
Taking ratio for both N values,

A. (log log N)! → 8!/4!

B. (log log N )log N → 2736
C. (log log N )log log log N → 32
D. (log N )log log N → 248
E. 2√log log N →≈ 2

© Copyright GATE Overflow. All rights reserved.

32 1 Algorithms (323)

Option B = (log log N)log N asymptotically grows the fastest, as for the same change in N , the
value increased the most (growth) for option B and this growth is monotonic (continuous).

 23 votes -- Kapil Phulwani (50k points)

1.3.21 Asymptotic Notations: TIFR2018-A-3 top [Link]

n = 22
k
Take

(A) gives n = 22
k
Now, option

√log22k √2k
Option (B) gives 2 =2
√log(log22k ) √k
Option (C) gives 22 = 22
Now, check power only for (B) and (C).
Take log of power of both functions, i.e.,
−−
(log(2√k ) = O(log(√2k )
−
√k = O( k2 )

Clearly, (C) < (B) < (A).

Option (A) is answer.
 3 votes -- Anu007 (17k points)

1.3.22 Asymptotic Notations: TIFR2018-B-5 top [Link]


Selected Answer

By applying master theorem,

(a) T(n) = Θ(n2 )

(b) T(n) = Θ(n2 )
(c) T(n) = Θ(n2 *log n)
(d) T(n) = Θ(n2 )
C is growing fastest.

 5 votes -- Hemant Parihar (14.8k points)

1.4 Dynamic Programming(12) top

1.4.1 Dynamic Programming: GATE2008-80 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 33

The subset-sum problem is defined as follows. Given a set of n positive integers,

S = {a1 , a2 , a3 , … , an } , and positive integer W , is there a subset of S whose elements sum to
W ? A dynamic program for solving this problem uses a 2-dimensional Boolean array, X, with n
rows and W + 1 columns. X[i, j], 1 ≤ i ≤ n, 0 ≤ j ≤ W , is TRUE, if and only if there is a subset
of {a1 , a2 , … , ai } whose elements sum to j.

Which of the following is valid for 2 ≤ i ≤ n, and ai ≤ j ≤ W ?

A. X[i, j] = X[i − 1, j] ∨ X[i, j − ai ]
B. X[i, j] = X[i − 1, j] ∨ X[i − 1, j − ai ]
C. X[i, j] = X[i − 1, j] ∧ X[i, j − ai ]
D. X[i, j] = X[i − 1, j] ∧ X[i − 1, j − ai ]

gate2008 algorithms normal dynamic-programming

Answer

1.4.2 Dynamic Programming: GATE2008-81 top [Link]

The subset-sum problem is defined as follows. Given a set of n positive integers,

S = {a1 , a2 , a3 , … , an } , and positive integer W , is there a subset of S whose elements sum to
W ? A dynamic program for solving this problem uses a 2-dimensional Boolean array, X, with n
rows and W + 1 columns. X[i, j], 1 ≤ i ≤ n, 0 ≤ j ≤ W , is TRUE, if and only if there is a subset
of {a1 , a2 , … , ai } whose elements sum to j.

Which entry of the array X, if TRUE, implies that there is a subset whose elements sum to W ?
A. X[1, W]
B. X[n, 0]
C. X[n, W]
D. X[n − 1, n]

gate2008 algorithms normal dynamic-programming

Answer

1.4.3 Dynamic Programming: GATE2009-53 top [Link]

A sub-sequence of a given sequence is just the given sequence with some elements (possibly none
or all) left out. We are given two sequences X[m] and Y [n] of lengths m and n, respectively with
indexes of X and Y starting from 0.

We wish to find the length of the longest common sub-sequence (LCS) of X[m] and Y [n] as
l(m, n), where an incomplete recursive definition for the function I(i, j) to compute the length of
the LCS of X[m] and Y [n] is given below:

l(i,j) = 0, if either i = 0 or j = 0
= expr1, if i,j > 0 and X[i-1] = Y[j-1]
= expr2, if i,j > 0 and X[i-1] ≠ Y[j-1]

Which one of the following options is correct?

A. expr1 = l (i − 1, j) + 1
B. expr1 = l (i, j − 1)
C. expr2 = max (l (i − 1, j) , l (i, j − 1))
expr2 = max (l (i − 1, j − 1) , l (i, j))
© Copyright GATE Overflow. All rights reserved.
34 1 Algorithms (323)

D. expr2 = max (l (i − 1, j − 1) , l (i, j))

gate2009 algorithms normal dynamic-programming recursion

Answer

1.4.4 Dynamic Programming: GATE2009-54 top [Link]

A sub-sequence of a given sequence is just the given sequence with some elements (possibly none
or all) left out. We are given two sequences X[m] and Y [n] of lengths m and n, respectively with
indexes of X and Y starting from 0.

We wish to find the length of the longest common sub-sequence (LCS) of X[m] and Y [n] as
l(m, n), where an incomplete recursive definition for the function I(i, j) to compute the length of
the LCS of X[m] and Y [n] is given below:

l(i, j) = 0, if either i = 0 or j = 0
= expr1, if i, j > 0 and X[i − 1] = Y [j − 1]
= expr2, if i, j > 0 and X[i − 1] ≠ Y [j − 1]
The value of l(i, j) could be obtained by dynamic programming based on the correct recursive
definition of l(i, j) of the form given above, using an array L[M, N], where M = m + 1 and
N = n + 1 , such that L[i, j] = l(i, j).
Which one of the following statements would be TRUE regarding the dynamic programming solution
for the recursive definition of l(i, j)?

A. All elements of L should be initialized to 0 for the values of l(i, j) to be properly computed.
B. The values of l(i, j) may be computed in a row major order or column major order of L[M, N].
C. The values of l(i, j) cannot be computed in either row major order or column major order of
L[M, N].
D. L[p, q] needs to be computed before L[r, s] if either p < r or q < s.

gate2009 normal algorithms dynamic-programming recursion

Answer

1.4.5 Dynamic Programming: GATE2010-34 top [Link]

The weight sequence a0 , a1 , … , an−1

of a of real numbers is defined as
a0 + a1 /2 + ⋯ + an−1 /2 n−1
. A subsequence of a sequence is obtained by deleting some
elements from the sequence, keeping the order of the remaining elements the same. Let X denote
the maximum possible weight of a subsequence of ao , a1 , … , an−1 and Y the maximum possible
weight of a subsequence of a1 , a2 , … , an−1 . Then X is equal to

A. max(Y , a0 + Y )
B. max(Y , a0 + Y /2)
C. max(Y , a0 + 2Y )
D. a0 + Y /2

gate2010 algorithms dynamic-programming normal

Answer

1.4.6 Dynamic Programming: GATE2011-25 top [Link]

An algorithm to find the length of the longest monotonically increasing sequence of numbers in an

A[0 : n − 1]
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 35

array A[0 : n − 1] is given below.

Let Li , denote the length of the longest monotonically increasing sequence starting at index i in the
array.

Initialize Ln−1 = 1.
For all i such that 0 ≤ i ≤ n − 2

Li = {
1 + Li+1 if A[i] < A[i+1]
1 Otherwise
Finally, the length of the longest monotonically increasing sequence is max (L0 , L1 , … , Ln−1 ).
Which of the following statements is TRUE?

A. The algorithm uses dynamic programming paradigm

B. The algorithm has a linear complexity and uses branch and bound paradigm
C. The algorithm has a non-linear polynomial complexity and uses branch and bound paradigm
D. The algorithm uses divide and conquer paradigm

gate2011 algorithms easy dynamic-programming

Answer

1.4.7 Dynamic Programming: GATE2011-38 top [Link]

Four Matrices M1 , M2 , M3 and M4 of dimensions p × q, q × r, r × s and s × t respectively can

be multiplied in several ways with different number of total scalar multiplications. For example when
multiplied as ((M1 × M2 ) × (M3 × M4 )) , the total number of scalar multiplications is
pqr + rst + prt. When multiplied as (((M1 × M2 ) × M3 ) × M4 ) , the total number of scalar
multiplications is pqr + prs + pst.

If p = 10, q = 100, r = 20, s = 5 and t = 80, then the minimum number of scalar multiplications
needed is

A. 248000
B. 44000
C. 19000
D. 25000
gate2011 algorithms dynamic-programming normal

Answer

1.4.8 Dynamic Programming: GATE2014-2-37 top [Link]

Consider two strings A="qpqrr" and B="pqprqrp". Let x be the length of the longest common
subsequence (not necessarily contiguous) between A and B and let y be the number of such
longest common subsequences between A and B. Then x + 10y = ___.

gate2014-2 algorithms normal numerical-answers dynamic-programming

Answer

1.4.9 Dynamic Programming: GATE2014-3-37 top [Link]

Suppose you want to move from 0 to 100 on the number line. In each step, you either move right

© Copyright GATE Overflow. All rights reserved.

36 1 Algorithms (323)

by a unit distance or you take a shortcut. A shortcut is simply a pre-specified pair of integers
i, j with i < j . Given a shortcut (i, j), if you are at position i on the number line, you may directly
move to j. Suppose T (k) denotes the smallest number of steps needed to move from k to 100.
Suppose further that there is at most 1 shortcut involving any number, and in particular, from 9
there is a shortcut to 15. Let y and z be such that T (9) = 1 + min(T (y), T (z)). Then the value of
the product yz is _____.

gate2014-3 algorithms normal numerical-answers dynamic-programming

Answer

1.4.10 Dynamic Programming: GATE2016-2-14 top [Link]

The Floyd-Warshall algorithm for all-pair shortest paths computation is based on

A. Greedy paradigm.
B. Divide-and-conquer paradigm.
C. Dynamic Programming paradigm.
D. Neither Greedy nor Divide-and-Conquer nor Dynamic Programming paradigm.

gate2016-2 algorithms dynamic-programming easy

Answer

1.4.11 Dynamic Programming: GATE2016-2-38 top [Link]

L e t A1 , A2 , A3 and A4 be four matrices of dimensions 10 × 5, 5 × 20, 20 × 10 and 10 × 5,

respectively. The minimum number of scalar multiplications required to find the product A1 A2 A3 A4
using the basic matrix multiplication method is _________.
gate2016-2 dynamic-programming algorithms normal numerical-answers

Answer

1.4.12 Dynamic Programming: GATE2018-31 top [Link]

Assume that multiplying a matrix G1 of dimension p × q with another matrix G2 of dimension q × r

requires pqr scalar multiplications. Computing the product of n matrices G1 G2 G3 … Gn can be
done by parenthesizing in different ways. Define Gi Gi+1 as an explicitly computed pair for a
given paranthesization if they are directly multiplied. Fr example, in the matrix multiplication chain
G1 G2 G3 G4 G5 G6 using parenthesization (G1 (G2 G3 ))(G4 (G5 G6 )), G2 G3 and G5 G6 are only
explicitly computed pairs.

Consider a matrix multiplication chain F1 F2 F3 F4 F5 , where matrices F1 , F2 , F3 , F4 and F5 are of

dimensions 2 × 25, 25 × 3, 3 × 16, 16 × 1 and 1 × 1000, respectively. In the parenthesization of
F1 F2 F3 F4 F5 that minimizes the total number of scalar multiplications, the explicitly computed
pairs is/are

A. F1 F2 and F3 F4 only
B. F2 F3 only
C. F3 F4 only
D. F2 F2 and F4 F5 only

gate2018 algorithms dynamic-programming

Answer

Answers: Dynamic Programming

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 37

1.4.1 Dynamic Programming: GATE2008-80 top [Link]


Selected Answer

This is analogous to the dynamic programming solution to 0/1 knapsack problem.

Consider the capacity of the knapsack, i.e., W to be analogous to J (the total sum here).
The solution exploits the optimal substructure of the problem.

At each stage we can have 2 options:

Case (1): Either we take an item(in this question either we consider the element Ai ) along with
the total solution to previous sub-problem(total solution here means the total sum obtained till
previous sub-problem)

in which case we choose A[i − 1][j − ai ]

A[i − 1] indicates we are considering solution to previous subproblem and
A[j − ai ] means we have considered element ai and now remaining sum is J − ai which has to
be further considered.

Case (2): Or we do not consider the item(in this case the element ai ) in which case we only
consider the solution to previous subproblem, which is,

A[i − 1][J]
Since the whole solution to this subset-sum problem is Logical OR(+) of cases 1 and 2, we
eliminate options C and D because both are considering the Logical AND of the two parts of the
solution.

Now, since here in the given question we are given a boolean array X[n][W + 1]
So, an entry X[i][j] is true only if sum j is possible with array elements from 0 to i.
So, for Each element of array, ai , we consider two possibilites:
(1) EIther we can ignore it and still get our possible sum, which is,

X[i − 1][j]
OR

(2) We could include element ai and then get our required sum, which is,

X[i − 1][j − ai ]
And finally, to get X[i][j], we take logical or of the above two cases.
Hence, answer is option B.

Reference :

Video:

By using the analogy of the problem and solution between subset-sum problem and 0/1 knapsack
problem, the above video clearly explains the how the solution to the problem is structured .

Video:

© Copyright GATE Overflow. All rights reserved.

38 1 Algorithms (323)

 8 votes -- Ayush Upadhyaya (9k points)

1.4.2 Dynamic Programming: GATE2008-81 top [Link]


Selected Answer

ANSWER is C.

If LAST ROW and LAST COLUMN entry is 1, then there exists a subset whose elements sum to W .

 11 votes -- Shivam Bhardwaj (289 points)

1.4.3 Dynamic Programming: GATE2009-53 top [Link]


Selected Answer

Answer is C. When the currently compared elements doesn't match, we have two possibilities for
the LCS, one including X[i] but not Y[j] and other including Y[j] but not X[i].

/* Returns length of LCS for X[0..m-1], Y[0..n-1] */

int lcs( char *X, char *Y, int m, int n )
{
if (m == 0 || n == 0)
return 0;
if (X[m-1] == Y[n-1])
return 1 + lcs(X, Y, m-1, n-1);
else
return max(lcs(X, Y, m, n-1), lcs(X, Y, m-1, n));
}

 20 votes -- Sona Praneeth Akula (4.2k points)

1.4.4 Dynamic Programming: GATE2009-54 top [Link]


Selected Answer

expr2 = max (l (i − 1, j) , l (i, j − 1))

When the currently compared elements doesn't match, we have two possibilities for the LCS, one
including X[i] but not Y [j] and other including Y [j] but not X[i].

/* Returns length of LCS for X[0..m-1], Y[0..n-1] */

int lcs( char *X, char *Y, int m, int n )
{
if (m == 0 || n == 0)
return 0;
if (X[m-1] == Y[n-1])
return 1 + lcs(X, Y, m-1, n-1);
else
return max(lcs(X, Y, m, n-1), lcs(X, Y, m-1, n));
}

Answer is B. Dynamic programming is used to save the previously found LCS. So, for any index
[p,q] all smaller ones should have been computed earlier. Option D is not correct as the condition
given requires even L[3,2] to be computed before L[2,4] which is not a necessity if we follow row-
major order.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 39

int lcs( char *X, char *Y, int m, int n )

{
int L[m+1][n+1];
int i, j;

/* Following steps build L[m+1][n+1] in bottom up fashion. Note

that L[i][j] contains length of LCS of X[0..i-1] and Y[0..j-1] */
for (i=0; i<=m; i++)
{
for (j=0; j<=n; j++)
{
if (i == 0 || j == 0)
L[i][j] = 0;

else if (X[i-1] == Y[j-1])

L[i][j] = L[i-1][j-1] + 1;

else
L[i][j] = max(L[i-1][j], L[i][j-1]);
}
}

/* L[m][n] contains length of LCS for X[0..n-1] and Y[0..m-1] */

return L[m][n];
}

 26 votes -- Arjun Suresh (350k points)

1.4.5 Dynamic Programming: GATE2010-34 top [Link]


Selected Answer

S = ⟨a0 , S1 ⟩
S1 = ⟨a1 , a2 , a3 … an−1 ⟩
Two possible cases arise:

1. a0 is included in the max weight subsequence of

S:
Y
In this case, X = weight((⟨a0 , S1 ⟩)) = a0 +
2
2. a0 is not included in the max weight subsequence of
S:
In this case, X = weight(S1 ) = Y
Since the value of a0 can be anything (negative or < Y
2 in general) {∵ ai ∈ R} , it is possible that
Y > a0 + Y
2 .

The maximum possible weight of a subsequence of S is given by:

X = max (Y , a0 + )
Y
2
Thus, option B is correct.

© Copyright GATE Overflow. All rights reserved.

40 1 Algorithms (323)

 50 votes -- Pragy Agarwal (20.6k points)

1.4.6 Dynamic Programming: GATE2011-25 top [Link]


Selected Answer

(A) is the answer.

The algorithm is storing the optimal solutions to subproblems at each point (for each i), and then
using it to derive the optimal solution of a bigger problem. And that is dynamic programming
approach. And the program has linear time complexity.

[Link]

Now, branch and bound comes when we explore all possible solutions (branch) and we
backtrack as soon as we realise we won't get a solution (in classical backtracking we will retreat
only when we won't find the solution). In backtracking : In each step, you check if this step
satisfies all the conditions.
If it does : you continue generating subsequent solutions
If not : you go one step backward to check for another path

So, backtracking gives all possible solutions while branch and bound will give only the optimal
one. [Link]

The given algorithm here is neither backtracking nor branch and bound. Because we are not
branching anywhere in the solution space.

And the algorithm is not divide and conquer as we are not dividing the problem and then merging
the solution as in the case of merge sort (where merge is the conquer step).

[Link]

 38 votes -- Arjun Suresh (350k points)

1.4.7 Dynamic Programming: GATE2011-38 top [Link]


Selected Answer

Answer is C.

Ordering: First Multiply M2 × M3 . This requires 100*20*5 multiplications.

Then Multiply M1 × (M2 × M3 ). This requires 10*100*5 multiplications.
Then Multiply (M1 × (M2 × M3 )) × M4 . This requires 10*5*8 multiplications.
Total 19000 Multiplications.

Brute Force approach - anyone can do.

No. of possible ordering for 4 matrices is C3 where C3 is the 3rd Catalan number and given by
1 2n
n = 3 in n+1 Cn = 5.
So, here we have

1. (M1 × M2 ) × (M3 × M4 )
2. (M1 × (M2 × M3 )) × M4
3. ((M1 × M2 ) × M3 ) × M4

1 ×( 2 ×( 3 × 4 ))
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 41

4. M1 × (M2 × (M3 × M4 ))
5. M1 × ((M2 × M3 ) × M4 ))
Each of these would give no. of multiplications required as

1. pqr + rst + prt

2. qrs + pqs + pst
3. pqr + prs + pst
4. rst + qrt + pqt
5. qrs + qst + pst
The last 2 are having qt terms which are the highest terms by far and hence we can avoid them
from consideration qt = 8000 multiplied by one other term would be larger than any value in
choice. So, just find the value of first 3 terms.

1. pqr + rst + prt = 20000 + 8000 + 16000 = 44000

2. qrs + pqs + pst = 10000 + 5000 + 4000 = 19000 - smallest value in choice, we can stop
here.
3. pqr + prs + pst

Dynamic Programming Solution (should know Matrix Chain Ordering algorithm)

Here we have a chain of length 4.

Dynamic programming solution of Matrix chain ordering has the solution

0 if i = j
m[i, j] = {
mini≤k<j m[i]k] + m[k + 1][j] + pi−1 pj pk if i < j
So, we can fill the following table starting with the diagonals and moving upward diagonally. Here
k < j but ≥ i.
j=1 j=2 j=3

i=1 0 p0 p1 p2 = 20000 min(10000 + p0 p1 p3 , 20000 + p0 + p2 p3 ) = 15000 min(18000 + p0 p1 p

i=2 0 p1 p2 p3 = 10000 min(10000 + p2 p3 p
i=3 0 p2 p3 p4 = 8000
i=4 0

Our required answer is given by m[1, 4] = 19000.

 25 votes -- Sona Praneeth Akula (4.2k points)

1.4.8 Dynamic Programming: GATE2014-2-37 top [Link]


Selected Answer

Answer is 34.
In first string, if we want to get 4 as maximum length then LCS should end with either "rr" or "qr".
Only 4 combinations are possible for LCS with length 4:

"qpqr"

"qqrr"

© Copyright GATE Overflow. All rights reserved.

42 1 Algorithms (323)

"pqrr"

"qprr"

Now, check for matching sequences in second string, except for "qqrr" all are possible.

 16 votes -- Anurag Semwal (8k points)

1.4.9 Dynamic Programming: GATE2014-3-37 top [Link]


Selected Answer

T (k) is the smallest number of steps needed to move from k to 100.

Now, it is given that y and z are two numbers such that,
T (9) = 1 + min(T (y), T (z)) , i.e.,
T (9) = 1 + min(Steps from y to 100, Steps from z to 100), where y and z are two possible
values that can be reached from 9.

One number that can be reached from 9 is 10, which is the number obtained if we simply move
one position right on the number line. Another number is 15, the shortcut path from 9, as given in
the question. So, we have two paths from 9, one is 10 and the other is 15.

Therefore, the value of y and z is 10 and 15 (either variable may take either of the values).
Thus, yz = 150.
 17 votes -- Divya Bharti (8.1k points)

T (9) = Distance from 9 to 100

T (9) = 1 + min(T (y), T (z)) = 1+min(Distance from y to 100 , Distance from z to 100)
There are only two such values where we can reach from 9 , one is simple step to right on number
line , i.e 10 and another is 15 (given shortcut)

Hence ,y = 10 , z = 15
yz = 10 × 15 = 150
 31 votes -- Srinath Jayachandran (3.7k points)

1.4.10 Dynamic Programming: GATE2016-2-14 top [Link]


Selected Answer

In Floyd Warshall's, we calculate all possibilities and select best one so its neither Divide & Conquer
nor Greedy but based on Dynamic Programming Paradigm.

 25 votes -- Anurag Semwal (8k points)

1.4.11 Dynamic Programming: GATE2016-2-38 top [Link]


Selected Answer

Answer is 1500.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 43

atrix Parenthesizing : A1 ((A2 A3 )A4 )

Check my solution below, using dynamic programming

A1 A2 A3 A4
20
10 × 5 5 × 20 10 × 5
× 10

A12 = 10 × 5 × 20 = 1000
A23 = 5 × 20 × 10 = 1000
A34 = 20 × 10 × 5 = 1000
A12 + A33 + 5 × 20 × 10 = 2000
A13 = min {
A11 + A23 + 10 × 5 × 10 = 1500
A23 + A44 + 5 × 10 × 5 = 1250
A24 = min {
A22 + A34 + 5 × 20 × 5 = 1500
⎧
⎪ A11 + A24 + 10 × 5 × 5 = 1500
A14 = min ⎨ A12 + A34 + 10 × 20 × 5 ⩾ 2000
⎩
⎪
A13 + A44 + 10 × 20 × 5 = 2000
Answer is 1500.

 31 votes -- Akash Kanase (42.5k points)

1.4.12 Dynamic Programming: GATE2018-31 top [Link]


Selected Answer

If we multiply anything with F5 we will get much greater multiplication cost because F5 is
1 ∗ 1000 matrix so 1000 will play vital role in cost. So we will multiply F5 at very last step.
So, here is the sequence giving minimal cost:

(((F1 (F2 (F3 F4 ))(F5 )) = 48 + 75 + 50 + 2000 = 2173

Explicitly computed pairs is (F3 F4 )
 12 votes -- Digvijay (54.9k points)

1.5 Graph Algorithms(48) top

1.5.1 Graph Algorithms: GATE1994-1.22 top [Link]

Which of the following statements is false?

A. Optimal binary search tree construction can be performed efficiently using dynamic programming

B. Breadth-first search cannot be used to find connected components of a graph

© Copyright GATE Overflow. All rights reserved.

44 1 Algorithms (323)

C. Given the prefix and postfix walks over a binary tree, the binary tree cannot be uniquely
constructed.

D. Depth-first search can be used to find connected components of a graph

gate1994 algorithms normal graph-algorithms

Answer

1.5.2 Graph Algorithms: GATE1994-24 top [Link]

An independent set in a graph is a subset of vertices such that no two vertices in the subset are
connected by an edge. An incomplete scheme for a greedy algorithm to find a maximum
independent set in a tree is given below:

V: Set of all vertices in the tree;

I := ϕ
while V ≠ ϕ do
begin
select a vertex u ∊ V such that
_______;
V := V - {u};
if u is such that
________then I := I ∪ {u}
end;
Output(I);

a. Complete the algorithm by specifying the property of vertex u in each case.

b. What is the time complexity of the algorithm?

gate1994 algorithms graph-algorithms normal

Answer

1.5.3 Graph Algorithms: GATE1996-17 top [Link]

Let G be the directed, weighted graph shown in below figure

We are interested in the shortest paths from A.

a. Output the sequence of vertices identified by the Dijkstra’s algorithm for single source shortest
path when the algorithm is started at node A

b. Write down sequence of vertices in the shortest path from A to E

c. What is the cost of the shortest path from A to E?

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 45

gate1996 algorithms graph-algorithms normal

Answer

1.5.4 Graph Algorithms: GATE1998-1.21, ISRO2008-16 top

[Link]

Which one of the following algorithm design techniques is used in finding all pairs of shortest
distances in a graph?

A. Dynamic programming

B. Backtracking

C. Greedy

D. Divide and Conquer

gate1998 algorithms graph-algorithms easy isro2008

Answer

1.5.5 Graph Algorithms: GATE2000-1.13 top [Link]

The most appropriate matching for the following pairs

X: depth first search 1: heap

Y: breadth-first search 2: queue
Z: sorting 3: stack

is:

A. X - 1, Y - 2, Z-3
B. X - 3, Y - 1, Z-2
C. X - 3, Y - 2, Z-1
D. X - 2, Y - 3, Z-1
gate2000 algorithms easy graph-algorithms

Answer

1.5.6 Graph Algorithms: GATE2001-2.14 top [Link]

Consider an undirected, unweighted graph G. Let a breadth-first traversal of G be done starting

from a node r. Let d(r, u) and d(r, v) be the lengths of the shortest paths from r to u and v
respectively in G. If u is visited before v during the breadth-first traversal, which of the following
statements is correct?

A. d(r, u) < d(r, v)

B. d(r, u) > d(r, v)
C. d(r, u) ≤ d(r, v)
D. None of the above

gate2001 algorithms graph-algorithms normal

Answer

1.5.7 Graph Algorithms: GATE2002-12 top [Link]

© Copyright GATE Overflow. All rights reserved.

46 1 Algorithms (323)

Fill in the blanks in the following template of an algorithm to compute all pairs shortest path lengths
in a directed graph G with n ∗ n adjacency matrix A. A[i, j] equals 1 if there is an edge in G from
i to j, and 0 otherwise. Your aim in filling in the blanks is to ensure that the algorithm is correct.
INITIALIZATION: For i = 1 ... n
{For j = 1 ... n
{ if a[i,j] = 0 then P[i,j] =_______ else P[i,j] =_______;}
}

ALGORITHM: For i = 1 ... n

{For j = 1 ... n
{For k = 1 ... n
{P[__,__] = min{_______,______}; }
}
}

a. Copy the complete line containing the blanks in the Initialization step and fill in the blanks.
b. Copy the complete line containing the blanks in the Algorithm step and fill in the blanks.
c. Fill in the blank: The running time of the Algorithm is O(___).

gate2002 algorithms graph-algorithms time-complexity normal descriptive

Answer

1.5.8 Graph Algorithms: GATE2003-21 top [Link]

Consider the following graph:

Among the following sequences:

I. abeghf
II. abfehg
III. abfhge
IV. afghbe

Which are the depth-first traversals of the above graph?

A. I, II and IV only
B. I and IV only
C. II, III and IV only
D. I, III and IV only

gate2003 algorithms graph-algorithms normal

Answer

1.5.9 Graph Algorithms: GATE2003-67 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 47

Let G = (V , E) be an undirected graph with a subgraph G1 = (V1 , E1 ) . Weights are assigned to

edges of G as follows.

w(e) = {
0, if e ∈ E1
1, otherwise
A single-source shortest path algorithm is executed on the weighted graph (V , E, w) with an
arbitrary vertex v1 of V1 as the source. Which of the following can always be inferred from the path
costs computed?

A. The number of edges in the shortest paths from v1 to all vertices of G

B. G1 is connected

C. V1 forms a clique in G
D. G1 is a tree

gate2003 algorithms graph-algorithms normal

Answer

1.5.10 Graph Algorithms: GATE2003-70 top [Link]

Let G = (V , E) be a directed graph with n vertices. A path from vi to vj in G is a sequence of

vertices (vi , vi+1 , … , vj ) such that (vk , vk + 1) ∈ E for all k in i through j − 1 . A simple path is
a path in which no vertex appears more than once.

Let A be an n × n array initialized as follows:

A[j, k] = {
1, if (j, k) ∈ E
0, otherwise
Consider the following algorithm:

for i=1 to n
for j=1 to n
for k=1 to n
A[j,k] = max(A[j,k], A[j,i] + A[i,k]);

Which of the following statements is necessarily true for all j and k after termination of the above
algorithm?

A. A[j, k] ≤ n
B. If A[j, j] ≥ n − 1 then G has a Hamiltonian cycle
C. If there exists a path from j to k, A[j, k] contains the longest path length from j to k
D. If there exists a path from j to k, every simple path from j to k contains at most A[j, k] edges

gate2003 algorithms graph-algorithms normal

Answer

1.5.11 Graph Algorithms: GATE2004-44 top [Link]

Suppose we run Dijkstra’s single source shortest-path algorithm on the following edge-weighted
directed graph with vertex P as the source.

© Copyright GATE Overflow. All rights reserved.

48 1 Algorithms (323)

In what order do the nodes get included into the set of vertices for which the shortest path distances
are finalized?

A. P, Q, R, S, T , U
B. P, Q, R, U, S, T
C. P, Q, R, U, T , S
D. P, Q, T , R, U, S
gate2004 algorithms graph-algorithms normal

Answer

1.5.12 Graph Algorithms: GATE2004-81 top [Link]

Let G1 = (V , E1 ) and G2 = (V , E2 ) be connected graphs on the same vertex set V with more
than two vertices. If G1 ∩ G2 = (V , E1 ∩ E2 ) is not a connected graph, then the graph
G1 ∪ G2 = (V , E1 ∪ E2 )
A. cannot have a cut vertex
B. must have a cycle
C. must have a cut-edge (bridge)
D. has chromatic number strictly greater than those of G1 and G2
gate2004 algorithms graph-algorithms normal

Answer

1.5.13 Graph Algorithms: GATE2004-IT-56 top [Link]

Consider the undirected graph below:

Using Prim's algorithm to construct a minimum spanning tree starting with node A, which one of the
following sequences of edges represents a possible order in which the edges would be added to
construct the minimum spanning tree?

A. (E, G), (C, F), (F, G), (A, D), (A, B), (A, C)
B. (A, D), (A, B), (A, C), (C, F), (G, E), (F, G)
C. (A, B), (A, D), (D, F), (F, G), (G, E), (F, C)
D. (A, D), (A, B), (D, F), (F, C), (F, G), (G, E)

gate2004-it algorithms graph-algorithms normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 49

1.5.14 Graph Algorithms: GATE2005-38 top [Link]

Let G(V , E) be an undirected graph with positive edge weights. Dijkstra’s single source shortest
path algorithm can be implemented using the binary heap data structure with time complexity:

A. O (|V |2 )

B. O (|E| + |V | log |V |)
C. O (|V | log |V |)
D. O ((|E| + |V |) log |V |)

gate2005 algorithms graph-algorithms normal

Answer

1.5.15 Graph Algorithms: GATE2005-82a top [Link]

Let s and t be two vertices in a undirected graph G = (V , E) having distinct positive edge weights.
L e t [X, Y ] be a partition of V such that s ∈ X and t ∈ Y . Consider the edge e having the
minimum weight amongst all those edges that have one vertex in X and one vertex in Y .

The edge e must definitely belong to:

A. the minimum weighted spanning tree of G

B. the weighted shortest path from s to t

C. each path from s to t
D. the weighted longest path from s to t
gate2005 algorithms graph-algorithms normal

Answer

1.5.16 Graph Algorithms: GATE2005-82b top [Link]

Let s and t be two vertices in a undirected graph G = (V , E) having distinct positive edge weights.
L e t [X, Y ] be a partition of V such that s ∈ X and t ∈ Y . Consider the edge e having the
minimum weight amongst all those edges that have one vertex in X and one vertex in Y .

Let the weight of an edge e denote the congestion on that edge. The congestion on a path is defined
to be the maximum of the congestions on the edges of the path. We wish to find the path from s to t
having minimum congestion. Which of the following paths is always such a path of minimum
congestion?

A. a path from s to t in the minimum weighted spanning tree

B. a weighted shortest path from s to t
C. an Euler walk from s to t
D. a Hamiltonian path from s to t
gate2005 algorithms graph-algorithms normal

Answer

© Copyright GATE Overflow. All rights reserved.

50 1 Algorithms (323)

1.5.17 Graph Algorithms: GATE2005-IT-14 top [Link]

In a depth-first traversal of a graph G with n vertices, k edges are marked as tree edges. The
number of connected components in G is

A. k
B. k+1
C. n−k−1
D. n−k
gate2005-it algorithms graph-algorithms normal

Answer

1.5.18 Graph Algorithms: GATE2005-IT-15 top [Link]

In the following table, the left column contains the names of standard graph algorithms and the right
column contains the time complexities of the algorithms. Match each algorithm with its time
complexity.

A:O
(m log n)
B:O
1. Bellman-Ford algorithm
2. Kruskal’s algorithm (n3 )
3. Floyd-Warshall algorithm C:O
4. Topological sorting (nm)
D:O
(n + m)
A. 1→ C, 2 → A, 3 → B, 4 → D
B. 1→ B, 2 → D, 3 → C, 4 → A
C. 1→ C, 2 → D, 3 → A, 4 → B
D. 1→ B, 2 → A, 3 → C, 4 → D
gate2005-it algorithms graph-algorithms normal

Answer

1.5.19 Graph Algorithms: GATE2005-IT-84a top [Link]

A sink in a directed graph is a vertex i such that there is an edge from every vertex j ≠ i to i and
there is no edge from i to any other vertex. A directed graph G with n vertices is represented by its
adjacency matrix A, where A[i][j] = 1 if there is an edge directed from vertex i to j and 0
otherwise. The following algorithm determines whether there is a sink in the graph G.

i = 0;
do {
j = i + 1;
while ((j < n) && E1) j++;
if (j < n) E2;
} while (j < n);
flag = 1;
for (j = 0; j < n; j++)
if ((j! = i) && E3) flag = 0;
if (flag) printf("Sink exists");
else printf ("Sink does not exist");

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 51

Choose the correct expressions for E1 and E2

A. E1 : A[i][j] and E2 : i = j ;
B. E1 : !A[i][j] and E2 : i = j + 1 ;
C. E1 : !A[i][j] and E2 : i = j ;
D. E1 : A[i][j] and E2 : i = j + 1 ;

gate2005-it algorithms graph-algorithms normal

Answer

1.5.20 Graph Algorithms: GATE2005-IT-84b top [Link]

A sink in a directed graph is a vertex i such that there is an edge from every vertex j ≠ i to i and
there is no edge from i to any other vertex. A directed graph G with n vertices is represented by its
adjacency matrix A, where A[i][j] = 1 if there is an edge directed from vertex i to j and 0
otherwise. The following algorithm determines whether there is a sink in the graph G.

i = 0;
do {
j = i + 1;
while ((j < n) && E1) j++;
if (j < n) E2;
} while (j < n);
flag = 1;
for (j = 0; j < n; j++)
if ((j! = i) && E3) flag = 0;
if (flag) printf("Sink exists") ;
else printf ("Sink does not exist");

Choose the correct expression for E3

A. (A[i][j] && !A[j][i])
B. (!A[i][j] && A[j][i])
C. (!A[i][j] || A[j][i])
D. (A[i][j] || !A[j][i])

gate2005-it algorithms graph-algorithms normal

Answer

1.5.21 Graph Algorithms: GATE2006-12 top [Link]

To implement Dijkstra’s shortest path algorithm on unweighted graphs so that it runs in linear time,
the data structure to be used is:

A. Queue
B. Stack
C. Heap
D. B-Tree

gate2006 algorithms graph-algorithms easy

Answer

1.5.22 Graph Algorithms: GATE2006-48 top [Link]

Let T be a depth first search tree in an undirected graph G. Vertices u and ν are leaves of this tree
T . The degrees of both u and ν in G are at least 2. which one of the following statements is true?
G
© Copyright GATE Overflow. All rights reserved.
52 1 Algorithms (323)

A. There must exist a vertex w adjacent to both u and ν in G

B. There must exist a vertex w whose removal disconnects u and ν in G
C. There must exist a cycle in G containing u and ν
D. There must exist a cycle in G containing u and all its neighbours in G
gate2006 algorithms graph-algorithms normal

Answer

1.5.23 Graph Algorithms: GATE2006-IT-46 top [Link]

Which of the following is the correct decomposition of the directed graph given below into its
strongly connected components?

A. {P, Q, R, S} , {T } , {U} , {V }
B. {P, Q, R, S, T , V } , {U}
C. {P, Q, S, T , V } , {R} , {U}
D. {P, Q, R, S, T , U, V }

gate2006-it algorithms graph-algorithms normal

Answer

1.5.24 Graph Algorithms: GATE2006-IT-47 top [Link]

Consider the depth-first-search of an undirected graph with 3 vertices P , Q, and R. Let discovery
time d(u) represent the time instant when the vertex u is first visited, and finish time f(u)
represent the time instant when the vertex u is last visited. Given that

d(P) f(P)
=5 = 12
units units
d(Q) f(Q)
=6 = 10
units units
d(R) f(R)
= 14 = 18
unit units

Which one of the following statements is TRUE about the graph?

A. There is only one connected component

B. There are two connected components, and P and R are connected
C. There are two connected components, and Q and R are connected
D. There are two connected components, and P and Q are connected

gate2006-it algorithms graph-algorithms normal

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 53

Answer

1.5.25 Graph Algorithms: GATE2007-41 top [Link]

In an unweighted, undirected connected graph, the shortest path from a node S to every other node
is computed most efficiently, in terms of time complexity, by

A. Dijkstra’s algorithm starting from S.

B. Warshall’s algorithm.

C. Performing a DFS starting from S.

D. Performing a BFS starting from S.
gate2007 algorithms graph-algorithms easy

Answer

1.5.26 Graph Algorithms: GATE2007-5 top [Link]

Consider the DAG with V = {1, 2, 3, 4, 5, 6} shown below.

Which of the following is not a topological ordering?

A. 123456
B. 132456
C. 132465
D. 324165
gate2007 algorithms graph-algorithms

Answer

1.5.27 Graph Algorithms: GATE2007-IT-24 top [Link]

A depth-first search is performed on a directed acyclic graph. Let d[u] denote the time at which
vertex u is visited for the first time and f[u] the time at which the DFS call to the vertex u
terminates. Which of the following statements is always TRUE for all edges (u, v) in the graph ?

A. d[u] < d[v]

B. d[u] < f[v]
C. f[u] < f[v]
D. f[u] > f[v]

gate2007-it algorithms graph-algorithms normal

Answer

© Copyright GATE Overflow. All rights reserved.

54 1 Algorithms (323)

1.5.28 Graph Algorithms: GATE2007-IT-3, UGCNET-June2012-

III-34 top [Link]

Consider a weighted, undirected graph with positive edge weights and let uv be an edge in the
graph. It is known that the shortest path from the source vertex s to u has weight 53 and the
shortest path from s to v has weight 65. Which one of the following statements is always TRUE?

A. Weight (u, v) ≤ 12
B. Weight (u, v) = 12
C. Weight (u, v) ≥ 12
D. Weight (u, v) > 12

gate2007-it algorithms graph-algorithms normal ugcnetjune2012iii

Answer

1.5.29 Graph Algorithms: GATE2008-19 top [Link]

The Breadth First Search algorithm has been implemented using the queue data structure. One
possible order of visiting the nodes of the following graph is:

A. MNOPQR
B. NQMPOR
C. QMNPRO
D. QMNPOR
gate2008 normal algorithms graph-algorithms

Answer

1.5.30 Graph Algorithms: GATE2008-45 top [Link]

Dijkstra's single source shortest path algorithm when run from vertex a in the above graph,
computes the correct shortest path distance to

A. only vertex a
B. only vertices a, e, f, g, h
C. only vertices a, b, c, d
D. all the vertices

gate2008 algorithms graph-algorithms normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 55

1.5.31 Graph Algorithms: GATE2008-7 top [Link]

The most efficient algorithm for finding the number of connected components in an undirected graph
on n vertices and m edges has time complexity

A. Θ(n)
B. Θ(m)
C. Θ(m + n)
D. Θ(mn)

gate2008 algorithms graph-algorithms time-complexity normal

Answer

1.5.32 Graph Algorithms: GATE2008-IT-47 top [Link]

Consider the following sequence of nodes for the undirected graph given below:

1. abefdgc
2. abefcgd
3. adgebcf
4. adbcgef
A Depth First Search (DFS) is started at node a. The nodes are listed in the order they are first
visited. Which of the above is/are possible output(s)?

A. 1 and 3 only
B. 2 and 3 only
C. 2, 3 and 4 only
D. 1, 2 and 3 only
gate2008-it algorithms graph-algorithms normal

Answer

1.5.33 Graph Algorithms: GATE2009-13 top [Link]

Which of the following statement(s) is/are correct regarding Bellman-Ford shortest path algorithm?

P: Always finds a negative weighted cycle, if one exists.

Q: Finds whether any negative weighted cycle is reachable from the source.

A. P only
B. Q only
C. Both P and Q
D. Neither P nor Q

© Copyright GATE Overflow. All rights reserved.

56 1 Algorithms (323)

gate2009 algorithms graph-algorithms normal

Answer

1.5.34 Graph Algorithms: GATE2012-40 top [Link]

Consider the directed graph shown in the figure below. There are multiple shortest paths between
vertices S and T . Which one will be reported by Dijkstra’s shortest path algorithm? Assume that, in
any iteration, the shortest path to a vertex v is updated only when a strictly shorter path to v is
discovered.

A. SDT
B. SBDT
C. SACDT
D. SACET
gate2012 algorithms graph-algorithms normal

Answer

1.5.35 Graph Algorithms: GATE2013-19 top [Link]

What is the time complexity of Bellman-Ford single-source shortest path algorithm on a complete
graph of n vertices?

A. θ(n2 )
B. θ(n2 log n)
C. θ(n3 )
D. θ(n3 log n)

gate2013 algorithms graph-algorithms normal

Answer

1.5.36 Graph Algorithms: GATE2014-1-11 top [Link]

Let G be a graph with n vertices and m [Link] is the tightest upper bound on the running time
of Depth First Search on G, when G is represented as an adjacency matrix?

A. Θ(n)
B. Θ(n + m)
C. Θ(n2 )
D. Θ(m2 )

gate2014-1 algorithms graph-algorithms normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 57

1.5.37 Graph Algorithms: GATE2014-1-13 top [Link]

Consider the directed graph below given.

Which one of the following is TRUE?

A. The graph does not have any topological ordering.

B. Both PQRS and SRQP are topological orderings.
C. Both PSRQ and SPRQ are topological orderings.
D. PSRQ is the only topological ordering.

gate2014-1 graph-algorithms easy

Answer

1.5.38 Graph Algorithms: GATE2014-2-14 top [Link]

Consider the tree arcs of a BFS traversal from a source node W in an unweighted, connected,
undirected graph. The tree T formed by the tree arcs is a data structure for computing

A. the shortest path between every pair of vertices.

B. the shortest path from W to every vertex in the graph.
C. the shortest paths from W to only those nodes that are leaves of T.
D. the longest path in the graph.

gate2014-2 algorithms graph-algorithms normal

Answer

1.5.39 Graph Algorithms: GATE2014-3-13 top [Link]

Suppose depth first search is executed on the graph below starting at some unknown vertex.
Assume that a recursive call to visit a vertex is made only after first checking that the vertex has not
been visited earlier. Then the maximum possible recursion depth (including the initial call) is
_________.

gate2014-3 algorithms graph-algorithms numerical-answers normal

Answer

1.5.40 Graph Algorithms: GATE2015-1-45 top [Link]

Let G = (V , E) be a simple undirected graph, and s be a particular vertex in it called the source.
For x ∈ V , let d(x) denote the shortest distance in G from s to x. A breadth first search (BFS) is
(u, v) G
© Copyright GATE Overflow. All rights reserved.
58 1 Algorithms (323)

performed starting at s. Let T be the resultant BFS tree. If (u, v) is an edge of G that is not in T,
then which one of the following CANNOT be the value of d(u) − d(v)?

A. −1
B. 0
C. 1
D. 2
gate2015-1 algorithms graph-algorithms normal

Answer

1.5.41 Graph Algorithms: GATE2016-1-11 top [Link]

Consider the following directed graph:

The number of different topological orderings of the vertices of the graph is _____________.

gate2016-1 algorithms graph-algorithms normal numerical-answers

Answer

1.5.42 Graph Algorithms: GATE2016-2-11 top [Link]

Breadth First Search (BFS) is started on a binary tree beginning from the root vertex. There is a
vertex t at a distance four from the root. If t is the nth vertex in this BFS traversal, then the
maximum possible value of n is __________
gate2016-2 algorithms graph-algorithms normal numerical-answers

Answer

1.5.43 Graph Algorithms: GATE2016-2-41 top [Link]

In an adjacency list representation of an undirected simple graph G = (V , E) , each edge (u, v) has
two adjacency list entries: [v] in the adjacency list of u, and [u] in the adjacency list of v. These are
called twins of each other. A twin pointer is a pointer from an adjacency list entry to its twin. If
|E| = m and |V | = n , and the memory size is not a constraint, what is the time complexity of the
most efficient algorithm to set the twin pointer in each entry in each adjacency list?

A. Θ (n2 )
B. Θ (n + m)
C. Θ (m2 )
D. Θ (n4 )

gate2016-2 algorithms graph-algorithms normal

Answer

1.5.44 Graph Algorithms: GATE2017-1-26 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 59

Let G = (V , E) be any connected, undirected, edge-weighted graph. The weights of the edges in
E are positive and distinct. Consider the following statements:
I. Minimum Spanning Tree of G is always unique.
II. Shortest path between any two vertices of G is always unique.

Which of the above statements is/are necessarily true?

A. I only
B. II only
C. both I and II
D. neither I nor II

gate2017-1 algorithms graph-algorithms normal

Answer

1.5.45 Graph Algorithms: GATE2017-2-15 top [Link]

The Breadth First Search (BFS) algorithm has been implemented using the queue data structure.
Which one of the following is a possible order of visiting the nodes in the graph below?

A. MNOPQR
B. NQMPOR
C. QMNROP
D. POQNMR
gate2017-2 algorithms graph-algorithms

Answer

1.5.46 Graph Algorithms: Gate2000-2.19 top [Link]

Let G be an undirected graph. Consider a depth-first traversal of G, and let T be the resulting
depth-first search tree. Let u be a vertex in G and let v be the first new (unvisited) vertex visited
after visiting u in the traversal. Which of the following statement is always true?

A. {u, v} must be an edge in G, and u is a descendant of v in T

B. {u, v} must be an edge in G, and v is a descendant of u in T
C. If {u, v} is not an edge in G then u is a leaf in T
D. If {u, v} is not an edge in G then u and v must have the same parent in T

gate2000 algorithms graph-algorithms normal

Answer

1.5.47 Graph Algorithms: TIFR2013-B-5 top [Link]

Given a weighted directed graph with n vertices where edge weights are integers (positive, zero,
or negative), determining whether there are paths of arbitrarily large weight can be performed in
time

A. O(n)
O(n. log(n)) O(n)
© Copyright GATE Overflow. All rights reserved.
60 1 Algorithms (323)

B. O(n. log(n)) but not O(n)

C. O(n1.5 ) but not O(n log n)
D. O(n3 ) but not O(n1.5 )
E. O(2n ) but not O(n3 )

tifr2013 algorithms graph-algorithms

Answer

1.5.48 Graph Algorithms: TIFR2014-B-3 top [Link]

Consider the following directed graph.

Suppose a depth-first traversal of this graph is performed, assuming that whenever there is a
choice, the vertex earlier in the alphabetical order is to be chosen. Suppose the number of tree
edges is T , the number of back edges is B and the number of cross edges is C . Then

a. B = 1, C = 1 , and T = 4.
b. B = 0, C = 2 , and T = 4.
c. B = 2, C = 1 , and T = 3.
d. B = 1, C = 2 , and T = 3.
e. B = 2, C = 2 , and T = 1.
tifr2014 algorithms graph-algorithms

Answer

Answers: Graph Algorithms

1.5.1 Graph Algorithms: GATE1994-1.22 top [Link]


Selected Answer

Answer is B.
A. True.
B. False.
C. True.
D. True.

 11 votes -- Rajarshi Sarkar (34.1k points)

1.5.2 Graph Algorithms: GATE1994-24 top [Link]


Selected Answer

a. While adding vertex u to I it should not have an edge with any node in I .
b. The algorithm runs till V is empty (in O(n) time) and is checking u with each vertex v in set I
(in O(n) time). So, overall complexity O(n2 ).

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 61

 15 votes -- Rajarshi Sarkar (34.1k points)

1.5.3 Graph Algorithms: GATE1996-17 top [Link]


Selected Answer

DIJKSTRA(G, w, s)
1 INITIALIZE-SINGLE-SOURCE(G, s)
2 S=∅
3 Q = G. V
4 while Q ≠ ∅
5 u = EXTRACT-MIN(Q)
6 S = S ∪ {u}
7 for each vertex v ∈ G. Adj[u]
8 RELAX(u, v, w)
Correct Solutions:
(A).
⎧0 ∞ ∞ ∞ ∞ ∞ ⎫
Q = ⎨A , B , C , D , E , F ⎬
⎩ ⎭

First we visit A as it is the one with smallest distance 0

Relax operation updates distances to B,C,F as 6, 90, 70 respectively
⎧ 6 90 ∞ ∞ 70 ⎫
Q = ⎨B , C , D , E , F ⎬
⎩ ⎭
B is next visited as its distance is now 6
Relax operation updates distance to D as 41 + 6 = 47
⎧ 90 47 ∞ 70 ⎫
Q=⎨C , D, E , F ⎬
⎩ ⎭
D is next visited as its distance is now 47
Relax operation updates distance to C as 47 + 12 = 59
⎧ 59 ∞ 70 ⎫
Q=⎨C , E , F ⎬
⎩ ⎭
C is next visited as its distance is now 59
Relax operation updates distance to F as 59 + 10 = 69
⎧ ∞ 69 ⎫
Q=⎨E , F ⎬
⎩ ⎭
F is next visited as its distance is now 69
Relax operation updates distance to E as 69 + 15 = 84
⎧ 84 ⎫
Q=⎨E ⎬
⎩ ⎭
Finally E is visited.
So, the sequence of node visits are A, B, D, C, F, E

(B). A A−B−D−C−F−E
© Copyright GATE Overflow. All rights reserved.
62 1 Algorithms (323)

(B). Sequence of vertices in the shortest path from A to E: A − B − D − C − F − E

(C). Cost of the shortest path from A to E = 84.

 3 votes -- Manu Thakur (39.6k points)

1.5.4 Graph Algorithms: GATE1998-1.21, ISRO2008-16 top

[Link]

Selected Answer

Answer is (A) because Floyd Warshall algorithm is used to find all shortest paths which is a
dynamic programming approach.

 15 votes -- shashi shekhar (569 points)

1.5.5 Graph Algorithms: GATE2000-1.13 top [Link]


Selected Answer

Answer is C.
X - 3 DFS uses stack implicitly
Y - 2 BFS uses queue explicitly in Algo
Z - 1 Heap-Heapsort

 16 votes -- Akash Kanase (42.5k points)

1.5.6 Graph Algorithms: GATE2001-2.14 top [Link]


Selected Answer

Answer is (C).
BFS is used to count shortest path from source (If all path costs are 1)

Now, if u is visited before v it means 2 things:

1. Either u is closer to v, or,
2. If u & v are same distance from r, then our BFS algo chose to visit u before v.

 24 votes -- Akash Kanase (42.5k points)

1.5.7 Graph Algorithms: GATE2002-12 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 63

INITIALIZATION: For i = 1 ... n

{For j = 1 ... n
{ if a[i,j] = 0 then P[i,j] =infinite
// i.e. if there is no direct path then put infinite
else P[i,j] =a[i,j];
}
}
ALGORITHM:
For i = 1 ... n
{For j = 1 ... n
{For k = 1 ... n
{
P[i, j] = min( p[i,j] , p[i,k] + p[k,j])
};
}
}

Time complexity O(n3 )

This algorithm is for weighted graph but it will work for unweighted graph too because if
p[i, j] = 1, p[i, k] = 1 and p[k, j] = 1 then according to the algorithm
p[i, j] = min(p[i, j], p[i, k] + p[k, j]) = min(1, 2) = 1
And all the other cases are also satisfied.(like as if p[i, j] was 0 in last iteration nd there exist a
path via k)

 12 votes -- Saurav Kumar Gupta (2.2k points)

1.5.8 Graph Algorithms: GATE2003-21 top [Link]


Selected Answer

For GATE purpose, without actually applying DFS, we can answer by just seeing options.
In DFS, we go in depth first i.e., one node to another in depth first order.

Here, abfehg is not possible as we can not go from f to e directly.

Thus, option (D) is correct.

In all the other options we can reach directly from the node to the next node.

So, just visualize and do.

 12 votes -- Monanshi Jain (9.3k points)

1.5.9 Graph Algorithms: GATE2003-67 top [Link]


Selected Answer

After applying the shortest path algorithm, check cost of vertex from source to every vertex in G1 .
If G1 is connected all these costs must be 0 as edge weights of subgraph G1 is 0 and that should
be the shortest path. If cost is not 0, to at least one vertex in G1 (not necessarily G), then G1 is
disconnected.

© Copyright GATE Overflow. All rights reserved.

64 1 Algorithms (323)

Answer is B.

 38 votes -- Anurag Semwal (8k points)

1.5.10 Graph Algorithms: GATE2003-70 top [Link]


Selected Answer

D is correct.

Consider a graph with 2 nodes and one edge from V1 to V2 ,

Running the above algorithm will result in A being

A 1 2
1 1 2
2 1 2

Clearly options B and C are wrong. Since

1. A[1][1] and A[2][2] > n − 1 and there exists no Hamiltonian cycle. Hence invalid.
2. The longest path between V1 and V2 is 1, but A[1][2] is 2, which is invalid. And no path
between V2 and V1 yet A[2][1] = 1 // it should be max cost path between j and k, not
path length.

Hence A or D could be valid.

Now consider a graph with 2 nodes and two edges, one from V1 and V2 and other form V2 and
V1 . Running the above algorithm will result in A being
A 1 2
1 2 3
2 3 4

Hence option A is invalid, as A[i][j] can be > n

D is correct

 25 votes -- ryan sequeira (3.4k points)

1.5.11 Graph Algorithms: GATE2004-44 top [Link]


Selected Answer

Answer is (B). In Dijkstra's algorithm at each point we choose the smallest weight edge which
starts from any one of the vertices in the shortest path found so far and add it to the shortest path.

 18 votes -- gate_asp (749 points)

1.5.12 Graph Algorithms: GATE2004-81 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 65

There are two connected graphs G1 and G2 , with same vertices. In least case, both graphs will
have n − 1 edges with n vertices as both the given graphs are connected.
When we UNION both the graphs as G = G1 ∪ G2 then G will be having at most (2n − 2) edges
in the best case when both the graphs don't have any common edges between them (or) at least
more than n − 1 edges if they have few common edges between them as the intersection of these
two graphs is not connected.
Suppose if both the graphs G1 and G2 have exactly the same edges then their intersection will be
a connected graph.

A graph with
n vertices and more than n − 1 edges will definitely have a cycle.
Hence (B) is the correct option.

 23 votes -- Vijay Thakur (17.1k points)

Take a tree for example

A. False. Every vertex of tree(other than leaves) is a cut vertex.

B. True.
C. False. Without E in G1 and G2, G1UG2 has no bridge.
D. False. G1UG2, G1, G2 three graphs have same chromatic number of 2.

 26 votes -- srestha (87k points)

1.5.13 Graph Algorithms: GATE2004-IT-56 top [Link]


Selected Answer

Answer is D.

A and B produce disconnected components with the GIVEN order in options which is NEVER
allowed by prims's algorithm.

C produces connected component every instant a new edge is added BUT when first vertex is
chosen(first vertex is chosen randomly) first edge must be the minimum weight edge that is
chosen . Therefore, (A, D) MUST be chosen BEFORE (A, B). Therefore, C is FALSE.

 16 votes -- Sandeep_Uniyal (7.5k points)

1.5.14 Graph Algorithms: GATE2005-38 top [Link]


Selected Answer

Option (D) : Binary heap. |E| decrease key operations and each taking O (log |V |) time + |V |
O (log |V |)
© Copyright GATE Overflow. All rights reserved.
66 1 Algorithms (323)

extract-min operations each taking O (log |V |).

Option (B) : Fibonacci heap. |E| decrease key operations and each taking O(1) time + |V |
extract-min operations each taking O (log |V |).

Option (A) : Array. Finding min-vertex in each iteration takes O(V ) and this needs to be done
|V | times.
Binomial Heap is same as Binary heap here, as the critical operations are decrease key and
extract-min.

 38 votes -- Gate Keeda (19.6k points)

1.5.15 Graph Algorithms: GATE2005-82a top [Link]


Selected Answer

For 82a: The answer should be Option A because edge e is the lightest safe edge connecting X
and Y so the minimum spanning tree of G must contain e (Greedy and optimal choice).
While option (B) might seem correct but it is not always true. One such case is when G is not
connected therefore there might not be any path between s and t.
Since the question is about definitely TRUE, (B) is incorrect and (A) is the only correct option.

Lets say AC = 1 , CD = 2 ,BD = 3 and AB = 4

Then if s = A and t = B then AC is the lightest edge crossing X and Y where X = A and
Y = C, B, D
But clearly AC is not on the shortest path from A to B. The shortest path is AB = 4.

 20 votes -- chandan1223 (217 points)

1.5.16 Graph Algorithms: GATE2005-82b top [Link]


Selected Answer

Here answer should be A.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 67

Here shortest path will give 6.

Spanning tree contains edges of weights 2,3,4 so congestion in this case is max(2, 3, 4), that is,
4. For path s to t, overall congestion is max(3, 4) = 4 but total weight is 7.
Option C and D are I think not related to this question.

 13 votes -- papesh (25.7k points)

1.5.17 Graph Algorithms: GATE2005-IT-14 top [Link]


Selected Answer

Tree edges are those edges which appear in the final DFS forest. For example in case of connected
graph (i.e. resulting DFS forest containing only one tree), if we run DFS over it, edges belonging to
the resulting DFS tree are called tree edges.

Let us assume the graph has x number of connected (or strongly connected in case of a directed
graph) components. And assume 1st component has K1 tree edges, 2nd component has K2 tree
edges and xth component has Kx tree edges.

Such that K1 + K2 + K3 + … + Kx = K ( = total)

Or in other way we can imagine like, the final DFS forest has x trees and those trees are having
K1 , K2 , K3 , … , Kx edges respectively.
Now we know that a tree having Kx edges contains Kx + 1 nodes. Similarly a tree having
K1 edges contains K1 + 1 nodes, etc. and so on.
So, Summation of nodes in each tree =n
(K1 + 1) + (K2 + 1) + (K3 + 1) + … + (Kx + 1) = n ⟹ (K1 + K2 + K3 + … + Kx ) +

 36 votes -- Debashish Deka (56.7k points)

© Copyright GATE Overflow. All rights reserved.

68 1 Algorithms (323)

1.5.18 Graph Algorithms: GATE2005-IT-15 top [Link]


Selected Answer

1. Bellman-Ford algorithm ⟹ option(C), O(nm). Assuming n as edges , m as vertices, for

every vertex we relax all edges. m ∗ n , O(mn).
2. Kruskal’s algorithm ⟹ Remaining Option (A) : O(m log n).
3. Floyd-Warshall algorithm ⟹ option (B), Dynamic Programming Algo, O(N 3 ).
4. Topological sorting ⟹ option(D), boils down to DFS, O(V + E) .

Answer
(A).

 15 votes -- Akash Kanase (42.5k points)

1.5.19 Graph Algorithms: GATE2005-IT-84a top [Link]


Selected Answer

If there is a sink in the graph, the adjacency matrix will contain all 1's (except diagonal) in one
column and all 0's (except diagonal) in the corresponding row of that vertex. The given algorithm
is a smart way of doing this as it finds the sink in O(n) time complexity.

The first part of the code, is finding if there is any vertex which doesn't have any outgoing edge to
any vertex coming after it in adjacency matrix. The smart part of the code is E2 , which makes
rows skip when there is no edge from i to it, making it impossible for them to form a sink. This is
done through

E1 : A[i][j]
and
E2 : i = j ;
E1 makes sure that there is no edge from i to j and i is a potential sink till A[i][j] becomes 1. If
A[i][j] becomes 1, i can no longer be a sink, similarly all previous j can also not be a sink (as
there was no edge from i to them and a sink requires an edge from all other vertices). Now, the
next potential candidate for sink is j. So, in E2 , we must make i = j.

So, answer is (C)

For E3 , [Link]

 28 votes -- Arjun Suresh (350k points)

1.5.20 Graph Algorithms: GATE2005-IT-84b top [Link]


Selected Answer

If there is a sink in the graph, the adjacency matrix will contain all 1s (except diagonal) in one
column and all 0s (except diagonal) in the corresponding row of that vertex. The given algorithm is
a smart way of doing this as it finds the sink in O(n) time complexity.

The first part of the code, is finding if there is any vertex which does not have any outgoing edge
to any vertex coming after it in adjacency matrix. The smart part of the code is E2 , which makes

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 69

rows skip when there is no edge from i to it, making it impossible them to form a sink. This is done
through

E1 : !A[i][j]
and
E2 : i = j ;
E1 makes sure that there is no edge from i to j and i is a potential sink till A[i][j] becomes 1. If
A[i][j] becomes 1, i can no longer be a sink, similarly all previous j can also not be a sink (as
there was no edge from i to them and a sink requires an edge from all other vertices). Now, the
next potential candidate for sink is j. So, in E2 , we must make i = j.

Now, the loop breaks when we found a potential sink- that is a vertex which does not have any
outgoing edge to any coming after it in adjacency matrix. So, if the column in which this vertex
comes is all 1s and the row is all 0s (except diagonal), this is the sink. Otherwise there is no sink in
the graph. So, E3 is checking this condition.

But in the code flag is used for storing the state that sink is present or not. And as per the usage
of flag in code, by default sink is considered present. So, the condition in E3 must make flag = 0,
if the found i is not a sink. So, the condition should be:

A[i][j] || !A[j][i]
So, (D) is the answer.

 32 votes -- Arjun Suresh (350k points)

1.5.21 Graph Algorithms: GATE2006-12 top [Link]


Selected Answer

Answer is A: Queue

We can find single source shortest path in unweighted graph by using Breadth First Search (BFS)
algorithm by using "Queue" data structure , in time O(m + n) (i.e. linear with respect to the
number of vertices and edges. )

 42 votes -- Mithlesh Upadhyay (5.8k points)

1.5.22 Graph Algorithms: GATE2006-48 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

70 1 Algorithms (323)

One diagram, which is eliminating option A, B, C.

Hence D is the answer.

 29 votes -- Ahwan Mishra (10.5k points)

1.5.23 Graph Algorithms: GATE2006-IT-46 top [Link]


Selected Answer

Here the answer is B.

A graph is said to be strongly connected if every vertex is reachable from every other vertex.

The strongly connected component is always maximal that is if x is strongly connected component
there should not exist another strongly connected component which contains x.

If we take R as a strongly connected component but which is part of PQRS and PQRS is part of
PQRSV T .

 22 votes -- papesh (25.7k points)

1.5.24 Graph Algorithms: GATE2006-IT-47 top [Link]


Selected Answer

As seen in question, after 10 we have to go for p again and since p is finished and then r is started
so r must be disconnected because if there is edges from q to r then r must be visited before q
and p end.

D is answer.

 15 votes -- Prashant Singh (59.8k points)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 71

1.5.25 Graph Algorithms: GATE2007-41 top [Link]


Selected Answer

Dijkastra and Warshall 's algorithm used only for weighted graph.

Both DFS and BFS can be used for finding path between 2 vertices in undirected and
unweighted graph but BFS can only give the shortest path as concerned in given question.
So, BFS is answer.

Note : Finding only path(DFS) and finding shortest path(BFS) matters a lot.

ust Read:

[Link]
BFS-What-are-the-applications-and-downsides-of-each

 19 votes -- Rajesh Pradhan (22.6k points)

1.5.26 Graph Algorithms: GATE2007-5 top [Link]


Selected Answer

Go with vertex with indegree 0. Remove the vertex with all edges going from it . Follow that
procedure.

We see 3 cannot come at first because indegree is not 0. So, D is answer here.
ALL other options are in Topological order.

Only 1 and 4 order matters for this question.

 17 votes -- Prashant Singh (59.8k points)

1.5.27 Graph Algorithms: GATE2007-IT-24 top [Link]


Selected Answer

I'm gonna disprove all wrong options here:

A. d[u] < d[v], Counter Example ⟹ Well if we directly start DFS on V first, then I call DFS on
X which visits U .
B. d[u] < f[v], Counter example ⟹ Same as A
C. f[u] < f[v], Counter example ⟹ Same as A again

So, answer is D.

© Copyright GATE Overflow. All rights reserved.

72 1 Algorithms (323)

 20 votes -- Akash Kanase (42.5k points)

1.5.28 Graph Algorithms: GATE2007-IT-3, UGCNET-June2012-

III-34 top [Link]


Selected Answer

C. Weight (u, v) ≥ 12
If weight(u, v) < 12, then the min. weight of (s, v) =weight of (s, u)+ weight of
(u, v) = 53 + (< 12) will be less than 65.
 27 votes -- Arjun Suresh (350k points)

1.5.29 Graph Algorithms: GATE2008-19 top [Link]


Selected Answer

A. MNOPQR: If you try to run BFS, after M, you must traverse NQR (In some order). Here, P is
traversed before Q, which is wrong.
B. NQMPOR: This is also not BFS. P is traversed before O.
C. QMNPRO: Correct.
D. QMNPOR: Incorrect. Because R needs to be traversed before O.(Because M is ahead of N in
queue).

Answer :- C

 17 votes -- Akash Kanase (42.5k points)

1.5.30 Graph Algorithms: GATE2008-45 top [Link]


Selected Answer

D. all the vertices. Just simulate the Dijkstra's algorithm on it. Dijkstra's algorithm is not meant for
graphs with negative-edge-weight-cycle, but here it does give the correct shortest path.

 24 votes -- Arjun Suresh (350k points)

1.5.31 Graph Algorithms: GATE2008-7 top [Link]


Selected Answer

Run DFS to find connected components. Its time complexity is Θ(m + n), hence (C) is the
answer.

 19 votes -- Happy Mittal (11.2k points)

1.5.32 Graph Algorithms: GATE2008-IT-47 top [Link]


Selected Answer

Answer: B

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 73

1. After f is visited, c or g should be visited next. So, the traversal is incorrect.

4. After c is visited, e or f should be visited next. So, the traversal is incorrect.
2 and 3 are correct.

 13 votes -- Rajarshi Sarkar (34.1k points)

1.5.33 Graph Algorithms: GATE2009-13 top [Link]


Selected Answer

Bellman-ford Algorithm
−−−−−−−−−−−−−−−−−−−
Single source shortest Path O(V E)
Relax every edge once in each iteration


E × (V − 1) = E. V − E = O(V . E)

at max (V − 1) edges can be there

As we can see that the last step is the verification step. In that step, values remained unchanged.
If there was a negative edge weight cycle reachable from source, then at verification step also,
those values will be different from the values above.

In case the cycle is not reachable from source then we can see that they will be at ∞ distance(or
cost) from the source from the beginning till the last step. As take anything away from the ∞ it
will still be infinite.

But it can also be the case that there are some points which are not forming a cycle and are still
unreachable from source, those also will be at ∞ distance from the source from the beginning till
end.

Hence, we won't be able to make a distinction among the cycle and such vertices. Thus, we say
that this algorithm can detect negative edge weight cycles only if they are reachable from the
source.

© Copyright GATE Overflow. All rights reserved.

74 1 Algorithms (323)

Answer is option B

 20 votes -- Amar Vashishth (30.5k points)

1.5.34 Graph Algorithms: GATE2012-40 top [Link]


Selected Answer

Relaxation at every vertex is as follows

Note the next vertex is taken out here:

A B C D E F
S4
3(by ∞ 7 ∞ ∞ ∞
s)
B 4(by 7
∞ ∞ ∞ ∞
s) ∵ (4 + 3 also = 7)(S → d)
5
A 7 ∞ ∞ ∞
(S → B → A)
6
C 7 ∞ ∞
(S → B → C)

7 8
E ∞
(S → D) (S → A →

D 12(S → B → D) 8

E 12
T 12
Now We see for S to T its (S → A → C → E → T )
which is Option : D

 29 votes -- Kalpish Singhal (2.1k points)

1.5.35 Graph Algorithms: GATE2013-19 top [Link]


Selected Answer

Time complexity of Bellman-Ford algorithm is Θ(|V ||E|)

|V | is number of vertices and |E|
where
is number of edges. If the graph is complete, the value of |E| becomes Θ (|V | ) . So overall time
2

complexity becomes Θ (|V |3 ) . And given here is n vertices. So, the answer ends up to be
Θ (n3 ).
 35 votes -- Gate Keeda (19.6k points)

1.5.36 Graph Algorithms: GATE2014-1-11 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 75


Selected Answer

Depth First Search of a graph takes O(m + n) time when the graph is represented using
adjacency list. In adjacency matrix representation, graph is represented as an n ∗ n matrix. To do
DFS, for every vertex, we traverse the row corresponding to that vertex to find all adjacent
vertices (In adjacency list representation we traverse only the adjacent vertices of the vertex).
Therefore time complexity becomes O(n2 ).

 7 votes -- Divya Bharti (8.1k points)

1.5.37 Graph Algorithms: GATE2014-1-13 top [Link]


Selected Answer

The C option has been copied wrongly

C. Both PSRQ and SPRQ are topological orderings

i. Apply DFS by choosing P or S as starting vertices

ii. As the vertex gets a finishing time assign it to the head of a linked list
iii. The linked list is your required topological ordering

 21 votes -- Akshay Jindal (457 points)

1.5.38 Graph Algorithms: GATE2014-2-14 top [Link]


Selected Answer

BFS always has a starting node. It does not calculate shortest path between every pair but it
computes shortest path between W and any other vertex.

 26 votes -- Digvijay (54.9k points)

1.5.39 Graph Algorithms: GATE2014-3-13 top [Link]


Selected Answer

Total 21 nodes are there. 2 nodes require back track here in this question.
So, max recursion depth is 21 − 2 = 19
( Do DFS from extreme ends such that max recursion depth will occur i.e. take leftmost top node
as initial node for DFS as shown in below image)

Note:- Backtrack means it reduces recursion depth in stack.

© Copyright GATE Overflow. All rights reserved.

76 1 Algorithms (323)

 29 votes -- Rajesh Pradhan (22.6k points)

1.5.40 Graph Algorithms: GATE2015-1-45 top [Link]


Selected Answer

2 is the answer.
d(u) − d(v) = 0 is possible when both u and v have an edge from t and t is in the shortest path
from s to u or v.

d(u) − d(v) = 1 is possible when v and t are in the shortest path from s to u and both t and v
are siblings- same distance from s to both t and v causing t − u edge to be in BFS tree and not
v − u.
d(u) − d(v) = −1 is possible as explained above by interchanging u and v.
d(u) − d(v) = 2 is not possible. This is because on BFS traversal we either visit u first or v. Let's
take u first. Now, we put all neighbors of u on queue. Since v is a neighbour and v is not visited
before as assumed, d(v) will become d(u) + 1. Similarly, for v being visited first.

 30 votes -- Arjun Suresh (350k points)

1.5.41 Graph Algorithms: GATE2016-1-11 top [Link]


Selected Answer

Here, start with

a and end with
f.

a _ _ _ _
f
Blank spaces are to be filled with
b,
c,
d,
e such that
b comes before
c, and
d comes before
e.
Number of ways to arrange
b,
c,
d,
e such that
b comes before
c and

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 77

d comes before
e, will be =
4!/(2! ∗ 2!) = 6

 64 votes -- Abhilash Panicker (9.5k points)

In topological sorting all nodes are like tasks and edges show the dependency among the tasks.

Node i to j an edge is there means task i must complete before task j.(in the mean time some
other task may get complete after task i and before task j..but task i and j sequence need to be
maintained)

Here in following 6 ways all the 6 tasks can get completed.

 36 votes -- Rajesh Pradhan (22.6k points)

1.5.42 Graph Algorithms: GATE2016-2-11 top [Link]


Selected Answer

No of nodes at level 0(root) of tree ⇒1

No of nodes at level 1 of tree ⇒ 2
No of nodes at level 2 of tree ⇒ 4
No of nodes at level 3 of tree ⇒ 8
No of nodes at level 4 of tree ⇒ 16
Last node in level 4th is the node we are looking for ⇒ 1 + 2 + 4 + 8 + 16 ⇒ 31
 28 votes -- Akash Kanase (42.5k points)

1.5.43 Graph Algorithms: GATE2016-2-41 top [Link]


Selected Answer

Applying BFS on undirected graph gives you twin pointer. Visit every vertex level-wise. For every
vertex, fill adjacent vertex in the adjacency list. BFS takes Θ (n + m) time.

© Copyright GATE Overflow. All rights reserved.

78 1 Algorithms (323)

Take extra field for storing number of linked lists for particular vertex. Take extra m + n time( m
vertex and n edges).

So, option B is the answer.

 29 votes -- Prashant Singh (59.8k points)

1.5.44 Graph Algorithms: GATE2017-1-26 top [Link]


Selected Answer

Answer is A.

MST is not unique only when edges are not distinct. Here the edges are distinct. Be careful for
the keyword DISTINCT.
Shortest Path can be different even if the edges are distinct. Example is shown below. Shortest
path from A to C is not unique here.

 24 votes -- Ahwan Mishra (10.5k points)

1.5.45 Graph Algorithms: GATE2017-2-15

top [Link]


Selected Answer

In BFS, starting from a node, we traverse all node adjacent to it at first then repeat same for
next nodes.

Here you can see that only

(D) is following BFS sequence properly.
As per BFS, if we start from M then RQN (immediate neighbors of M) have to come after it
in any order but in A here, O comes in between. So, it is not BFS.
As per BFS, if we start from N then QMO has to come after it in any order but in B here, P
comes. So, it is not BFS.
As per BFS, if we start from Q then MNOP has to come after it in any order but in C here, R
comes. So, it is not BFS.

But D is following the sequences.

So, D is the correct answer.

 12 votes -- Aboveallplayer (18.4k points)

1.5.46 Graph Algorithms: Gate2000-2.19 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 79


Selected Answer

Let this be the DFS order of the tree, then,

u = D and v = F
So, we conclude that

1. It is not necessary that there is an edge between them.

2. If there is no edge then u must be leaf i.e. D is leaf here.
3. It is not always possible that u and v have same parent. But they have same ancestor.

 34 votes -- Prashant Singh (59.8k points)

1.5.47 Graph Algorithms: TIFR2013-B-5 top [Link]


Selected Answer

I think arbitrary large weights means having positive weight cycle.

So, Bellman Ford algorithm can be used.

O(V E)
Changing sign of weights of edges.

 5 votes -- papesh pathare (599 points)

1.5.48 Graph Algorithms: TIFR2014-B-3 top [Link]


Selected Answer

Since they said that whenever there is a choice we will have to select the node which is
alphabetically earlier, therefore we choose the starting node as A.

The tree then becomes A−B−E−C . Therefore number of tree edges is 3, that is, (T = 3)
.

Now, there is one cycle B − E − C , so, we will get a back edge from C to B while performing
DFS. Hence B = 1.

Now, D becomes disconnected node and it can only contribute in forming cross edge . There are 2
cross edges D − A , D − B. Therefore C = 2 .

Answer is Option D.

Correct me if am going wrong.

© Copyright GATE Overflow. All rights reserved.

80 1 Algorithms (323)

 19 votes -- Riya Roy(Arayana) (7.2k points)

1.6 Graph Connectivity(1) top

1.6.1 Graph Connectivity: GATE2018-43 top [Link]

Let G be a graph with 100! vertices!, with each vertex labelled by a distinct permutation od the
numbers 1, 2, ..., 100. There is an edge between vertices u and v if and only if the label of u can be
obtained by swapping two adjacent numbers in the label of v. Let y denote the degree of a vertex in
G, and z denote the number of connected components in G. Then y + 10z = ____
gate2018 algorithms graph-algorithms graph-connectivity numerical-answers

Answer

Answers: Graph Connectivity

1.6.1 Graph Connectivity: GATE2018-43 top [Link]


Selected Answer

Answer: 109

Explanation:

We have to find 2 things here, the degree of every vertex(which will be same for all vertices) and
number of connected components.

Instead of 100, let's solve this by taking lesser value, say 4.

With 4! vertices, each vertex is a permutation of {1, 2, 3, 4}. So, we have vertices like {1, 2, 3, 4},
{1, 3, 2, 4}, {4, 1, 3, 2}, ... etc.
Here {1, 2, 3, 4}will be connected with
{2, 1, 3, 4}
{1, 3, 2, 4}
{1, 2, 4, 3}
To get this list, just take 2 adjacent numbers and swap them. eg. {1, 2, 3, 4} swap 1 and 2 to get
{2, 1, 3, 4}.
The given 3 are the only permutations we can get by swapping only 2 adjacent numbers from
{1, 2, 3, 4}. So, the degree of vertex {1, 2, 3, 4} will be 3. Similarly for any vertex it's degree will
be 3.

Here we got "3" because we can chose any 3 pairs of adjacenet numbers. So, with n, we have
n − 1 adjacent pairs to swap. So, degree will be n-1.
In our question, degree will be 100 − 1 = 99
Now let's see how many connected components we have.

It will be 1. Why?
If one can reach from one vertex to any other vertex, then that means that the graph is

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 81

connected.

Now if we start with a vertex say {1, 2, 3, 4} we can reach to other vertex, say {4, 3, 2, 1} by the
following path:

{1234} -> {1243} -> {1423} -> {4123} -> {4132} -> {4312} -> {4321}
Just take two adjacent numbers and swap them. With this operation you can create any
permutation, from any given initial permutation.

This way you can show that from any given vertex we can reach any other vertex. This shows that
the graph is connected and the number of connected components is 1.

y = 99 and z = 1

y + 10z = 99 + 10*1 = 109

 26 votes -- Rishabh Gupta (13.9k points)

1.7 Greedy Algorithm(8) top

1.7.1 Greedy Algorithm: CMI2015-B-04 top [Link]

You are given n positive integers, d1 , d2 … dn , each greater than 0. Design a greedy algorithm to
test whether these integers correspond to the degrees of some n-vertex simple undirected graph
G = (V , E) . [A simple graph has no self-loops and at most one edge between any pair of vertices].
cmi2015 descriptive algorithms greedy-algorithm

Answer

1.7.2 Greedy Algorithm: GATE1999-2.20 top [Link]

The minimum number of record movements required to merge five files A (with 10 records), B (with
20 records), C (with 15 records), D (with 5 records) and E (with 25 records) is:
A. 165
B. 90
C. 75
D. 65
gate1999 algorithms normal greedy-algorithm

Answer

1.7.3 Greedy Algorithm: GATE2003-69 top [Link]

The following are the starting and ending times of activities A, B, C, D, E, F, G and H respectively
in chronological order: “as bs cs ae ds ce es fs be de gs ee fe hs ge he ” . Here, xs denotes the
starting time and xe denotes the ending time of activity X. We need to schedule the activities in a
set of rooms available to us. An activity can be scheduled in a room only if the room is reserved for
the activity for its entire duration. What is the minimum number of rooms required?

A. 3
B. 4
C. 5
D. 6
gate2003 algorithms normal greedy-algorithm

© Copyright GATE Overflow. All rights reserved.

82 1 Algorithms (323)

Answer

1.7.4 Greedy Algorithm: GATE2005-84a top [Link]

We are given 9 tasks T1 , T2 , … , T9 . The execution of each task requires one unit of time. We can
execute one task at a time. Each task Ti has a profit Pi and a deadline di . Profit Pi is earned if the
task is completed before the end of the dith unit of time.

Task T1 T2 T3 T4 T5 T6 T7 T8 T9
Profit 15 20 30 18 18 10 23 16 25
Deadline 7 2 5 3 4 5 2 7 3

Are all tasks completed in the schedule that gives maximum profit?

A. All tasks are completed

B. T1 and T6 are left out

C. T1 and T8 are left out

D. T4 and T6 are left out

gate2005 algorithms greedy-algorithm process-schedule normal

Answer

1.7.5 Greedy Algorithm: GATE2005-84b top [Link]

We are given 9 tasks T1 , T2 , … , T9 . The execution of each task requires one unit of time. We can
execute one task at a time. Each task Ti has a profit Pi and a deadline di . Profit Pi is earned if the
task is completed before the end of the dith unit of time.

Task T1 T2 T3 T4 T5 T6 T7 T8 T9
Profit 15 20 30 18 18 10 23 16 25
Deadline 7 2 5 3 4 5 2 7 3

What is the maximum profit earned?

A. 147
B. 165
C. 167
D. 175
gate2005 algorithms greedy-algorithm process-schedule normal

Answer

1.7.6 Greedy Algorithm: GATE2006-IT-48 top [Link]

The characters a to h have the set of frequencies based on the first 8 Fibonacci numbers as follows
a : 1, b : 1, c : 2, d : 3, e : 5, f : 8, g : 13, h : 21
A Huffman code is used to represent the characters. What is the sequence of characters
corresponding to the following code?

110111100111010

fdheg
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 83

A. fdheg
B. ecgdf
C. dchfg
D. fehdg

gate2006-it algorithms greedy-algorithm normal

Answer

1.7.7 Greedy Algorithm: GATE2007-76 top [Link]

have probabilities 12 , 14 , 18 , 16
1 1 1
Suppose the letters a, b, c, d, e, f , 32 , 32 , respectively.

Which of the following is the Huffman code for the letter a, b, c, d, e, f ?

A. 0, 10, 110, 1110, 11110, 11111
B. 11, 10, 011, 010, 001, 000
C. 11, 10, 01, 001, 0001, 0000
D. 110, 100, 010, 000, 001, 111
gate2007 algorithms greedy-algorithm normal

Answer

1.7.8 Greedy Algorithm: GATE2018-48 top [Link]

Consider the weights and values of items listed below. Note that there is only one unit of each item.

Item Weight (in Value (in

number Kgs) rupees)
1 10 60
2 7 28
3 4 20
4 2 24
The task is to pick a subset of these items such that their total weight is no more than 11 Kgs and
their total value is maximized. Moreover, no item may be split. The total value of items picked by an
optimal algorithm is denoted by Vopt . A greedy algorithm sorts the items by their value-to-weight
ratios in descending order and packs them greedily, starting from the first item in the ordered list.
The total value of items picked by the greedy algorithm is denoted by Vgreedy .

The value of Vopt − Vgreedy is ____

gate2018 algorithms greedy-algorithm numerical-answers

Answer

Answers: Greedy Algorithm

1.7.1 Greedy Algorithm: CMI2015-B-04 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

84 1 Algorithms (323)

1. Sort the degrees in non-increasing order.

2. Pick up the highest degree ( let us say it a), remove it from the list of degrees and subtract 1
from next a degrees in list.
3. Repeat Step 2 until :

If we get all 0 entries in list ⇒ Simple Graph exits

If we get a negative entry or not enough entries to subtract 1 in step 2 ⇒ Simple
Graph does not exist

Read More : [Link]

Let's take a example : 3, 2, 1, 2

Step 1 : Sort the degree sequence : 3, 2, 2, 1
Step 2: Pick 3, Remove 3 from list and from next 3 elements subtract 1, Result : (1, 1, 0)
Again repeat step 2 : select 1, Remove 1 and from next 1 subtract 1, Result : (0, 0, 0)
Thus, a simple graph exists for the following degree-sequence.

 15 votes -- Manish Joshi (27.9k points)

1.7.2 Greedy Algorithm: GATE1999-2.20 top [Link]


Selected Answer

Arrange files in increasing order of records:

5 10 15 20 25
D A C B E

75
30 45

C B E
15 15 20 25

D A
5 10
No. of movements = 15 + 30 + 45 + 75 = 165.
 23 votes -- Pooja Palod (31.3k points)

1.7.3 Greedy Algorithm: GATE2003-69 top [Link]


Selected Answer

Solution: B

The problem can be modeled as a graph coloring problem. Construct a graph with one node
corresponding to each activity A, B, C, D, E, F, G and H . Connect the activities that occur

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 85

between the start and end time of an activity. Now, the chromatic number of the graph is the
number of rooms required.

 28 votes -- Gowthaman Arumugam (1.4k points)

1.7.4 Greedy Algorithm: GATE2005-84a top [Link]


Selected Answer

Step -1 Sort the tasks in decreasing order of profit and if any conflict arises between two or more
tasks,resolve them by sorting them on basis of having greater deadline first(Because we have
more time to complete the task with greater deadline and same profit).

Step 2- Since Maximum deadline given is 7, so we consider we have 7 time slots ranging from
0 − 7 where a task Ti having deadline say 2 can be filled in slots either 0 − 1 or 1 − 2 and not
beyond 2 because this task has deadline of 2 time units, so this task has to be completed by
atmost time T = 2 .

Now according to question, since Each task completes in Unit time, so a single tasks takes only one
slot as shown.

Now Take the first task in the list i.e. T3 which has a deadline of 5, so it can be completed in
maximum 5 time units, so place it in slot 4 − 5 which is the maximum deadline by which this task
can be completed.

Task T9 with deadline 3 is similarly placed in slot 2 − 3.

Task T7 with deadline 2 is placed in slot 1 − 2.

© Copyright GATE Overflow. All rights reserved.

86 1 Algorithms (323)

Now for task T2 having deadline 2 can be placed in either 0 − 1 or 1 − 2 (Occupied by T7 ). So T2

will occupy slot 0 − 1.

Task T5 with deadline 4 is placed in slot 3 − 4.

Now comes task T4 which has deadline 3 can be put in slots 0 − 1 or 1 − 2 or 2 − 3 and not
beyond [Link], all such slots are occupied so
T4 will be left out.
Task T8 with deadline 7 goes in slot 6 − 7.
Task T1 with deadline 7 can be placed in slot 5 − 6.
Now all time slots are full.

So, Task
T6 will be left out.
So, option (d) is the answer.

 15 votes -- Ayush Upadhyaya (9k points)

1.7.5 Greedy Algorithm: GATE2005-84b top [Link]


Selected Answer

The most important statement in question is

each task requires one unit of time

This shows that we can greedily choose the better task and that should give us the optimal
solution. The best task would be the one with maximum profit. Thus we can sort the tasks based
on deadline and then profit as follows:

Task T7 T2 T9 T4 T5 T3 T6 T8 T1
Deadline 2 2 3 3 4 5 5 7 7
0 --T7 -- 1 -- T2 -- 2 -- T9 -- 3 -- T5 -- 4 -- T3 -- 5 -- T8 -- 6 -- T1 -- 7
so we know that T4 and T6 are left

so profit will not include T4 and T6 = 15 + 20 + 30 + 18 + 16 + 23 + 25 = 147

A is answer

 20 votes -- Prashant Singh (59.8k points)

1.7.6 Greedy Algorithm: GATE2006-IT-48 top [Link]


Selected Answer

Answer is A. Huffman's tree is as follows. The two least frequent characters are taken as the
children of a newly made node and the frequency of the newly made node is made equal to the
sum of those two child nodes. Then the same procedure is repeated till all nodes are finished.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 87

110111100111010 = 110111100111010 = fdheg

 31 votes -- Arjun Suresh (350k points)

1.7.7 Greedy Algorithm: GATE2007-76 top [Link]


Selected Answer

Based on the probabilities, we can say the probable frequency of the letters will be

16, 8, 4, 2, 1, 1
Now, the Huffman tree can be constructed as follows:

So, A is the answer for 76.

[Link]

 22 votes -- Arjun Suresh (350k points)

1.7.8 Greedy Algorithm: GATE2018-48 top [Link]


Selected Answer

Voptimal is clearly 60. You can go for brute force or by normal intuition you can get it.
Now solving for Vgreedy.

© Copyright GATE Overflow. All rights reserved.

88 1 Algorithms (323)

Item
Weight Value Value/Weight
name
1 10 60 6
2 7 28 4
3 4 20 5
4 2 24 12
Sort them in descending order of Value/Weight as per the question.

Item
Weight Value Value/Weight
name
4 2 24 12
1 10 60 6
3 4 20 5
2 7 28 4
Now start picking items.(Note: You cannot take a fraction of the given weight as per the
question). Max weight size is given as 11(Inclusive).

Item 4 is picked. Weight remaining = 11 − 2 = 9kg.

Item 1 cannot be picked as 10kg > 9kg.
Item 3 can be picked as 4kg < 9kg. Weight Remaining = 9 − 4 = 5kg
Item 2 cannot be picked as 7kg > 5kg.
So, item 4 and Item 3 are picked. Their values are 24 and 20 respetively.
So, Vgreedy = 24 + 20 = 44
Voptimal - Vgreedy= 60 − 44 = 16 answer.

 12 votes -- Ruturaj Mohanty (337 points)

1.8 Huffman Code(2) top

1.8.1 Huffman Code: GATE2007-77 top [Link]

have probabilities 12 , 14 , 18 , 16
1 1 1
Suppose the letters a, b, c, d, e, f , 32 , 32 , respectively.

What is the average length of the Huffman code for the letters a, b, c, d, e, f ?
A. 3
B. 2.1875
C. 2.25
D. 1.9375
gate2007 algorithms greedy-algorithm normal huffman-code

Answer

1.8.2 Huffman Code: GATE2017-2-50 top [Link]

A message is made up entirely of characters from the set X = {P, Q, R, S, T } . The table of
probabilities for each of the characters is shown below:

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 89

Character Probability
P 0.22
Q 0.34
R 0.17
S 0.19
T 0.08
T otal 1.00
If a message of 100 characters over X is encoded using Huffman coding, then the expected length
of the encoded message in bits is ______.

gate2017-2 huffman-code numerical-answers algorithms

Answer

Answers: Huffman Code

1.8.1 Huffman Code: GATE2007-77 top [Link]


Selected Answer

Answer should be D.
Letter Probability
a 1/2
b 1/4
c 1/8
d 1/16
e 1/32
f 1/32

Prefix Code Code Length

a=0 1

b=10 2

c=110 3

d=1110 4

e=11110 5

© Copyright GATE Overflow. All rights reserved.

90 1 Algorithms (323)

f=11111 5

Avg length
1 1 1 1 1 1 16+16+12+8+5+5
= 2 ×1+ 4 ×2+ 8 ×3+ 16 ×4+ 32 ×5+ 32 ×5 = 32 = 1.9375

 23 votes -- sonam vyas (14.8k points)

1.8.2 Huffman Code: GATE2017-2-50 top [Link]


Selected Answer

X = {P, Q, R, S, T }

∴ Expected length of encoded message

= (22 × 2) + (34 × 2) + (17 × 3) + (19 × 2) + (8 × 3)bits = 44 + 68 + 51 + 38 + 24bits = 225

 6 votes -- Akash Dinkar (23.8k points)

1.9 Identify Function(37) top

1.9.1 Identify Function: CMI2010-B-07b top [Link]

A finite sequence of bits is represented as a list with values from the set {0, 1}. For example,
[0, 1, 0], [1, 0, 1, 1], . . . .
[ ] denotes the empty list, and [b] is the list consisting of one bit b. The function length(l) returns
the length (number of bits) in the list l. For a nonempty list l, head(l) returns the first element of l,
tail(l)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 91

and tail(l) returns the list obtained by removing the first element from l. The operator ++ denotes
list concatenation.
For example:

head([0, 1, 0]) = 0, tail([0, 1, 0]) = [1, 0],

head([1]) = 1, tail([1]) = [ ], and
[0, 1, 0] + +[1] = [0, 1, 0, 1].
Consider the following functions:
op takes as input two bits and returns a bit.
op(a,b)
if (a = b) return(0)
else return(1)
endif

mystery1 takes as input two lists and returns a list.

mystery1(s,t)
if (length(s) != length(t)) then return(t)
else if (length(s) = 0) then return(s)
else return([op(head(s),head(t))] ++ mystery1(tail(s),tail(t)))
endif
endif

mystery2 takes as input two lists and outputs a list.

mystery2(s,t)
if (length(t) = 0) then return(s)
else return( mystery1(mystery2(s,tail(t)),mystery2(s,tail(t))))
endif

i. Suppose s = t = 110100100. What are the first two bits of mystery2(s, t)?

descriptive cmi2010 algorithms identify-function

Answer

1.9.2 Identify Function: CMI2013-A-09 top [Link]

The below question is based on the following program.

procedure mystery (A : array [1..100] of int)

int i,j,position,tmp;
begin
for j := 1 to 100 do
position := j;
for i := j to 100 do
if (A[i] > A[position]) then
position := i;
endfor
tmp := A[j];
A[j] := A[position];
A[position] := tmp;
endfor
end

When the procedure terminates, the array A has been:

A. Reversed
B. Sorted in descending order
C. Left unaltered
D. Sorted in ascending order

cmi2013 algorithms identify-function

© Copyright GATE Overflow. All rights reserved.

92 1 Algorithms (323)

Answer

1.9.3 Identify Function: GATE1990-11b top [Link]

The following program computes values of a mathematical function f(x). Determine the form of
f(x).
main ()
{
int m, n; float x, y, t;
scanf ("%f%d", &x, &n);
t = 1; y = 0; m = 1;
do
{
t *= (-x/m);
y += t;
} while (m++ < n);
printf ("The value of y is %f", y);
}

gate1990 descriptive algorithms identify-function

Answer

1.9.4 Identify Function: GATE1991-03-viii top [Link]

Choose the correct alternatives (more than one may be correct) and write the corresponding letters
only:

Consider the following Pascal function:

Function X(M:integer):integer;
Var i:integer;
Begin
i := 0;
while i*i < M
do i:= i+1
X := i
end

The function call X(N), if N is positive, will return

−−
A. ⌊√N ⌋
−−
B. ⌊√N ⌋ + 1
−−
C. ⌈√N ⌉
−−
D. ⌈√N ⌉ + 1
E. None of the above

gate1991 algorithms easy identify-function

Answer

1.9.5 Identify Function: GATE1993-7.4 top [Link]

What does the following code do?

var a, b: integer;
begin
a:=a+b;

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 93

b:=a-b;
a:a-b;
end;

A. exchanges a and b
B. doubles a and stores in b
C. doubles b and stores in a
D. leaves a and b unchanged
E. none of the above

gate1993 algorithms identify-function easy

Answer

1.9.6 Identify Function: GATE1994-6 top [Link]

What function of x, n is computed by this program?

Function what(x, n:integer): integer:
Var
value : integer
begin
value := 1
if n > 0 then
begin
if n mod 2 =1 then
value := value * x;
value := value * what(x*x, n div 2);
end;
what := value;
end;

gate1994 algorithms identify-function normal

Answer

1.9.7 Identify Function: GATE1995-1.4 top [Link]

In the following Pascal program segment, what is the value of X after the execution of the program
segment?

X := -10; Y := 20;
If X > Y then if X < 0 then X := abs(X) else X := 2*X;

A. 10
B. −20
C. −10
D. None

gate1995 algorithms identify-function easy

Answer

1.9.8 Identify Function: GATE1995-2.3 top [Link]

Assume that X and Y are non-zero positive integers. What does the following Pascal program
segment do?

while X <> Y do

© Copyright GATE Overflow. All rights reserved.

94 1 Algorithms (323)

if X > Y then
X := X - Y
else
Y := Y - X;
write(X);

A. Computes the LCM of two numbers

B. Divides the larger number by the smaller number

C. Computes the GCD of two numbers

D. None of the above

gate1995 algorithms identify-function normal

Answer

1.9.9 Identify Function: GATE1995-4 top [Link]

a. Consider the following Pascal function where A and B are non-zero positive integers. What is the
value of GET (3, 2)?

function GET(A,B:integer): integer;

begin
if B=0 then
GET:= 1
else if A < B then
GET:= 0
else
GET:= GET(A-1, B) + GET(A-1, B-1)
end;

b. The Pascal procedure given for computing the transpose of an N × N, (N > 1) matrix A of
integers has an error. Find the error and correct it. Assume that the following declaration are
made in the main program

const
MAXSIZE=20;
type
INTARR=array [1..MAXSIZE,1..MAXSIZE] of integer;
Procedure TRANSPOSE (var A: INTARR; N : integer);
var
I, J, TMP: integer;
begin
for I:=1 to N – 1 do
for J:=1 to N do
begin
TMP:= A[I, J];
A[I, J]:= A[J, I];
A[J, I]:= TMP
end
end;

gate1995 algorithms identify-function normal

Answer

1.9.10 Identify Function: GATE1998-2.12 top [Link]

What value would the following function return for the input x = 95?

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 95

Function fun (x:integer):integer;

Begin
If x > 100 then fun = x – 10
Else fun = fun(fun (x+11))
End;

A. 89
B. 90
C. 91
D. 92
gate1998 algorithms recursion identify-function normal

Answer

1.9.11 Identify Function: GATE1999-2.24 top [Link]

Consider the following C function definition

int Trial (int a, int b, int c)
{
if ((a>=b) && (c<b)) return b;
else if (a>=b) return Trial(a, c, b);
else return Trial(b, a, c);
}

The functional Trial:

A. Finds the maximum of a, b, and c

B. Finds the minimum of a, b, and c
C. Finds the middle number of a, b, c
D. None of the above

gate1999 algorithms identify-function normal

Answer

1.9.12 Identify Function: GATE2000-2.15 top [Link]

Suppose you are given an array s[1....n] and a procedure reverse (s, i, j) which reverses the order
of elements in s between positions i and j (both inclusive). What does the following sequence do,
where 1 ⩽ k ⩽ n :

reverse (s, 1, k);

reverse (s, k+1, n);
reverse (s, 1, n);

A. Rotates s left by k positions

B. Leaves s unchanged
C. Reverses all elements of s
D. None of the above

gate2000 algorithms normal identify-function

Answer

1.9.13 Identify Function: GATE2003-1 top [Link]

C
© Copyright GATE Overflow. All rights reserved.
96 1 Algorithms (323)

Consider the following C function.

For large values of y, the return value of the function f best approximates

float f,(float x, int y) {

float p, s; int i;
for (s=1,p=1,i=1; i<y; i++) {
p *= x/i;
s += p;
}
return s;
}

A. xy
B. ex
C. ln(1 + x)
D. xx
gate2003 algorithms identify-function normal

Answer

1.9.14 Identify Function: GATE2003-88 top [Link]

In the following C program fragment, j, k, n and TwoLog_n are integer variables, and A is an array
of integers. The variable n is initialized to an integer ⩾ 3, and TwoLog_n is initialized to the value of
2∗ ⌈log2 (n)⌉
for (k = 3; k <= n; k++)
A[k] = 0;
for (k = 2; k <= TwoLog_n; k++)
for (j = k+1; j <= n; j++)
A[j] = A[j] || (j%k);
for (j = 3; j <= n; j++)
if (!A[j]) printf("%d", j);

The set of numbers printed by this program fragment is

A. {m ∣ m ≤ n, (∃i) [m = i!]}
B. {m ∣ m ≤ n, (∃i) [m = i2 ]}
C. {m ∣ m ≤ n, m is prime}
D. { }

gate2003 algorithms identify-function normal

Answer

1.9.15 Identify Function: GATE2004-41 top [Link]

Consider the following C program

main()
{
int x, y, m, n;
scanf("%d %d", &x, &y);
/* Assume x>0 and y>0*/
m = x; n = y;
while(m != n)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 97

{
if (m > n)
m = m-n;
else
n = n-m;
}
printf("%d", n);
}

The program computes

A. x + y using repeated subtraction

B. x mod y using repeated subtraction

C. the greatest common divisor of x and y

D. the least common multiple of x and y

gate2004 algorithms normal identify-function

Answer

1.9.16 Identify Function: GATE2004-42 top [Link]

What does the following algorithm approximate? (Assume m > 1, ϵ > 0 ).

x = m;
y = 1;
While (x-y > ϵ)
{
x = (x+y)/2;
y = m/x;
}
print(x);

A. log m
B. m2
1
C. m2
1
D. m3
gate2004 algorithms identify-function normal

Answer

1.9.17 Identify Function: GATE2005-31 top [Link]

Consider the following C-program:

void foo (int n, int sum) {

int k = 0, j = 0;
if (n == 0) return;
k = n % 10; j = n/10;
sum = sum + k;
foo (j, sum);
printf ("%d,",k);
}

int main() {
int a = 2048, sum = 0;
foo(a, sum);

© Copyright GATE Overflow. All rights reserved.

98 1 Algorithms (323)

printf("%d\n", sum);
}

What does the above program print?

A. 8, 4, 0, 2, 14
B. 8, 4, 0, 2, 0
C. 2, 0, 4, 8, 14
D. 2, 0, 4, 8, 0
gate2005 algorithms identify-function recursion normal

Answer

1.9.18 Identify Function: GATE2005-IT-57 top [Link]

What is the output printed by the following program?

#include <stdio.h>

int f(int n, int k) {

if (n == 0) return 0;
else if (n % 2) return f(n/2, 2*k) + k;
else return f(n/2, 2*k) - k;
}

int main () {
printf("%d", f(20, 1));
return 0;
}

A. 5
B. 8
C. 9
D. 20
gate2005-it algorithms identify-function normal

Answer

1.9.19 Identify Function: GATE2006-50 top [Link]

A set X can be represented by an array x[n] as follows:

x [i] = {
1 if i ∈ X
0 otherwise
Consider the following algorithm in which x, y, and z are Boolean arrays of size n:
algorithm zzz(x[], y[], z[]) {
int i;
for(i=0; i<n; ++i)
z[i] = (x[i] ∧ ~y[i]) ∨ (~x[i] ∧ y[i]);
}

The set Z computed by the algorithm is:

A. (X ∪ Y )
B. (X ∩ Y )
(X − Y ) ∩ (Y − X)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 99

C. (X − Y ) ∩ (Y − X)
D. (X − Y ) ∪ (Y − X)

gate2006 algorithms identify-function normal

Answer

1.9.20 Identify Function: GATE2006-53 top [Link]

Consider the following C-function in which a[n] and b[m] are two sorted integer arrays and
c[n + m] be another integer array,
void xyz(int a[], int b [], int c []){
int i,j,k;
i=j=k=0;
while ((i<n) && (j<m))
if (a[i] < b[j]) c[k++] = a[i++];
else c[k++] = b[j++];
}

Which of the following condition(s) hold(s) after the termination of the while loop?

i. j < m, k = n + j − 1 and a[n − 1] < b[j] if i = n

ii. i < n, k = m + i − 1 and b[m − 1] ≤ a[i] if j = m
A. only (i)
B. only (ii)
C. either (i) or (ii) but not both
D. neither (i) nor (ii)

gate2006 algorithms identify-function normal

Answer

1.9.21 Identify Function: GATE2006-IT-52 top [Link]

The following function computes the value of (m

n) correctly for all legal values m and n (
m ≥ 1, n ≥ 0 and m > n)
int func(int m, int n)
{
if (E) return 1;
else return(func(m -1, n) + func(m - 1, n - 1));
}

In the above function, which of the following is the correct expression for E?

A. (n == 0)||(m == 1)
B. (n = = 0) && (m = = 1)
C. (n == 0)||(m == n)
D. (n = = 0) && (m = = n)

gate2006-it algorithms identify-function normal

Answer

1.9.22 Identify Function: GATE2008-IT-82 top [Link]

© Copyright GATE Overflow. All rights reserved.

100 1 Algorithms (323)

Consider the code fragment written in C below :

void f (int n)
{
if (n <=1) {
printf ("%d", n);
}
else {
f (n/2);
printf ("%d", n%2);
}
}

What does f(173) print?

A. 010110101
B. 010101101
C. 10110101
D. 10101101
gate2008-it algorithms recursion identify-function normal

Answer

1.9.23 Identify Function: GATE2008-IT-83 top [Link]

Consider the code fragment written in C below :

void f (int n)
{
if (n <= 1) {
printf ("%d", n);
}
else {
f (n/2);
printf ("%d", n%2);
}
}

Which of the following implementations will produce the same output for f(173) as the above code?

P1 P2

void f (int n)
{
void f (int n)
if (n <=1) {
{
printf ("%d", n);
if (n/2) {
}
f(n/2);
else {
}
printf ("%d", n%2);
printf ("%d", n%2);
f (n/2);
}
}
}

A. Both P1 and P2
B. P2 only
C. P1 only
D. Neither P1 nor P2

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 101

gate2008-it algorithms recursion identify-function normal

Answer

1.9.24 Identify Function: GATE2009-18 top [Link]

Consider the program below:

#include <stdio.h>
int fun(int n, int *f_p) {
int t, f;
if (n <= 1) {
*f_p = 1;
return 1;
}
t = fun(n-1, f_p);
f = t + *f_p;
*f_p = t;
return f;
}

int main() {
int x = 15;
printf("%d/n", fun(5, &x));
return 0;
}

The value printed is:

A. 6
B. 8
C. 14
D. 15
gate2009 algorithms recursion identify-function normal

Answer

1.9.25 Identify Function: GATE2010-35 top [Link]

What is the value printed by the following C program?

#include<stdio.h>

int f(int *a, int n)

{
if (n <= 0) return 0;
else if (*a % 2 == 0) return *a+f(a+1, n-1);
else return *a - f(a+1, n-1);
}

int main()
{
int a[] = (12, 7, 13, 4, 11, 6);
printf("%d", f(a, 6));
return 0;
}

A. −9
B. 5
C. 15
D. 19

© Copyright GATE Overflow. All rights reserved.

102 1 Algorithms (323)

gate2010 algorithms recursion identify-function normal

Answer

1.9.26 Identify Function: GATE2011-48 top [Link]

Consider the following recursive C function that takes two arguments.

unsigned int foo(unsigned int n, unsigned int r) {

if (n>0) return ((n%r) + foo(n/r, r));
else return 0;
}

What is the return value of the function foo when it is called as foo(345, 10)?
A. 345
B. 12
C. 5
D. 3
gate2011 algorithms recursion identify-function normal

Answer

1.9.27 Identify Function: GATE2011-49 top [Link]

Consider the following recursive C function that takes two arguments.

unsigned int foo(unsigned int n, unsigned int r) {

if (n>0) return ((n%r) + foo(n/r, r));
else return 0;
}

What is the return value of the function foo when it is called as foo(513, 2)?
A. 9
B. 8
C. 5
D. 2
gate2011 algorithms recursion identify-function normal

Answer

1.9.28 Identify Function: GATE2013-31 top [Link]

Consider the following function:

int unknown(int n){

int i, j, k=0;
for (i=n/2; i<=n; i++)
for (j=2; j<=n; j=j*2)
k = k + n/2;
return (k);

The return value of the function is

A. Θ(n2 )
2
Θ( log n)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 103

B. Θ(n2 log n)
C. Θ(n3 )
D. Θ(n3 log n)

gate2013 algorithms identify-function normal

Answer

1.9.29 Identify Function: GATE2014-1-41 top [Link]

Consider the following C function in which size is the number of elements in the array E:

int MyX(int *E, unsigned int size)

{
int Y = 0;
int Z;
int i, j, k;

for(i = 0; i< size; i++)

Y = Y + E[i];

for(i=0; i < size; i++)

for(j = i; j < size; j++)
{
Z = 0;
for(k = i; k <= j; k++)
Z = Z + E[k];
if(Z > Y)
Y = Z;
}
return Y;
}

The value returned by the function MyX is the

A. maximum possible sum of elements in any sub-array of array E.

B. maximum element in any sub-array of array E.
C. sum of the maximum elements in all possible sub-arrays of array E.
D. the sum of all the elements in the array E.

gate2014-1 algorithms identify-function normal

Answer

1.9.30 Identify Function: GATE2014-2-10 top [Link]

Consider the function func shown below:

int func(int num) {

int count = 0;
while (num) {
count++;
num>>= 1;
}
return (count);
}

The value returned by func(435) is ________

gate2014-2 algorithms identify-function numerical-answers easy

© Copyright GATE Overflow. All rights reserved.

104 1 Algorithms (323)

Answer

1.9.31 Identify Function: GATE2014-3-10 top [Link]

Let A be the square matrix of size n × n. Consider the following pseudocode. What is the expected
output?

C=100;
for i=1 to n do
for j=1 to n do
{
Temp = A[i][j]+C;
A[i][j] = A[j][i];
A[j][i] = Temp -C;
}
for i=1 to n do
for j=1 to n do
output (A[i][j]);

A. The matrix A itself

B. Transpose of the matrix A
C. Adding 100 to the upper diagonal elements and subtracting 100 from lower diagonal elements of
A
D. None of the above

gate2014-3 algorithms identify-function easy

Answer

1.9.32 Identify Function: GATE2015-1-31 top [Link]

Consider the following C function.

int fun1 (int n) {

int i, j, k, p, q = 0;
for (i = 1; i < n; ++i)
{
p = 0;
for (j = n; j > 1; j = j/2)
++p;
for (k = 1; k < p; k = k * 2)
++q;
}
return q;
}

Which one of the following most closely approximates the return value of the function fun1?

A. n3
B. n(log n)2
C. n log n
D. n log(log n)

gate2015-1 algorithms normal identify-function

Answer

1.9.33 Identify Function: GATE2015-2-11 top [Link]

Consider the following C function.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 105

int fun(int n) {
int x=1, k;
if (n==1) return x;
for (k=1; k<n; ++k)
x = x + fun(k) * fun (n-k);
return x;
}

The return value of fun(5) is ______.

gate2015-2 algorithms identify-function recurrence normal numerical-answers

Answer

1.9.34 Identify Function: GATE2015-3-49 top [Link]

Suppose c = ⟨c[0], … , c[k − 1]⟩ is an array of length k, where all the entries are from the set
{0, 1}. For any positive integers a and n, consider the following pseudocode.
DOSOMETHING (c, a, n)

z←1
for i ← 0 to k − 1
do z ← z2 mod n
if c[i]=1
then z ← (z × a) mod n
return z

I f k = 4, c = ⟨1, 0, 1, 1⟩, a = 2, and n = 8, then the output of DOSOMETHING(c, a, n) is

_______.

gate2015-3 algorithms identify-function normal numerical-answers

Answer

1.9.35 Identify Function: TIFR2010-B-24 top [Link]

Consider the following program operating on four variables u, v, x, y , and two constants X and Y.
x, y, u, v:= X, Y, Y, X;
While (x ≠ y)
do
if (x > y) then x, v := x - y, v + u;
else if (y > x) then y, u:= y - x, u + v;
od;
print ((x + y) / 2); print ((u + v) / 2);

Given X > 0 ∧ Y > 0 , pick the true statement out of the following:
A. The program prints gcd(X, Y ) and the first prime larger than both X and Y .
B. The program prints gcd(X, Y ) followed by lcm(X, Y ).
C. The program prints gcd(X, Y ) followed by 12 × lcm(X, Y ) .
D. The program prints 12 × gcd(X, Y ) followed by 12 × lcm(X, Y ) .
E. The program does none of the above.

tifr2010 algorithms identify-function

Answer

© Copyright GATE Overflow. All rights reserved.

106 1 Algorithms (323)

1.9.36 Identify Function: TIFR2014-B-2 top [Link]

Consider the following code.

def brian(n):
count = 0

while ( n ! = 0 )
n = n & ( n-1 )
count = count + 1

return count

Here n is meant to be an unsigned integer. The operator & considers its arguments in binary and
computes their bit wise AND. For example, 22 & 15 gives 6, because the binary (say 8-bit)
representation of 22 is 00010110 and the binary representation of 15 is 00001111, and the bit-wise
AND of these binary strings is 00000110, which is the binary representation of 6. What does the
function brian return?

a. The highest power of 2 dividing n, but zero if n is zero.

b. The number obtained by complementing the binary representation of n.
c. The number of ones in the binary representation of n.
d. The code might go into an infinite loop for some n.
e. The result depends on the number of bits used to store unsigned integers.

tifr2014 algorithms identify-function

Answer

1.9.37 Identify Function: TIFR2017-A-12 top [Link]

Consider the following program modifying an n × n square matrix A:

for i=1 to n:
for j=1 to n:
temp=A[i][j]+10
A[i][j]=A[j][i]
A[j][i]=temp-10
end for
end for

Which of the following statements about the contents of matrix A at the end of this program must
be TRUE?

A. the new A is the transpose of the old A

B. all elements above the diagonal have their values increased by 10 and all the values below have
their values decreased by 10
C. all elements above the diagonal have their values decreased by 10 and all the values below have
their values increased by 10
D. the new matrix A is symmetric, that is, A[i][j] = A[j][i] for all 1 ≤ i, j ≤ n
E. A remains unchanged

tifr2017 algorithms identify-function

Answer

Answers: Identify Function

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 107

1.9.1 Identify Function: CMI2010-B-07b top [Link]

Here, the op function takes two bits as input and outputs their XOR(⊕).
Let x1 and x2 be the two input bits in the function op.

Then, the output is x3 s.t. x3 = x1 ⊕ x2 → (output of op function)

Now, the mystery 1 function takes two lists s and t as input and outputs another list u (say) such
that

u = s⊕t → (output of mystery 1 function)

and the output of mystery 2 function is


u = s ⊕ s ⊕ s ⊕ s … ⊕ s 2length(t) times

i. Taking XOR of s, 2length(t) times does not change the length(s). So, our answer will be
length(s).
ii. In this problem the two bits of length s are 11, so taking the XOR of 1, even number of times
gives the output 0. So, our answer is 00.

 3 votes -- Kushagra Chatterjee (8.2k points)

1.9.2 Identify Function: CMI2013-A-09 top [Link]


Selected Answer

Answer is B. Sorted in descending order ( selection sorting algorithm is used ).

 8 votes -- Dhananjay Kumar Sharma (25.7k points)

1.9.3 Identify Function: GATE1990-11b top [Link]


Selected Answer

The returned value is

∑ni=1 (−1)n xi!

i

 2 votes -- Arjun Suresh (350k points)

1.9.4 Identify Function: GATE1991-03-viii top [Link]


Selected Answer

For N = 9 , it returns 3.
For N = 10 it returns 4.
For N = 16 it returns 4.
For N = 17 it returns 5.

© Copyright GATE Overflow. All rights reserved.

108 1 Algorithms (323)

So answer should be C.

 10 votes -- Taymiyyah Bhat (3.5k points)

1.9.5 Identify Function: GATE1993-7.4 top [Link]


Selected Answer

Answer is simply A i.e. it swaps the values of the two.. Take any two values for A and B. and
perform the given operations over them.

 16 votes -- Gate Keeda (19.6k points)

1.9.6 Identify Function: GATE1994-6 top [Link]


Selected Answer

answer - xn
 7 votes -- Ankit Rokde (9k points)

1.9.7 Identify Function: GATE1995-1.4 top [Link]


Selected Answer

Answer of X remains unchanged. As the if condition becomes false.

X := -10

Answer is C . This is classic example of if − else issue. Always else matches for nesting to the
closest if in C Programming & Pascal .
[Link]

if (x>y)
{
if (x<0)
x=abs(x)
else
x=2*x
}

 21 votes -- Akash Kanase (42.5k points)

1.9.8 Identify Function: GATE1995-2.3 top [Link]


Selected Answer

Answer: C

Let X = 3 and Y = 7 .
1st pass: X = 3 , Y = 4
2nd pass: X = 3 , Y = 1
3rd pass: X = 2 , Y = 1
4th pass: X = 1 , Y = 1
write (X), which writes 1.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 109

Ref: [Link]

 11 votes -- Rajarshi Sarkar (34.1k points)

1.9.9 Identify Function: GATE1995-4 top [Link]


Selected Answer

a. =3
b.

begin
for I:=2 to N do
for J:=1 to ( I-1) do
begin
TMP:= A[I, J];
A[I, J]:= A[J, I];
A[J, I]:= TMP
end

Should be the condition...

 8 votes -- papesh (25.7k points)

1.9.10 Identify Function: GATE1998-2.12 top [Link]


Selected Answer

Value returned by fun(95) = fun(fun(106))

= fun(96)
= fun(fun(107))
= fun(97)
= fun(fun(108))
= fun(98)
= fun(fun(109))
= fun(99)
= fun(fun(110))
= fun(100)
= fun(fun(111))
= fun(101) = 91.
 23 votes -- Digvijay (54.9k points)

1.9.11 Identify Function: GATE1999-2.24 top [Link]


Selected Answer

abc Return
The final return statement is
1 1 1 c < b, so this never returns.
Answer D.

 37 votes -- Arjun Suresh (350k points)

© Copyright GATE Overflow. All rights reserved.

110 1 Algorithms (323)

1.9.12 Identify Function: GATE2000-2.15 top [Link]


Selected Answer

Answer is A.

Effect of the above 3 reversals for any K is equivalent to left rotation of the array of size n by k.
Let , S[1......7]
1 2 3 4 5 6 7

so,n = 7 ,k = 2
reverse (S, 1, 2) we get [2, 1, 3, 4, 5, 6, 7]
reverse (S, 3, 7) we get [2, 1, 7, 6, 5, 4, 3]
reverse (S, 1, 7) we get [3, 4, 5, 6, 7, 1, 2]
Hence, option (A) rotates s left by k positions and is correct.

 29 votes -- Kalpna Bhargav (3.3k points)

1.9.13 Identify Function: GATE2003-1 top [Link]


Selected Answer

A simplified version of the given program can be written as:

float f(float x, int y) {

float p=1, s=1;
int i;
for (i=1; i<y; i++) {
p = p * (x/i);
s = s + p;
}
return s;
}

We can take p = 1, s = 1 initialization outside of for loop because there is no condition checking in
for loop involving p, s .
i p = p*(x/i) s = s+p
1 x 1+x
x2
1+x
2 2 2
+ x2
1+x
2
3 x3 + x2
6
x3
+ 6

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 111

1+x
2
+ x2
4 x4
x3
24 + 6
x4
+ 24
xn
n n! ex
∞
xn x2 x3 x4 xn
ex =∑ =1+x+ + + +…+
n=0
n! 2 6 24 n!

Hence, option B is answer.

 16 votes -- Rajesh Pradhan (22.6k points)

1.9.14 Identify Function: GATE2003-88 top [Link]


Selected Answer

The nested loop is taking all integers from 2 to 2 ∗ log2 n. Take all their non-multiples before n,
and make the corresponding entry in A as 1. For example, for 2, and n = 10, A[3], A[5], A[7],
and A[9]are made 1. Similarly for 3, 4, ... till 2 ∗ log n. So, if any entry A[p] is 1 means it must
be a multiple of 2, 3, . . . . 2 ∗ log2 n , which is (2 log n)! and is greater than n. So, for no index p,
A[p] will be 0. So, answer is D.
Suppose the line

A[j] = A[j] || (j%k);

is replaced with

A[j] = A[j] || !(j%k);

Now, the nested loop is taking all integers from 2 to log2 n , take all their multiples before n, and
make the corresponding entry in A as 1. For example, for 2, and n = 10, A[4], A[6], A[8] and
A[10] are made 1. Similarly for 3, 4, . . . till 2 ∗ log n . So, for all non-prime indices of A, we will
have a 1, and for prime indices we have a 0. And we print i if A[j] is 0 meaning j is prime.

 20 votes -- Arjun Suresh (350k points)

1.9.15 Identify Function: GATE2004-41 top [Link]


Selected Answer

It is an algorithm for gcd computation.

Here, while loop executes until m = n.
We can test by taking any two numbers as m, n.
Answer will be (C).

Ref: [Link]

© Copyright GATE Overflow. All rights reserved.

112 1 Algorithms (323)

 9 votes -- srestha (87k points)

1.9.16 Identify Function: GATE2004-42 top [Link]


Selected Answer

By putting y = m/x into x = (x + y)/2

x = (x + m/x)/2

⟹ 2x2 = x2 + m
⟹ x = m1/2
We can also check by putting 2 or 3 different values also.
 30 votes -- gate_asp (749 points)

1.9.17 Identify Function: GATE2005-31 top [Link]


Selected Answer

Option is D.

foo is printing the lowest digit. But the printf inside it is after the recursive call. This forces the
output to be in reverse order

2, 0, 4, 8
The final value sum printed will be 0 as C uses pass by value and hence the modified value inside
foo won't be visible inside main.

 16 votes -- anshu (3.3k points)

1.9.18 Identify Function: GATE2005-IT-57 top [Link]


Selected Answer

See the following calling sequence:

Hence, answer is option C.

 18 votes -- Rajesh Pradhan (22.6k points)

1.9.19 Identify Function: GATE2006-50 top

[Link]

Selected Answer

Option (D)

In the given algorithm the for loop contains a logical expression

z[i] = (x[i] ∧ ~y[i]) ∨ (~x[i] ∧ y[i]);

The equivalent set representation of a given logical expression if we assume

z[i] = Z, x[i] = X, y[i] = Y

© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 113

z[i] = Z, x[i] = X, y[i] = Y then

Z = (X ∧ ¬Y ) ∨ (¬X ∧ Y )
⟹ Z = (X − Y ) ∪ (Y − X)[∵ A ∧ ¬B = A − B]

 13 votes -- Prasanna Ranganathan (4.7k points)

1.9.20 Identify Function: GATE2006-53 top [Link]


Selected Answer

The while loop adds elements from a and b (whichever is smaller) to c and terminates when either
of them exhausts. So, when loop terminates either i = n or j = m .

Suppose i = n. This would mean all elements from array a are added to c => k must be
incremented by n. c would also contain j elements from array b. So, number of elements in c
would be n + j and hence k = n + j .

Similarly, when j = m, k = m + i .
Hence, option (D) is correct. (Had k started from −1 and not 0 and we used + + k inside loop,
answer would have been option (C))

 41 votes -- Arjun Suresh (350k points)

1.9.21 Identify Function: GATE2006-IT-52 top [Link]


Selected Answer

Answer: C

Because (m0 ) = 1 and (nn) = 1.

 21 votes -- Rajarshi Sarkar (34.1k points)

1.9.22 Identify Function: GATE2008-IT-82 top [Link]


Selected Answer

Answer: D

The function prints the binary equivalent of the number n.

Binary equivalent of 173 is 10101101.

 16 votes -- Rajarshi Sarkar (34.1k points)

1.9.23 Identify Function: GATE2008-IT-83 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

114 1 Algorithms (323)

Here, P1 and P2 will print opposite in direction as shown in diagram.

And given code fragment will print like P1 and not like P2
Hence, answer will be (C).

 21 votes -- srestha (87k points)

1.9.24 Identify Function: GATE2009-18 top [Link]


Selected Answer

The answer is B.
Let the address of x be 1000.
1.f(5, 1000) = 8
2.f(4, 1000) = 5
3.f(3, 1000) = 3
4.f(2, 1000) = 2
5.f(1, 1000) = 1.
The evaluation is done from 5 to 1. Since recursion is used.
 11 votes -- Gate Keeda (19.6k points)

1.9.25 Identify Function: GATE2010-35 top [Link]


Selected Answer

Suppose int array takes 4 bytes for each element and stored at base address 100.
Follow below image. Red color shows the return value.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 115

So, 15 is the answer.

 12 votes -- Rajesh Pradhan (22.6k points)

It will print
12 + (7 − (13 − (4 + (11 − (6 + 0)))))
= 12 + (7 − (13 − (4 + (11 − 6)))))
= 12 + 7 − 13 + 9
= 15
 23 votes -- gatecse (18k points)

1.9.26 Identify Function: GATE2011-48 top [Link]


Selected Answer

Red color represents return values.

Answer is
12.

 17 votes -- Rajesh Pradhan (22.6k points)

1.9.27 Identify Function: GATE2011-49 top [Link]


Selected Answer

The function returns the sum of digits in a binary representation of the given number

so 1+0+0+0+0+0+0+0+0+1 =2
 16 votes -- Sandeep_Uniyal (7.5k points)

© Copyright GATE Overflow. All rights reserved.

116 1 Algorithms (323)

1.9.28 Identify Function: GATE2013-31 top [Link]


Selected Answer

The outer loop is running for n/2 times and inner loop is running for log2 n times (each iteration
doubles j and j stops at n means log2 n times j loop will iterate).

Now in each iteration k is incremented by n/2. So, overall k will be added n/2 ∗ logn ∗ n/2 with
an initial value of 0. So, final value of k will be Θ(n2 logn)

 41 votes -- Arjun Suresh (350k points)

1.9.29 Identify Function: GATE2014-1-41 top [Link]


Selected Answer

Answer is (A) maximum possible sum of elements in any sub-array of array E.

int MyX ( int * E, unsigned int size )

{
int Y= 0;
int z;
int i, j,k;

//calculate sum of the elements of the array E and stores it in Y

for i 0;i<size;i++)
Y = Y+E[i];

//calculate the sum of all possible subaarays

//(starting from postion 0..n-1)
for (i=0;i<size;i++)
for(j=i;j<size ;j++)
{
z = 0;
for(k=i; k<=j;k++)
z=z+E[k];

//checks whether sum of elements of each subarray is greater

//than the current max, if so, then assign it to currentmax
if(z>Y)
Y = z;
}
//ultimately returns the maximum possible sum of elements
//in any sub array of given array E
return Y;
}

 25 votes -- Kalpna Bhargav (3.3k points)

1.9.30 Identify Function: GATE2014-2-10 top [Link]


Selected Answer

Answer is 9.

435 − (110110011)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 117

435 − (110110011)
num >>= 1; implies a num is shifted one bit right in every while loop [Link] loop is
executed 9 times successfully and 10th time num is zero.

So count is incremented 9 times.

Note:

Shifting a number "1"bit position to the right will have the effect of dividing by 2:

8 >> 1 = $4 // In binary: (00001000) >> 1 = (00000100)

 23 votes -- Prasanna Ranganathan (4.7k points)

1.9.31 Identify Function: GATE2014-3-10 top [Link]


Selected Answer

A.
In the computation of given pseudo code for each row and column of Matrix A, each upper
triangular element will be interchanged by its mirror image in the lower triangular and after that
the same lower triangular element will be again re-interchanged by its mirror image in the upper
triangular, resulting the final computed Matrix A same as input Matrix A.

 24 votes -- Gate Keeda (19.6k points)

1.9.32 Identify Function: GATE2015-1-31 top [Link]


Selected Answer

i loop is executing n times. j loop is executing log n times for each i, and so value of p is log n. k
loop is executing log p times, which is log log n times for each iteration of i. In each of these q is
incremented. So, over all iterations of i, q will be incremented n log log n times. So, D choice.

 41 votes -- Arjun Suresh (350k points)

1.9.33 Identify Function: GATE2015-2-11 top [Link]


Selected Answer

fun(1) = 1;
fun(2) = 1 + fun(1) ∗ fun(1) = 1 + 1 = 2;
fun(3) = 1 + fun(1) ∗ fun(2) + fun(2) ∗ fun(1) = 5;
fun(4) = 1 + fun(1) ∗ fun(3) + fun(2) ∗ fun(2) + fun(3) ∗ fun(1) = 1 + 5 + 4 + 5 = 15;
fun(5) = 1 + fun(1) ∗ fun(4) + fun(2) ∗ fun(3) + fun(3) ∗ fun(2) + fun(4) ∗ fun(1) = 1

More formal way:

© Copyright GATE Overflow. All rights reserved.

118 1 Algorithms (323)

The recurrence relation is

1, n = 1
f(n) = {
1 + ∑n−1
i=1 f(i) × f(n − i), n > 1
f(1) = 1
f(2) = 1 + f(1). f(1) = 1 + 1.1 = 2
f(3) = 1 + f(1). f(2) + f(2). f(1) = 1 + 1.2 + 2.1 = 5
f(4) = 1 + f(1). f(3) + f(2). f(2) + f(3). f(2) = 1 + 1.5 + 2.2 + 5.1 = 15
f(5) = 1 + f(1). f(4) + f(2). f(3) + f(3). f(2) + f(4). f(1) = 1 + 1.15 + 2.5 + 5.2 + 15.1 =

 53 votes -- Arjun Suresh (350k points)

1.9.34 Identify Function: GATE2015-3-49 top [Link]


Selected Answer

Initially k = 4, c = [1, 0, 1, 1], a = 2, n = 8.

Now let's iterate through the function step by step :

z=1 (at the start of do-something)

i = 0 (start of external for loop)

In the do loop

z=1∗1 (non zero value so considered as true and continue)

c[0] = 1, so in the if clause, z = 1 ∗ 2%8 = 2

In the do loop

z = 2 ∗ 2%8 = 4 (since now z = 2 ) (non zero value so considered as true and continue)
c[0] = 1, so in the if clause, z = 4 ∗ 2%8 = 0
Now no need to check further :

Reason : All the operations that update Z are multiplicative operations and hence the value of Z
will never change from 0.

 26 votes -- Tamojit Chatterjee (2.3k points)

1.9.35 Identify Function: TIFR2010-B-24 top [Link]

It prints with gcd(x, y) and lcm(x, y).

Consider x, y, u, v = 17, 3, 3, 17.
X = 14, v = 20
X = 11, v = 23

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 119

X = 8 , v = 26
X = 5 , v = 29
X = 2 , v = 32
Y = 1 , u = 35
X = 1 , v = 67
This is the value obtained.

Lastly, printing (x + y)/2 and (v + u)/2 gives 1 and 51.

 8 votes -- zambus (279 points)

1.9.36 Identify Function: TIFR2014-B-2 top [Link]


Selected Answer

Option C. It returns no of 1's in binary representation of n.

Here, n&(n − 1) reset rightmost bit of n in each iteration.

For example,

Suppose n = 15 = 00001111(binary)
n − 1 = 14(00001110)
00001111
^ 00001110
---------------------
00001110

 6 votes -- Avdhesh Singh Rana (3.2k points)

1.9.37 Identify Function: TIFR2017-A-12 top [Link]


Selected Answer

[Link] i=1 to n:
2. for j=1 to n:
3. temp=A[i][j]+10
4. A[i][j]=A[j][i]
5. A[j][i]=temp-10
6. end for
[Link] for

The 3, 4, 5 lines swap A[j][i] and A[i][j].

The same variables are swapped twice. For eg when: i = 5, j = 10. A[10][5] and A[5][10] will be
swapped. They will be swapped again when i = 10 , j = 5 .

Two times swap of same elements will lead to A remaining unchanged.

Hence, E is correct.

© Copyright GATE Overflow. All rights reserved.

120 1 Algorithms (323)

 6 votes -- tarun_svbk (1.5k points)

1.10 Minimum Maximum(4) top

1.10.1 Minimum Maximum: GATE2014-1-39 top [Link]

The minimum number of comparisons required to find the minimum and the maximum of 100
numbers is ________
gate2014-1 algorithms numerical-answers normal minimum-maximum

Answer

1.10.2 Minimum Maximum: TIFR2014-B-10 top [Link]

Given a set of n distinct numbers, we would like to determine both the smallest and the
largest number. Which of the following statements is TRUE?

A. These two elements can be determined using O (log100 n) comparisons.

B. O (log100 n) comparisons do not suffice, however these two elements can be determined using
n + O(log n) comparisons.
C. n + O(log n) comparisons do not suffice, however these two elements can be determined using
3⌈n/2⌉ comparisons.
D. 3⌈n/2⌉ comparisons do not suffice, however these two elements can be determined using
2(n − 1) comparisons.
E. None of the above.

tifr2014 algorithms minimum-maximum

Answer

1.10.3 Minimum Maximum: TIFR2014-B-6 top [Link]

Consider the problem of computing the minimum of a set of n distinct numbers. We choose
a permutation uniformly at random (i.e., each of the n! permutations of ⟨1, . . . . , n⟩ is chosen with
probability (1/n!) and we inspect the numbers in the order given by this permutation. We maintain
a variable MIN that holds the minimum value seen so far. MIN is initialized to ∞ and if we see a
value smaller than MIN during our inspection, then MIN is updated. For example, in the
inspection given by the following sequence, MIN is updated four times.

5942680317
What is the expected number of times MIN is updated?

A. O(1)
B. Hn = ∑ni=1 1/i
−
C. √n
D. n/2
E. n
tifr2014 algorithms minimum-maximum

Answer

1.10.4 Minimum Maximum: TIFR2014-B-9 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 121

Given a set of n distinct numbers, we would like to determine the smallest three numbers in this
set using comparisons. Which of the following statements is TRUE?

A. These three elements can be determined using O (log2 n) comparisons.

B. O (log2 n) comparisons do not suffice, however these three elements can be determined using
n + O(1) comparisons.
C. n + O(1) comparisons do not suffice, however these three elements can be determined using
n + O(log n) comparisons.
D. n + O(log n) comparisons do not suffice, however these three elements can be determined
using O(n) comparisons.
E. None of the above.

tifr2014 algorithms minimum-maximum

Answer

Answers: Minimum Maximum

1.10.1 Minimum Maximum: GATE2014-1-39 top [Link]


Selected Answer

We can solve this question by using Tournament Method Technique -

1. To find the smallest element in the array will take n − 1 comparisons = 99.
2. To find the largest element -

a. After the first round of Tournament , there will be exactly n/2 numbers = 50 that will loose the
round.
b. So, the biggest looser (the largest number) should be among these 50 [Link] find the
largest number will take n/2 − 1 comparisons = 49.

Total 99 + 49 = 148.

 37 votes -- Harsh181996 (4k points)

1.10.2 Minimum Maximum: TIFR2014-B-10 top [Link]


Selected Answer

I think answer will be C.

To be accurate, it will need 3n/2 − 2 comparisons .

 8 votes -- Pranay Datta (10.2k points)

1.10.3 Minimum Maximum: TIFR2014-B-6 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

122 1 Algorithms (323)

Let us consider 3 numbers {1, 2, 3}

We will consider the permutation along with min no of times MIN is updated .

Permutation : No of times MIN updated (Minimum)

1, 2, 3 1
1, 3, 2 1
2, 1, 3 2
2, 3, 1 2
3, 1, 2 2
3, 2, 1 3
Total number of times MIN updated is : 11.
Average no of times MIN updated is : (11/6)
Now going by the options i am getting B .

H3 = 1 + 1/2 + 1/3 = 11/6 .

H3 is the answer and that is option B .

 17 votes -- Riya Roy(Arayana) (7.2k points)

1.10.4 Minimum Maximum: TIFR2014-B-9 top [Link]


Selected Answer

Option (C) is correct. Reason is as follows :

Here, at first level we are Given n elements, out of which we have to find smallest 3 numbers.
We compare 2 − 2 elements as shown in figure & get n/2 elements at Second level.
Note: Minimum element is in these
n/2 elements.
n/2
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 123
n/2
So, comparisons for this is n/2.
Similarly for next level we have n/4 Comparisons & n/2 elements and so on.
Total Comparisons till now is n/2 + n/4 + n/8+. . . . +4 + 2 + 1 = (2log n − 1) = n − 1 {Use
G.P. sum}

Now we have to get smallest 3.

We have 1st smallest at last level already =>0 Comparison for this.

=> 2nd & 3rd smallest can be found in O(log n) time as shown below:

Minimum Element must have descended down from some path from top to Bottom.

=> SQUARES represent Candidates for 2nd minimum.

Every element that is just below m1(first minimum) is a candidate for second minimum.

So, O(log n) Comparisons for finding second smallest.

=> Similarly for 3rd minimum we get O(log n) time.. As, every element that is just below 1st &
2nd minimum is a candidate for 3rd minimum.

 18 votes -- Himanshu Agarwal (15.3k points)

1.11 Minimum Spanning Trees(2) top

1.11.1 Minimum Spanning Trees: CMI2012-B-05b top [Link]

Given an undirected weighted graph G = (V , E) with non-negative edge weights, we can compute
a minimum cost spanning tree T = (V , E ′ ) . We can also compute, for a given source vertex sϵV ,
the shortest paths from s to every other vertex in V . We now increase the weight of every edge in
the graph by 1. Are the following true or false, regardless of the structure of G? Give a
mathematically sound argument if you claim the statement is true or a counterexample if the
statement is false.

i. All the shortest paths from s to the other vertices are unchanged.

cmi2012 descriptive algorithms graph-algorithms minimum-spanning-trees

© Copyright GATE Overflow. All rights reserved.

124 1 Algorithms (323)

Answer

1.11.2 Minimum Spanning Trees: GATE2018-47 top [Link]

Consider the following undirected graph G:

Choose a value for x that will maximize the number of minimum weight spanning trees (MWSTs) of
G. The number of MWSTs of G for this value of x is ____.
gate2018 algorithms graph-algorithms minimum-spanning-trees numerical-answers

Answer

Answers: Minimum Spanning Trees

1.11.1 Minimum Spanning Trees: CMI2012-B-05b top [Link]


Selected Answer

The given statement "All the shortest paths from s to the other vertices are unchanged." is false .
From the above graph it is clear that the shortest path from S to D is
S ⟹ A ⟹ B ⟹ C ⟹ D and the cost is 6.
Now, we increment the edge cost of all the edges by 1.

After incrementation, the shortest path from S to D gets changed. Now the shortest path becomes
S ⟹ E ⟹ D and shortest path cost is 9. The above graph is the proof.

 5 votes -- balaeinstein (1.3k points)

1.11.2 Minimum Spanning Trees: GATE2018-47 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 125

Number of possible MSTs increase, when we have multiple edges with same edge weights.

To maximize the number of MST, x should be 5.

In the question, number of MST is asked for the value of X.
So, number of MST = 2 × 2 = 4 (Answer)
(Because one 4 forms cycle, cant be included in any way. Now from two 4 and 5 we can select one
in 2 × 2 = 4 ways)

 11 votes -- Ahwan Mishra (10.5k points)

1.12 Numerical Computation(2) top

1.12.1 Numerical Computation: GATE2014-1-37 top [Link]

There are 5 bags labeled 1 to 5. All the coins in a given bag have the same weight. Some bags have
coins of weight 10 gm, others have coins of weight 11 gm. I pick 1, 2, 4, 8, 16 coins respectively
from bags 1 to 5 Their total weight comes out to 323 gm. Then the product of the labels of the bags
having 11 gm coins is ___.
gate2014-1 algorithms numerical-answers normal numerical-computation

Answer

1.12.2 Numerical Computation: TIFR2014-B-20 top [Link]

Consider the following game. There is a list of distinct numbers. At any round, a player
arbitrarily chooses two numbers a, b from the list and generates a new number c by subtracting the
smaller number from the larger one. The numbers a and b are put back in the list. If the number c is
non-zero and is not yet in the list, c is added to the list. The player is allowed to play as many
rounds as the player wants. The score of a player at the end is the size of the final list.

Suppose at the beginning of the game the list contains the following numbers: 48, 99, 120, 165 and
273. What is the score of the best player for this game?
A. 40
B. 16
C. 33
D. 91
E. 123
tifr2014 algorithms numerical-computation

Answer

Answers: Numerical Computation

1.12.1 Numerical Computation: GATE2014-1-37 top [Link]


Selected Answer

Suppose X is the number of coins of 11 gm and Y is the number of 10 gm coins.

According to question,

11X + 10Y = 323 → (1)

X + Y = 31 → (2)
© Copyright GATE Overflow. All rights reserved.
126 1 Algorithms (323)

X + Y = 31 → (2)
Solving (1), (2), we get X = 13 and Y = 18. So, here number of coins of 11 gm is 13 and
the only possible combination for 13 coins is

1 coin from bag1

4 coins from bag3
8 coins from bag4
So, product of label of bags will be = 1 × 3 × 4 = 12.

 40 votes -- sonam vyas (14.8k points)

1.12.2 Numerical Computation: TIFR2014-B-20 top [Link]


Selected Answer

Option D is correct

Here the list is (48, 99, 120, 165, 273).

Gcd(48, 99) = 3 ,means if we subtract (99 − 48 = 51) then that number is also % 3,
So the numbers like (3, 6, 9 − − − −99) are added. Total numbers = 99/3 = 33
//y Gcd(48, 120) = 24,so the numbers %24 are added like (24, 48, − − −120). Total numbers
= 120/24 = 5
//y Gcd(48, 165) = 3,so the numbers (3, 6, 9, − − 24 − −48 − −99 − −120 − − − 165) are
added. Totally, 165/3 = 55

At end, Gcd(48, 273) = 3,so the numbers

(3, 6, 9 − −24 − − − 48 − −99 − − − 120 − −165 − − − 273) are added(which covers all
the above numbers)

So total numbers added to this list = 273/3 = 91

 10 votes -- venky.victory35 (753 points)

1.13 P Np Npc Nph(12) top

1.13.1 P Np Npc Nph: CMI2010-A-10 top [Link]

Consider the following statements.

1. NP-complete problems are those that we know we can never solve efficiently.
2. If we find an efficient algorithm for one NP-complete problem, then we can solve all NP-complete
problems efficiently.
3. Checking whether a number is a prime is an NP-complete problem.
Then:
A. 1 and 2 are true but 3 is false.
B. 1 and 3 are false but 2 is true.
C. 2 and 3 are true but 1 is false.
D. All three statements are false.

cmi2010 algorithms p-np-npc-nph

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 127

Answer

1.13.2 P Np Npc Nph: CMI2015-A-06 top [Link]

Suppose we have constructed a polynomial time reduction from problem A to problem B. Which of
the following can we infer from this fact?

A. If the best algorithm for B takes exponential time, there is no polynomial time algorithm for A.
B. If the best algorithm for A takes exponential time, there is no polynomial time algorithm for B.
C. If we have a polynomial time algorithm for A, we must also have a polynomial time algorithm for
B.
D. If we don’t know whether there is a polynomial time algorithm for B, there cannot be a
polynomial time algorithm for A.

cmi2015 algorithms p-np-npc-nph

Answer

1.13.3 P Np Npc Nph: CMI2017-A-10 top [Link]

We have constructed a polynomial time reduction from problem A to problem B. Which of the
following is a valid inference?

A. If the best algorithm for B takes exponential time, then there is no polynomial time algorithm for
A
B. If the best algorithm for A takes exponential time, then there is no polynomial time algorithm for
B.
C. If we have a polynomial time algorithm for A, then we must also have a polynomial time
algorithm for B
D. If we don’t know whether there is a polynomial time algorithm for B, then there cannot be a
polynomial time algorithm for A.

cmi2017 algorithms theory-of-computation p-np-npc-nph

Answer

1.13.4 P Np Npc Nph: GATE1992-02,vi top [Link]

Choose the correct alternatives (more than one may be correct) and write the corresponding letters
only:

Which of the following problems is not NP -hard?

a. Hamiltonian circuit problem
b. The 0/1 Knapsack problem
c. Finding bi-connected components of a graph
d. The graph coloring problem

gate1992 p-np-npc-nph algorithms

Answer

1.13.5 P Np Npc Nph: GATE2003-12 top [Link]

Ram and Shyam have been asked to show that a certain problem Π is NP-complete. Ram shows a
polynomial time reduction from the 3-SAT problem to Π, and Shyam shows a polynomial time
reduction from Π to 3-SAT. Which of the following can be inferred from these reductions?

© Copyright GATE Overflow. All rights reserved.

128 1 Algorithms (323)

A. Π is NP-hard but not NP-complete

B. Π is in NP, but is not NP-complete
C. Π is NP-complete
D. Π is neither NP-hard, nor in NP
gate2003 algorithms p-np-npc-nph normal

Answer

1.13.6 P Np Npc Nph: GATE2004-30, ISRO2017-10 top [Link]

The problem 3-SAT and 2-SAT are

A. both in P
B. both NP complete

C. NP -complete and in P respectively

D. undecidable and NP complete respectively

gate2004 algorithms p-np-npc-nph easy isro2017

Answer

1.13.7 P Np Npc Nph: GATE2006-16, ISRO-DEC2017-27 top

[Link]

Let S be an NP-complete problem and Q and R be two other problems not known to be in NP. Q is
polynomial time reducible to S and S is polynomial-time reducible to R. Which one of the following
statements is true?

A. R is NP-complete
B. R is NP-hard
C. Q is NP-complete
D. Q is NP-hard

gate2006 algorithms p-np-npc-nph normal isrodec2017

Answer

1.13.8 P Np Npc Nph: GATE2008-44 top [Link]

The subset-sum problem is defined as follows: Given a set S of n positive integers and a positive
integer W , determine whether there is a subset of S whose elements sum to W . An algorithm Q
solves this problem in O(nW) time. Which of the following statements is false?

A. Q solves the subset-sum problem in polynomial time when the input is encoded in unary
B. Q solves the subset-sum problem in polynomial time when the input is encoded in binary
C. The subset sum problem belongs to the class NP
D. The subset sum problem is NP-hard

gate2008 algorithms p-np-npc-nph normal

Answer

1.13.9 P Np Npc Nph: TIFR2010-B-39 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 129

Suppose a language L is NP complete. Then which of the following is FALSE?

A. L ∈ NP
B. Every problem in P is polynomial time reducible to L.
C. Every problem in NP is polynomial time reducible to L.
D. The Hamilton cycle problem is polynomial time reducible to L.
E. P ≠ NP and L ∈ P.

tifr2010 algorithms p-np-npc-nph

Answer

1.13.10 P Np Npc Nph: TIFR2011-B-37 top [Link]

Given an integer n ≥ 3, consider the problem of determining if there exist integers a, b ≥ 2 such
that n = ab . Call this the forward problem. The reverse problem is: given a and b, compute ab (mod
b). Note that the input length for the forward problem is ⌊log n⌋ + 1, while the input length for the
reverse problem is ⌊log a⌋ + ⌊log b⌋ + 2. Which of the following statements is TRUE?

a. Both the forward and reverse problems can be solved in time polynomial in the lengths of their
respective inputs.
b. The forward problem can be solved in polynomial time, however the reverse problem is NP -
hard.
c. The reverse problem can be solved in polynomial time, however the forward problem is NP -
hard.
d. Both the forward and reverse problem are NP -hard.
e. None of the above.

tifr2011 algorithms p-np-npc-nph

Answer

1.13.11 P Np Npc Nph: TIFR2013-B-7 top [Link]

Which of the following is not implied by P = NP ?

a. 3SAT can be solved in polynomial time.
b. Halting problem can be solved in polynomial time.
c. Factoring can be solved in polynomial time.
d. Graph isomorphism can be solved in polynomial time.
e. Travelling salesman problem can be solved in polynomial time.

tifr2013 algorithms p-np-npc-nph

Answer

1.13.12 P Np Npc Nph: TIFR2017-B-2 top [Link]

Consider the following statements:

i. Checking if a given undirected graph has a cycle is in P

ii. Checking if a given undirected graph has a cycle is in NP
iii. Checking if a given directed graph has a cycle is in P
iv. Checking if a given directed graph has a cycle is in NP
Which of the above statements is/are TRUE? Choose from the following options.

A. Only i and ii
B. Only ii and iv

© Copyright GATE Overflow. All rights reserved.

130 1 Algorithms (323)

C. Only ii, iii, and iv

D. Only i, ii and iv
E. All of them

tifr2017 algorithms p-np-npc-nph

Answer

Answers: P Np Npc Nph

1.13.1 P Np Npc Nph: CMI2010-A-10 top [Link]

Option is B.
 2 votes -- shubham (221 points)

1.13.2 P Np Npc Nph: CMI2015-A-06 top [Link]


Selected Answer

Problem
A reduces to Problem
B,
Option (B) If the best algorithm for A takes exponential time, there is no polynomial time
algorithm for B. will be correct.

 8 votes -- Muktinath Vishwakarma (35.4k points)

1.13.3 P Np Npc Nph: CMI2017-A-10 top [Link]

A is reducible to B implies B is as tough as A. (A cannot be harder than B)

Option A - False. If A is polynomial then B must be Polynomial (A polynomial algorithm can be
easily converted into exponential. Converse is not true).

Option B- True. As per first line above, if A is expositional then B cannot be a polynomial time.
Option C - False. If we have polynomial time algorithm for A then we can have polynomial,
expositional, sub expositional algorithm for B.

Option D- False. A can be polynomial. B can be harder than polynomial ( as per first line, B is as
hard as A can be termed as B = Ω(A)).

 3 votes -- Keval Malde (16.1k points)

1.13.4 P Np Npc Nph: GATE1992-02,vi top [Link]


Selected Answer

a. Is NPC and hence NP hard.

b. Is again NP hard (optimization version is NP hard and decision version is NPC).
Ref: [Link]
is-np-complete

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 131

c. Is in P. See the algorithm here based on

DFS: [Link]
d. NPC and hence NP hard.

 11 votes -- Arjun Suresh (350k points)

1.13.5 P Np Npc Nph: GATE2003-12 top [Link]


Selected Answer

C. For a problem to be NP-Complete, it must be NP-hard and it must also be in NP.

Ram's reduction shows that Π is NP hard because it must be at least as hard as 3-SAT which is a
known NP-Complete problem. Here, Π need not be NP-Complete.

Now, Shyam's reduction shows that 3-SAT problem is at least as hard as Π or equivalently Π is
not harder than 3-SAT. Since 3-SAT is in NP, Π must also be in NP.

Thus, Π is NP-hard and also in NP ⟹ Π is NP-Complete.

 10 votes -- Anurag Semwal (8k points)

1.13.6 P Np Npc Nph: GATE2004-30, ISRO2017-10 top [Link]


Selected Answer

Option is C.
[Link]

 9 votes -- anshu (3.3k points)

1.13.7 P Np Npc Nph: GATE2006-16, ISRO-DEC2017-27 top

[Link]

Selected Answer

Answer B.

As S is NPC i.e NP -Hard and NP .

We know that, if NP -Hard problem is reducible to another problem in Polynomial Time, then
that problem is also NP -Hard which means every NP problem can be reduced to this
problem in Polynomial Time.

Therefore, R is NP -Hard.
Now Q is reduced to S in polynomial time.
If Q is reducible to S in polynomial time, Q could be NP because all NP problems can be
reduced to S. Since Q could be NP therefore Q could be P also as P is subset of NP . Also Q
could be NPC because every NPC problem can be reduced to another NPC problem in
polynomial time.

So, nothing can be concluded about Q.

© Copyright GATE Overflow. All rights reserved.

132 1 Algorithms (323)

 11 votes -- Mehak Sharma (1.6k points)

1.13.8 P Np Npc Nph: GATE2008-44 top [Link]


Selected Answer

Subset problem is NP-Complete - there is reduction proof but I don't remember (Can see the below
link). So, (C) and (D) are true as an NPC problem is in NP as well as NPH.

[Link]

Now, complexity of Q is O(nW), where W is an integer.

A. Input is encoded in unary. So, length of input is equal to the value of the input. So, complexity
= O(nW) where both n and W are linear multiples of the length of the inputs. So, the
complexity is polynomial in terms of the input length. So, (A) is true.
B. Input is encoded in binary. So, length of W will be lgW . (for W = 1024, input length will be
just 10). So, now W is exponential in terms of the input length of W and O(nW) also
becomes exponential in terms of the input lengths. So, Q is not a polynomial time algorithm.
So, (B) is false.

 17 votes -- Arjun Suresh (350k points)

1.13.9 P Np Npc Nph: TIFR2010-B-39 top [Link]


Selected Answer

Option E leads to a contradiction, hence is false.

We know that L is NPC, hence ∈ NP. If P ≠ NP, then L can't be in P

 6 votes -- Pragy Agarwal (20.6k points)

1.13.10 P Np Npc Nph: TIFR2011-B-37 top [Link]


Selected Answer

The reverse problem can be solved in polynomial time as ab requires at most log b recursive calls
using the approach given below:

pow(int a, int b)
{
if(b%2)
return a* pow(a*a, b/2);
else
return pow(a*a, b/2);
}

Now, the forward problem is also solvable in polynomial time. We need to check for all the roots of
n − 1
(from √n till n log n ) whether it is an integer . But each of these check can be done in log n
time using a binary search on the set of integers from 2..n and so, the overall complexity will be
(log n)2 which is polynomial in log n (log n is the size of input). So, (a) must be the answer.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 133

 6 votes -- gatecse (18k points)

1.13.11 P Np Npc Nph: TIFR2013-B-7 top [Link]


Selected Answer

I believe Umang is right, option B is the correct answer.

Intractability : We are looking for EFFICIENT algorithms.

Intractable Problems are problems that are decidable, although the algorithm to decide that
problem might be efficient (P) or inefficient (NP), but at least an algorithm exists to solve these
problems.

Here we talk about efficient vs inefficient computations.

Thus the language of problems in P and NP classes is the language of Decidable

Problems i.e. Recursive Language.

Undecidability: We are looking for algorithms.

Undecidable Problems are problems for which there is no algorithm to solve these problems.

Here we talk about what can or can not be computed.

The language of Undecidable Problems are "Recursively Enumerable but not recursive
languages" & "Not Recursively Enumerable Languages".

Clearly we can talk about the intractability of any problem if we know at least one
algorithm to solve the problem, if there is no algorithm to solve a problem how can we talk about
efficiency?

Halting Problem is undecidable.

I guess, all other problems mentioned here are decidable.

I don't know the most efficient algorithms to solve these problems but at least I can say that Brute
force approach will work on all the other options except the Halting Problem.

What P = NP implies?

"Any problem that is solved by a non deterministic Turing machine in polynomial time
also be solved by some deterministic Turing machine in polynomial time, even if the
degree of the polynomial is higher."

and

There is neither a Non Deterministic Turing Machine nor Deterministic Turing Machine
that can solve the Halting Problem.

So any inference about P & NP is not going to affect the solvability of Halting Problem,
since it is undecidable.

 10 votes -- Anurag Pandey (13.6k points)

© Copyright GATE Overflow. All rights reserved.

134 1 Algorithms (323)

1.13.12 P Np Npc Nph: TIFR2017-B-2 top [Link]


Selected Answer

E. All of them. Because all of them can be solved by Depth first traversal.

Every P problem is a subset of NP.

 9 votes -- Motamarri Anusha (12.3k points)

1.14 Recurrence(32) top

1.14.1 Recurrence: GATE1987-10a top [Link]

Solve the recurrence equations:

T (n) = T (n − 1) + n
T (1) = 1

gate1987 algorithms recurrence

Answer

1.14.2 Recurrence: GATE1988-13iv top [Link]

Solve the recurrence equations:

T (n) = T ( n2 ) + 1
T (1) = 1

gate1988 descriptive algorithms recurrence

Answer

1.14.3 Recurrence: GATE1989-13b top [Link]

Find a solution to the following recurrence equation:

T (n) = √−
n + T ( n2 )

T (1) = 1
gate1989 descriptive algorithms recurrence

Answer

1.14.4 Recurrence: GATE1990-17a top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 135

Express T (n) in terms of the harmonic number Hn = ∑nt=1 1/i, n ≥ 1 , where T (n) satisfies the
recurrence relation,

T (n) = n T (n − 1)
n+1
+ 1 , for n ≥ ∑ and T (1) = 1

What is the asymptotic behaviour of T (n) as a function of n ?

gate1990 descriptive algorithms recurrence

Answer

1.14.5 Recurrence: GATE1992-07a top [Link]

Consider the function F(n) for which the pseudocode is given below :
Function F(n)
begin
F1 ← 1
if(n=1) then F ← 3
else
For i = 1 to n do
begin
C ← 0
For j = 1 to n – 1 do
begin C ← C + 1 end
F1 = F1 * C
end
F = F1
end

[n is a positive integer greater than zero]

(a) Derive a recurrence relation for F(n)

gate1992 algorithms recurrence descriptive

Answer

1.14.6 Recurrence: GATE1992-07b top [Link]

Consider the function F(n) for which the pseudocode is given below :
Function F(n)
begin
F1 ← 1
if(n=1) then F ← 3
else
For i = 1 to n do
begin
C ← 0
For j = 1 to n – 1 do
begin C ← C + 1 end
F1 = F1 * C
end
F = F1
end

[n is a positive integer greater than zero]

Solve the recurrence relation for a closed form solution of F(n).

© Copyright GATE Overflow. All rights reserved.

136 1 Algorithms (323)

gate1992 algorithms recurrence descriptive

Answer

1.14.7 Recurrence: GATE1993-15 top [Link]

Consider the recursive algorithm given below:

procedure bubblesort (n);

var i,j: index; temp : item;
begin
for i:=1 to n-1 do
if A[i] > A[i+1] then
begin
temp := A[i];
A[i] := A[i+1];
A[i+1] := temp;
end;
bubblesort (n-1)
end

Let an be the number of times the ‘if…then…’ statement gets executed when the algorithm is run
with value n. Set up the recurrence relation by defining an in terms of an−1 . Solve for an .

gate1993 algorithms recurrence normal

Answer

1.14.8 Recurrence: GATE1994-1.7, ISRO2017-14 top [Link]

The recurrence relation that arises in relation with the complexity of binary search is:

A. T (n) = 2T ( n2 ) + k, k is a constant
B. T (n) = T ( n2 ) + k, k is a constant
C. T (n) = T ( n2 ) + log n
D. T (n) = T ( n2 ) + n

gate1994 algorithms recurrence easy isro2017

Answer

1.14.9 Recurrence: GATE1996-2.12 top [Link]

The recurrence relation

T (1) = 2
T (n) = 3T ( n4 ) + n
has the solution T (n) equal to
A. O(n)
B. O(log n)

O (n 4 )
3
C.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 137

D. None of the above

gate1996 algorithms recurrence normal

Answer

1.14.10 Recurrence: GATE1997-4.6 top [Link]

Let T (n) be the function defined by T (1) = 1, T (n) = 2T (⌊ n2 ⌋) + √−

n for n ≥ 2.
Which of the following statements is true?

A. T (n) = O√−
n
B. T (n) = O(n)
C. T (n) = O(log n)
D. None of the above

gate1997 algorithms recurrence normal

Answer

1.14.11 Recurrence: GATE1998-6a top [Link]

Solve the following recurrence relation

xn = 2xn−1 − 1, n > 1
x1 = 2
gate1998 algorithms recurrence descriptive

Answer

1.14.12 Recurrence: GATE2002-1.3 top [Link]

The solution to the recurrence equation T (2k ) = 3T (2k−1 ) + 1, T (1) = 1 is

A. 2k
(3k+1 −1)
B. 2
log2 k
C. 3
D. 2log3 k
gate2002 algorithms recurrence normal

Answer

1.14.13 Recurrence: GATE2002-2.11 top [Link]

The running time of the following algorithm

Procedure A(n)
If n ⩽ 2 return (1) else return (A(⌈√−
n ⌉));
is best described by

O(n)
© Copyright GATE Overflow. All rights reserved.
138 1 Algorithms (323)

A. O(n)
B. O(log n)
C. O(log log n)
D. O(1)

gate2002 algorithms recurrence normal

Answer

1.14.14 Recurrence: GATE2003-35 top [Link]

Consider the following recurrence relation

T (1) = 1
−−−−−
T (n + 1) = T (n) + ⌊√n + 1 ⌋ for all n≥1
The value of T (m2 ) for m ≥ 1 is

6 (21m − 39) + 4
A. m
2
B. m
6 (4m − 3m + 5)
m 2.5 − 11m + 20) − 5
2 (3m
C.
3 2 5
6 (5m − 34m + 137m − 104) + 6
m
D.

gate2003 algorithms time-complexity recurrence

Answer

1.14.15 Recurrence: GATE2004-83, ISRO2015-40 top [Link]

The time complexity of the following C function is (assume n > 0)

int recursive (int n) {
if(n == 1)
return (1);
else
return (recursive (n-1) + recursive (n-1));
}

A. O(n)
B. O(n log n)
C. O(n2 )
D. O(2n )

gate2004 algorithms recurrence time-complexity normal isro2015

Answer

1.14.16 Recurrence: GATE2004-84 top [Link]

The recurrence equation

T (1) = 1
T (n) = 2T (n − 1) + n, n ≥ 2
evaluates to

A. 2n+1 − n − 2
n
−n
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 139

B. 2n − n
C. 2n+1 − 2n − 2
D. 2n + n
gate2004 algorithms recurrence normal

Answer

1.14.17 Recurrence: GATE2004-IT-57 top [Link]

Consider a list of recursive algorithms and a list of recurrence relations as shown below. Each
recurrence relation corresponds to exactly one algorithm and is used to derive the time complexity
of the algorithm.

Recursive Recurrence
Algorithm Relation
T (n) = T (n
P. Binary search I. − k) + T (k)
+ cn
T (n) = 2T (n
Q. Merge sort II.
− 1) + 1
T (n) = 2T (n
R. Quick sort III.
/2) + cn
Tower of T (n) = T (n/2)
S. IV.
Hanoi +1
Which of the following is the correct match between the algorithms and their recurrence relations?

A. P-II, Q-III, R-IV, S-I

B. P-IV, Q-III, R-I, S-II
C. P-III, Q-II, R-IV, S-I
D. P-IV, Q-II, R-I, S-III
gate2004-it algorithms recurrence normal

Answer

1.14.18 Recurrence: GATE2005-IT-51 top [Link]

Let T (n) be a function defined by the recurrence

T (n) = 2T (n/2) + √− n for n ≥ 2 and
T (1) = 1
Which of the following statements is TRUE?

A. T (n) = Θ(log n)
B. T (n) = Θ(√−n)
C. T (n) = Θ(n)
D. T (n) = Θ(n log n)

gate2005-it algorithms recurrence easy

Answer

© Copyright GATE Overflow. All rights reserved.

140 1 Algorithms (323)

1.14.19 Recurrence: GATE2006-51, ISRO2016-34 top [Link]

Consider the following recurrence:

T (n) = 2T (√−
n ) + 1, T (1) = 1

Which one of the following is true?

A. T (n) = Θ(log log n)

B. T (n) = Θ(log n)
C. T (n) = Θ(√−n)
D. T (n) = Θ(n)

algorithms recurrence isro2016 gate2006

Answer

1.14.20 Recurrence: GATE2008-78 top [Link]

Let xn denote the number of binary strings of length n that contain no consecutive 0s.
Which of the following recurrences does xn satisfy?

A. xn = 2xn−1
B. xn = x⌊n/2⌋ + 1
C. xn = x⌊n/2⌋ + n
D. xn = xn−1 + xn−2

gate2008 algorithms recurrence normal

Answer

1.14.21 Recurrence: GATE2008-79 top [Link]

Let xn denote the number of binary strings of length n that contain no consecutive 0s.
The value of x5 is

A. 5
B. 7
C. 8
D. 16
gate2008 algorithms recurrence normal

Answer

1.14.22 Recurrence: GATE2008-IT-44 top [Link]

When n = 22k for some k ⩾ 0, the recurrence relation

T (n) = √(2)T (n/2) + √n, T (1) = 1
evaluates to :

A. √(n)(log n + 1)
B. √(n) log n
√(n) log √(n)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 141

C. √(n) log √(n)

D. n log √n

gate2008-it algorithms recurrence normal

Answer

1.14.23 Recurrence: GATE2009-35 top [Link]

The running time of an algorithm is represented by the following recurrence relation:

n≤3
T (n) = {
n
T ( n3 ) + cn otherwise
Which one of the following represents the time complexity of the algorithm?

A. Θ(n)
B. Θ(n log n)
C. Θ(n2 )
D. Θ(n2 log n)

gate2009 algorithms recurrence time-complexity normal

Answer

1.14.24 Recurrence: GATE2012-16 top [Link]

The recurrence relation capturing the optimal execution time of the T owers of Hanoi problem
with n discs is

A. T (n) = 2T (n − 2) + 2
B. T (n) = 2T (n − 1) + n
C. T (n) = 2T (n/2) + 1
D. T (n) = 2T (n − 1) + 1

gate2012 algorithms easy recurrence

Answer

1.14.25 Recurrence: GATE2014-2-13 top [Link]

Which one of the following correctly determines the solution of the recurrence relation with
T (1) = 1?
n
T (n) = 2T ( ) + log n
2
A. Θ(n)
B. Θ(n log n)
C. Θ(n2 )
D. Θ(log n)

gate2014-2 algorithms recurrence normal

Answer

© Copyright GATE Overflow. All rights reserved.

142 1 Algorithms (323)

1.14.26 Recurrence: GATE2015-1-49 top [Link]

Let an represent the number of bit strings of length n containing two consecutive 1s. What is the
recurrence relation for an ?

A. an−2 + an−1 + 2n−2

B. an−2 + 2an−1 + 2n−2
C. 2an−2 + an−1 + 2n−2
D. 2an−2 + 2an−1 + 2n−2

gate2015-1 algorithms recurrence normal

Answer

1.14.27 Recurrence: GATE2015-3-39 top [Link]

Consider the following recursive C function.

void get(int n)
{
if (n<1) return;
get (n-1);
get (n-3);
printf("%d", n);
}

If get(6) function is being called in main() then how many times will the get() function be invoked
before returning to the main()?

A. 15
B. 25
C. 35
D. 45
gate2015-3 algorithms recurrence normal

Answer

1.14.28 Recurrence: GATE2016-2-39 top [Link]

The given diagram shows the flowchart for a recursive function A(n). Assume that all statements,
except for the recursive calls, have O(1) time complexity. If the worst case time complexity of this
function is O(nα ), then the least possible value (accurate up to two decimal positions) of α is
________.

Flow chart for Recursive Function A(n).

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 143

gate2016-2 algorithms time-complexity recurrence normal numerical-answers

Answer

1.14.29 Recurrence: GATE2017-2-30 top [Link]

Consider the recurrence function

2T (√−
n ) + 1, n>2
T (n) = {
2, 0<n≤2
Then T (n) in terms of θ notation is
A. θ(log log n)
B. θ(log n)
C. θ(√−n)
D. θ(n)

gate2017-2 algorithms recurrence

Answer

1.14.30 Recurrence: TIFR2014-B-11 top [Link]

Consider the following recurrence relation:

T ( nk ) + T ( 3n
4 )+n if n ≥ 2
T (n) = {
1 if n = 1
Which of the following statements is FALSE?

a. T (n) is O(n3/2 ) when k = 3.

b. T (n) is O(n log n) when k = 3.
c. T (n) is O(n log n) when k = 4.
d. T (n) is O(n log n) when k = 5.
e. T (n) is O(n) when k = 5.

tifr2014 algorithms recurrence

Answer

1.14.31 Recurrence: TIFR2015-B-1 top [Link]

Consider the following recurrence relation:

2T (⌊√−
n ⌋) + log n if n ≥ 2
T (n) = {
1 if n = 1
Which of the following statements is TRUE?

a. T (n) is O(log n).

b. T (n) is O(log n. log log n) but not O(log n).
c. T (n) is O(log3/2 n) but not O(log n. log log n).
d. T (n) is O(log2 n) but not O(log3/2 n).
e. T (n) is O(log2 n. log log n) but not O(log2 n).

© Copyright GATE Overflow. All rights reserved.

144 1 Algorithms (323)

tifr2015 algorithms recurrence time-complexity

Answer

1.14.32 Recurrence: TIFR2017-A-15 top [Link]

Let T (a, b) be the function with two arguments (both nonnegative integral powers of 2) defined by
the following reccurence:

T (a, b) = T ( a2 , b)+T (a, 2b ) if a, b ≥ 2 ;

T (a, 1) = T ( a2 , 1) if a ≥ 2 ;

T (1, b) = T (1, 2b ) if b ≥ 2 ;
T (1, 1) = 1.
What is T (2r , 2s ) ?
A. rs
B. r+s
2r + 2s
C. ( )
2r
r+s
D. ( )
r
E. 2r−s if r ≥ s, otherwise 2s−r
tifr2017 algorithms recurrence

Answer

Answers: Recurrence

1.14.1 Recurrence: GATE1987-10a top [Link]


Selected Answer

T (n) = T (n − 1) + n
= T (n − 2) + (n − 1) + n
= T (n − 3) + (n − 2) + (n − 1) + n
.
.
.
= T (n − k) + [(n − k + 1) + (n − k + 2) + … + (n − 1) + n]
Recurrence stops at,

n−k=1
k=n−1
n(n+1)
∴ T (n) = T (1) + [2 + 3 + … + n] = 1 + 2 + 3 + … + n = 2

PS: Unless explicitly asked for asymptotic bound, we should always try to get the exact answer.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 145

 19 votes -- kirti singh (3.8k points)

1.14.2 Recurrence: GATE1988-13iv top [Link]


Selected Answer

T (n) = T (n/2) + 1
= T (n/4) + 2
= T (n/8) + 3

= T (n/2k ) + k.

Recurrence stops when 2k >= n.

When 2k = n, k = lg n
So, T (n) = T (1) + lg n = 1 + lg n
PS: Unless explicitly asked for asymptotic bound, we should give exact answers for solutions of
recurrence equations.

 4 votes -- Arjun Suresh (350k points)

1.14.3 Recurrence: GATE1989-13b top [Link]

O(n^1/2)

 2 votes -- Purvi Agrawal (2.9k points)

1.14.4 Recurrence: GATE1990-17a top [Link]


Selected Answer

T (n) = n T (n − 1)
n+1
+1 → (1)

T (n − 1) = n−1 T (n − 2)
n
+1 → (2)

T (n − 2) = n−2 T (n − 3)
n−1
+1 → (3)

Substituting value of T (n − 1) from (2) in (1)

⟹ T (n) = n+1
n ∗ n−1 T (n − 2)
n
+ n+1
n +1

⟹ T (n) = n−1 T (n − 2)
n+1
+ n+1
n +1

Now substituting value of T (n − 2) in above equation

⟹ T (n) = n+1
n−1 ∗ n−2 T (n − 3)
n−1
+ n+1
n−1 + n+1
n +1

⟹ T (n) = n+1
T (n − 3) + n+1
+ n+1
+1
© Copyright GATE Overflow. All rights reserved.
146 1 Algorithms (323)

⟹ T (n) = n−2 T (n − 3)
n+1
+ n+1
n−1 + n+1
n +1

⋮
so on

T (n) = n+1
n−k+1
T (n − k) + n+1
n + n+1
n−1 +…+ n+1
n−k+2
+1

T (n) = n+1
n−k+1
T (n − k) + (n + 1) ∗ ( n1 + 1
n−1 +…+ 1
n−k+2
) +1

Now let n−k=1 so k = n − 1 , substitute value of k in above equation

⟹ T (n) = n+1
n−(n−1)+1
T (1) + (n + 1) ∗ ( n1 + 1
n−1 +…+ 1
n−(n−1)+2
)+1

⟹ T (n) = n+1
2 + (n + 1) × ( n1 + 1
n−1 + … + 13 ) + 1
1
⟹ T (n) = n+1
2 + (n + 1) × (Hn − 2 − 1) + 1

⟹ T (n) = n+1
2 + (n + 1) × Hn − n+1
2 − (n + 1) + 1

⟹ T(n) = (n + 1) × Hn − n
Now, Hn ≈ log n + γ
where γ is the Euler-Mascheroni constant.
T(n) = O(n log n)
 15 votes -- Digvijaysingh Gautam (9k points)

1.14.5 Recurrence: GATE1992-07a top [Link]


Selected Answer

1. The function F(n) is NOT a recursive function. You can't have a recurrence relation for it in the
first place!
2. F(n) calculates (n − 1)n .

The equivalent C++ code is as follows: (You can try it out here: [Link]

long F(long n) {
long F1 = 1;

if(n==1) { return 3; }
else {
for(long i = 1; i <= n; i++) {
long C = 0;
// Note: the belore For loop only has one line
for(long j = 1; j <= n-1; j++) { C = C+1; }
// At the end of this for loop, C will be = (n-1)
F1 = F1 * C;
}
}
return F1;
}

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 147

It is clear that the inner for loop can be replaced by a single statement as follows:

long F(long n) {
long F1 = 1;

if(n==1) { return 3; }
else {
for(long i = 1; i <= n; i++)
F1 = F1 * (n-1);
}
return F1;
}

And this calculates (n − 1)n

 17 votes -- Pragy Agarwal (20.6k points)

1.14.6 Recurrence: GATE1992-07b top [Link]


Selected Answer

Function F(n)
begin
F1 ← 1
if(n=1) then F ← 3 //if (n==1) then return 3
else
For i = 1 to n do
begin
C ← 0
For j = 1 to n – 1 do //inner loop runs n-1 times outer loop runs for n times
begin C ← C + 1 end //means C=n-1
F1 = F1 * C //means n-1 is getting multiplied n times so ans is (n-1)^n for n>=2
end
F = F1
end

 10 votes -- Rajesh Pradhan (22.6k points)

1.14.7 Recurrence: GATE1993-15 top [Link]


Selected Answer

an = an−1 + n − 1 (n − 1 comparisons for n numbers)

an = an−2 + (n − 2) + (n − 1)
an = an−3 + (n − 3) + (n − 2) + (n − 1)
.
.
.

an = an−n + (n − n) + (n − (n − 1))+. . . . +(n − 3) + (n − 2) + (n − 1)

an = 0 + 1 + 2+. . . . +(n − 3) + (n − 2) + (n − 1)
(n−1)×(n)
which given an = 2

© Copyright GATE Overflow. All rights reserved.

148 1 Algorithms (323)

 19 votes -- Rajarshi Sarkar (34.1k points)

1.14.8 Recurrence: GATE1994-1.7, ISRO2017-14 top [Link]


Selected Answer

It is B. searching for only one half of the list. leading to T (n/2)+ constant time in comparing and
finding mid element.

 26 votes -- Gate Keeda (19.6k points)

1.14.9 Recurrence: GATE1996-2.12 top [Link]


Selected Answer

Answer: A

According to Master theorem,

T (n) = aT ( nb ) + f(n) can be expressed as:
T (n) = [nlogb a ][T (1) + u(n)]
f(n)
where u(n) = Θ(h(n)) where h(n) = log a = logn 3 = n1−log4 3 as h(n) = nr where r > 0.
n b n 4
log
So, T (n) = [n b a ][T (1) + u(n)] = T (n) = [nlog4 3 ][T (1) + Θ(n1−log4 3 )] = Θ(n1 ).

 14 votes -- Rajarshi Sarkar (34.1k points)

1.14.10 Recurrence: GATE1997-4.6 top [Link]


Selected Answer

Answer is B.
using master method (case 1)

where a = 2, b = 2

O(n1/2) < O(nlogba)

O(n1/2) < O(nlog22)

 15 votes -- Ankit Rokde (9k points)

1.14.11 Recurrence: GATE1998-6a top [Link]


Selected Answer

T (n) = 2T (n − 1) − 1
= 2(2T (n − 2) − 1) − 1

= 22 T (n − 2) − 2 − 1

2
= (2T (n − 3) − 1) − 2 − 1
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 149

= 22 (2T (n − 3) − 1) − 2 − 1

= 23 T (n − 3) − 22 − 2 − 1
…
= 2n−1 T (n − (n − 1)) − (2n−2 + 2n−3 + ⋯ + 22 + 2 + 1)
a.(rn −1)
= 2n−1 × 2 − 2 2−1−1 ∵ T (1) = 2, Sn = r−1
n−1

= 2n − (2n−1 − 1)

= 2n−1 + 1
 20 votes -- srestha (87k points)

1.14.12 Recurrence: GATE2002-1.3 top [Link]


Selected Answer

Let x = 2k
T (x) = 3T ( x2 ) + 1
We can apply Master's theorem case 1 with a = 3 and b = 2 as f(x) = 1 = O (xlog2 3−ϵ ) , ϵ > 0
log2 3
T (x) = Θ (xlog2 3 ) = Θ (2k ) = Θ (2log2 3 ) = Θ (3k )
k
So,

So, only option possible is B.

We can also directly solve as follows:

T (x) = 3T ( x2 ) + 1
= 9T ( x4 ) + 1 + 3
⋮
= 3log2 2 + (1 + 3 + 9 + ⋯ + 3log2 2 −1 )
k k

(recursion depth is log2 x and x = 2k )

log 2k
= 3k + 3 3−1
2 −1

(Sum to n terms of GP with a = 1 and r = 3)

= 3k + 3 2−1
k

= 3.32−1
k

= 3 2−1
k+1

OR

T (2k ) = 3T (2k−1 ) + 1
= 32 T (2k−2 ) + 1 + 3

⋮
© Copyright GATE Overflow. All rights reserved.
150 1 Algorithms (323)

⋮
= 3k T (2k−k ) + (1 + 3 + 9 + ⋯ + 3k−1 )
(recursion depth is k)
= 3k + 33−1
k−1

(Sum to n terms of GP with a = 1 and r = 3)

= 3k + 3 2−1
k

= 3.32−1
k

3k+1 −1
= 2

 34 votes -- Arjun Suresh (350k points)

1.14.13 Recurrence: GATE2002-2.11 top [Link]


Selected Answer

The complexity will be the number of times the recursion happens which is equal to the number of
times we can take square root of n recursively, till n becomes 2.

T (n) = T (⌈√−
n ⌉) + 1

T (2) = 1
T (22 ) = T (2) + 1 = 2
T (22 ) = T (4) + 1 = 3
2

T (22 ) = T (16) + 1 = 4
3

So, T (n) = lg lg n + 1 = O (log log n)

Answer : Option C

 40 votes -- Arjun Suresh (350k points)

1.14.14 Recurrence: GATE2003-35 top [Link]


Selected Answer

−−−−
T (m2 ) = T (m2 − 1) + ⌊√(m2 )⌋
−−−−−−− −−−−
= T (m2 − 2) + ⌊√(m2 − 1) ⌋ + ⌊√(m2 )⌋
−−−−−−− −−−−−−− −−−−
= T (m2 − 3) + ⌊√(m2 − 2) ⌋ + ⌊√(m2 − 1) ⌋ + ⌊√(m2 )⌋

⋮
−−
− −−
− −−−−
= T (1) + ⌊√(2)⌋ + ⌊√(3)⌋ + … + ⌊√(m2 )⌋

= 3 × 1 + 5 × 2 + … + (2m − 1) × (m − 1) + m

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 151

(We are taking floor of square root of numbers, and between successive square roots number of
numbers are in the series 3, 5, 7 … like 3 numbers from 1..4, 5 numbers from 5 − 9 and so on).

We can try out options here or solve as shown at end:

Put m = 5, T (25) = 3 × 1 + 5 × 2 + 7 × 3 + 9 × 4 + 5 = 75
A. 59
B. 75
C. non-integer
D. 297.5

So, answer must be B.

T (m2 ) = 3 × 1 + 5 × 2 + ⋯ + (2m − 1) × (m − 1) + m
= m + ∑i = 1m−1 (2i + 1). (i)
= m + ∑i = 1m−1 2i2 + i
(m−1).m.(2m−1) (m−1)m
=m+ 3 + 2
= m6 (6 + 4m2 − 2m − 4m + 2 + 3m − 3)
= m6 (4m2 − 3m + 5)
n.(n+1)
[Sum of the first n natural numbers = 2 .
n.(n+1).(2n+1)
Sum of the squares of first n natural numbers = 6 .]

 19 votes -- Arjun Suresh (350k points)

1.14.15 Recurrence: GATE2004-83, ISRO2015-40 top [Link]


Selected Answer

Option is D.

int recursive (int n) {

if(n == 1)// takes constant time say 'A' time
return (1);// takes constant time say 'A' time
else
return (recursive (n-1) + recursive (n-1));//takes
//T(n-1) + T(n-1) time
}

T (n) = 2T (n − 1) + a is the recurrence equation found from the pseudo code .

Solving the Recurrence Equation By Back Substitution Method

T (n) = 2T (n − 1) + a → (1)
T (n − 1) = 2T (n − 2) + a
T (n − 2) = 2T (n − 3) + a
We can re-write Equation (1) as

© Copyright GATE Overflow. All rights reserved.

→ (2)
152 1 Algorithms (323)

T (n) = 2[2T (n − 2) + a] + a = 4T (n − 2) + 3a
= 2[2T (n − 3) + a] + 3a = 8T (n − 3) + 7a
→ (2)
⋮
= 2 T (n − k) + (2k − 1)a
k

On Substituting Limiting Condition

T (1) = 1 implies n−k=1 ⟹ k=n−1

Therefore, equation (2) becomes:
2n−1 + (2n−1 − 1)a = O(2n )

 22 votes -- pC (22k points)

1.14.16 Recurrence: GATE2004-84 top [Link]


Selected Answer

T (n) = 2T (n − 1) + n, n ⩾ 2, T (1) = 1

T (n) = n + 2(n − 1) + 22 (n − 2) + ⋯ + 2(n−1) (n − (n − 1))

= n(1 + 2 + ⋯ + 2n−1 ) − (1.2 + 2.22 + 3.23 + ⋯ + (n − 1).2n−1 )

= n(2n − 1) − (n.2n − 2n+1 + 2)

= 2n+1 − n − 2
 37 votes -- suraj (5.6k points)

T(1) = 1

T(2) = 4

T(3) = 11

T(4) = 26

T(5) = 57

T(6) = 120

T(7) = 247

So,

T (n) = 2n+1 − n − 2

 44 votes -- Arjun Suresh (350k points)

1.14.17 Recurrence: GATE2004-IT-57 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 153
Selected Answer

Answer is B.

 8 votes -- Sankaranarayanan P.N (11.5k points)

1.14.18 Recurrence: GATE2005-IT-51 top [Link]


Selected Answer

Option C is the answer. It can be done by Master's theorem.

nlogb a = nlog2 2 = n .

f(n) = √−
1
n = n2 .

1
So, f(n) = O (nlogb a−ϵ ) is true for any real ϵ, 0 < ϵ < 2 . Hence Master theorem Case 1
satisfied,

T (n) = Θ (nlogb a ) = Θ(n).

 22 votes -- Bhagirathi Nayak (14.1k points)

1.14.19 Recurrence: GATE2006-51, ISRO2016-34 top [Link]


Selected Answer

T (n) = 2T (n 2 ) + 1
1

1
= 2 (2T (n 22 ) + 1) + 1
1
= 4 × T (n 22 ) + 5
1
= 8 × T (n 23 ) + 13 ⋯

= 2(lg lg n) + 2 × lg lg n + 1 (Proved below)

= Θ(lg n)

1
n 2k = 2(Putting 2 so that we can take [Link] more step of recurrence can't change the complex

So, answer is B, T (n) = Θ(log n)

 33 votes -- Arjun Suresh (350k points)

1.14.20 Recurrence: GATE2008-78 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

154 1 Algorithms (323)

0 1−2
01 10 11 − 3
010 011 101 110 111 − 5
0101 0110 0111 1010 1011 1101 1110 1111 − 8
So, xn = xn−1 + xn−2 (For all the strings ending in 1, we get two new strings and for all strings
ending in 0, we get a new string. So, the new set of strings for n + 1, will have exactly n strings
ending in 1)

x5 = 8 + 5 = 13
 19 votes -- Arjun Suresh (350k points)

1.14.21 Recurrence: GATE2008-79 top [Link]


Selected Answer

Number of binary strings of length n that contain no consecutive 0s, following will be the required
recurrence relation:

T (n) = T (n − 1) + T (n − 2) n > 2
base conditionT (1) = 2 and T (2) = 3
T (1) = 2 There will be 2 strings of length 1, i.e 0 & 1
T (2) = 3 There will be 3 strings of length 2, i.e. 01, 10, 11
T (3) = T (1) + T (2) = 2 + 3 = 5
T (4) = T (3) + T (2) = 5 + 3 = 8
T (5) = T (4) + T (3) = 8 + 5 = 13
Hence, answer is 13, but no option matches!

 13 votes -- Vijay Thakur (17.1k points)

1.14.22 Recurrence: GATE2008-IT-44 top [Link]


Selected Answer

T (n) = √(2)T ( n2 ) + √−
n
–2 – −−
= √2 T ( 2n2 ) + √2√ n2 + √−
n
⋮
–lg n
= √2 T (1) + lg n√−
n
= √−
n + lg n√−n
= √−
n (lg n + 1)
If we use Master theorem we get option B. But one must know that Master theorem is used to find
the asymptotic bound and not an EXACT value. And in the question here it explicitly says
"evaluates to".

 40 votes -- Arjun Suresh (350k points)

1.14.23 Recurrence: GATE2009-35 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 155


Selected Answer

a = 1, b = 3, logb a = 0
So nlogb a = n0 = 1
f(n) = n
So, f(n) = Ω(1)
To, check Master theorem case 3, we need c > 0,
f(n/3) ≤ cf(n)
c=1
So using case three of master theorem

T (n) = Θ(f(n)) = Θ(n)

answer is A.

 19 votes -- Pooja Palod (31.3k points)

1.14.24 Recurrence: GATE2012-16 top [Link]


Selected Answer

Recurrence relation for Towers of Hanoi is

T (1) = 1
T (n) = 2T (n − 1) + 1
So Answer should be (D)

 25 votes -- Narayan Kunal (421 points)

1.14.25 Recurrence: GATE2014-2-13 top [Link]


Selected Answer

f(n) = log n
a = 2, b = 2 ⟹ nlogb a = n
So, f(n) = log n = O (n1−ϵ ), we can take any ϵ from 0-1 for example 0.5 which gives
log n = O(√(n)), whose proof is given
here: [Link]

So, Master theorem Case 1, and answer will be O (nlog2 2 ) = O(n)

Alternate way:

T (1) = 1T (2) = 2T (1) + log 2 = 3 = 3n − 2T (4) = 2T (2) + log 4 = 8 = 3n − 4T (8) = 2T (4

© Copyright GATE Overflow. All rights reserved.
156 1 Algorithms (323)

T (1) = 1T (2) = 2T (1) + log 2 = 3 = 3n − 2T (4) = 2T (2) + log 4 = 8 = 3n − 4T (8) = 2T (4

The second term being subtracted is growing at a lower rate than the first term. So, we can say
T (n) = O(n).

 28 votes -- Arjun Suresh (350k points)

1.14.26 Recurrence: GATE2015-1-49 top [Link]


Selected Answer

Counting the number of bit strings NOT containing two consecutive 1's. (It is easy to derive a
recurrence relation for the NOT case as shown below)

0 1
00 01 10 − 3 (append both 0 and 1 to any string ending in 0, and append 0 to any string
ending in 1)
000 001 010 100 101 − 5 (all strings ending in 0 give two strings and those ending in 1
give 1 string)
0000 0001 0010 0100 0101 1000 1001 1010 − 8
⋮
a′n = a′n−1 + a′n−2 (where an denote the number of bit strings of length n containing two
consecutive 1s)

2n − an = (2n−1 − an−1 ) + (2n−2 − an−2 )

an = 2n−2 (4 − 2 − 1) + an−1 + an−2
an = an−1 + an−2 + 2n−2
A is choice.

 39 votes -- Arjun Suresh (350k points)

1.14.27 Recurrence: GATE2015-3-39 top [Link]


Selected Answer

Answer : Option B

T(n) = T(n-1) + T(n-3) + 2, here T(n) denotes the number of times a recursive call is made
for input n. 2 denotes the two direct recursive calls.

T(n ≤ 0) = 0
T(1) = 2
T(2) = 4
T(3) = 6
T(4) = 10
T(5) = 16
T(6) = 24
So, answer is 24 + 1 call from main = 25.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 157

 41 votes -- Arjun Suresh (350k points)

1.14.28 Recurrence: GATE2016-2-39 top [Link]


Selected Answer

If they are asking for worst case complexity hence,

By calling A(n) we get A(n/2) 5 times,

A(n) = 5A(n/2) + O(1)

Hence, by applying masters theorem,
Case 1 : a > bk

nlog2 5
Thus value of alpha will be 2.32
 51 votes -- Shashank Chavan (3.3k points)

1.14.29 Recurrence: GATE2017-2-30 top [Link]


Selected Answer

T (n) = 2T (√−
n) + 1

Put, n = 2m

T (2m ) = 2T (2m/2 ) + 1
put, T (2m ) = s(m)
s(m) = 2s(m/2) + 1
Using case 1 of master method ,

= Θ(m) = Θ(log n)
[Link]

 26 votes -- jatin saini (4k points)

1.14.30 Recurrence: TIFR2014-B-11 top [Link]

We have to analyze it at three points k = 3, 4, 5.

At k = 4 it is standard quick sort kind of recurrence relation. So its complexity is O(n × log4 n)
Now if k > 4 then complexity will come less then O(n × log4 n) and vice versa for k < 4.
So option c, d are correct but b is false.

Although partial answer is achieved but do not know how part (a) and (e) could be investigated.

© Copyright GATE Overflow. All rights reserved.

158 1 Algorithms (323)

Some people are taking about Akra-Bazzi method. Some ex. of this method is mentioned in
following PDF.

[Link]

[Link]

 3 votes -- Chhotu Ram Chauhan (10.5k points)

1.14.31 Recurrence: TIFR2015-B-1 top [Link]


Selected Answer

Let n = 2k

T (2k ) = 2T (2k/2 ) + k

Let T (2k ) = S(k) ⟹ T (2k/2 ) = S(k/2)

S(k) = 2S(k/2) + k
This gives S(k) = Θ(k log k) (Master theorem case 2)
= Θ(log n log log n)
So ans is b

 12 votes -- Pooja Palod (31.3k points)

1.14.32 Recurrence: TIFR2017-A-15 top [Link]

Value of T (2r , 2s ) is nothing but the number of leaf nodes in its recursion tree. The reason for this
is that there is only one base condition T (1, 1) which returns 1 which are all ultimately added.
Here is an example recursion tree for T (22 , 23 ).

We will try to use some combinatorial technique to count the number of leaf nodes. One way of
counting the number of leaf nodes in the recurrence tree is by counting all possible paths that one
can take from the root to the leaf node. The reason for doing this would be clear in a moment. This
idea would be used to establish a link with a very common combinatorial problem.

Claim: If we consider a grid of size r × s then the possible paths from (r, s) to (0, 0) in that grid
such that we can either move backwards or downward is nothing but all possible paths in
recurrence tree from the root node to leaf node.

This happens to be true because arguments of recurrence relation are in powers of two, and at

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 159

each level of recurrence tree, the power of one of the arguments get reduced by one.

T (2r , 2s ) = T (2r−1 , 2s ) + T (2r , 2s−1 )

T (1, 2s ) = T (1, 2s−1 )
T (2r , 1) = T (2r−1 , 1)
T (1, 1) =1
Continuing from our previous example here is a grid of size 2 × 3 with two valid paths highlighted.

The number of valid paths possible in this grid will give us the count of the number of leaf nodes in
the recurrence tree. Interestingly this problem, when interpreted exactly in opposite way, happens
to be a very well known problem of combinatory.

In a grid of size r × s, count the number of possible paths from (0, 0) to (r, s) such that we can
either move forward(F) or upward(U).

Video:

So there are (r+s

r ) many different valid paths on this grid to reach (r, s) from (0, 0) or vice versa.
And in turn, there are same numbers of leaf nodes in the recursion tree. Therefore, we can
conclude with -

r+s
T (2r , 2s ) = ( )
r
Video:

 4 votes -- Prateek Dwivedi (4.4k points)

1.15 Searching(8) top

1.15.1 Searching: GATE1996-18 top [Link]

Consider the following program that attempts to locate an element x in an array a[] using binary
search. Assume N > 1 . The program is erroneous. Under what conditions does the program fail?

var i,j,k: integer; x: integer;

a: array; [1..N] of integer;
begin i:= 1; j:= n;
repeat
k:(i+j) div 2;
if a[k] < x then i:= k
else j:= k
until (a[k] = x) or (i >= j);

if (a[k] = x) then
writeln ('x is in the array')
else
writeln ('x is not in the array')
end;

© Copyright GATE Overflow. All rights reserved.

160 1 Algorithms (323)

gate1996 algorithms searching normal

Answer

1.15.2 Searching: GATE1996-2.13, ISRO2016-28 top [Link]

The average number of key comparisons required for a successful search for sequential search on n
items is

A. n
2
n−1
B. 2
n+1
C. 2
D. None of the above

gate1996 algorithms easy isro2016 searching

Answer

1.15.3 Searching: GATE2002-2.10 top [Link]

Consider the following algorithm for searching for a given number x in an unsorted array A[1..n]
having n distinct values:

1. Choose an i at random from 1..n

2. If A[i] = x, then Stop else Goto 1;

Assuming that x is present in A, what is the expected number of comparisons made by the
algorithm before it terminates?

A. n
B. n − 1
C. 2n
D. n
2

gate2002 searching normal

Answer

1.15.4 Searching: GATE2008-84 top [Link]

Consider the following C program that attempts to locate an element x in an array Y [ ] using binary
search. The program is erroneous.

f (int Y[10] , int x) {

int u, j, k;
i= 0; j = 9;
do {
k = (i+ j) / 2;
if( Y[k] < x) i = k;else j = k;
} while (Y[k] != x) && (i < j)) ;
if(Y[k] == x) printf(" x is in the array ") ;
else printf(" x is not in the array ") ;
}

On which of the following contents of Y and x does the program fail?

A. Y is [1 2 3 4 5 6 7 8 9 10]and x < 10
B. Y is [1 3 5 7 9 11 13 15 17 19]and x < 1
[2 2 2 2 2 2 2 2 2 2]
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 161

C. Y is [2 2 2 2 2 2 2 2 2 2]and x > 2
D. Y is [2 4 6 8 10 12 14 16 18 20]and 2 < x < 20 and x is even

gate2008 algorithms searching normal

Answer

1.15.5 Searching: GATE2008-85 top [Link]

Consider the following C program that attempts to locate an element x in an array Y [ ] using binary
search. The program is erroneous.

f (int Y[10] , int x) {

int u, j, k;
i= 0; j = 9;
do {
k = (i+ j) / 2;
if( Y[k] < x) i = k;else j = k;
} while (Y[k] != x) && (i < j)) ;
if(Y[k] == x) printf(" x is in the array ") ;
else printf(" x is not in the array ") ;
}

The correction needed in the program to make it work properly is

A. Change line 6 to: if (Y [k] < x)i = k + 1 ; else j = k − 1 ;

B. Change line 6 to: if (Y [k] < x)i = k − 1 ; else j = k + 1 ;
C. Change line 6 to: if (Y [k] < x)i = k; else j = k;
D. Change line 7 to: } while ((Y [k] == x)&&(i < j)) ;

gate2008 algorithms searching normal

Answer

1.15.6 Searching: GATE2017-1-48 top [Link]

Let A be an array of 31 numbers consisting of a sequence of 0's followed by a sequence of 1's. The
problem is to find the smallest index i such that A [i] is 1 by probing the minimum number of
locations in A. The worst case number of probes performed by an optimal algorithm is
____________.

gate2017-1 algorithms normal numerical-answers searching

Answer

1.15.7 Searching: TIFR2010-B-29 top [Link]

Suppose you are given an array A with 2n numbers.

The numbers in odd positions are sorted in ascending order, that is,
A[1] ≤ A[3] ≤ … ≤ A[2n − 1].
The numbers in even positions are sorted in descending order, that is, A[2] ≥ A[4] ≥ … ≥ A[2n].
What is the method you would recommend for determining if a given number is in the array?

A. Sort the array using quick-sort and then use binary search.
B. Merge the sorted lists and perform binary search.
C. Perform a single binary search on the entire array.

© Copyright GATE Overflow. All rights reserved.

162 1 Algorithms (323)

D. Perform separate binary searches on the odd positions and the even positions.
E. Search sequentially from the end of the array.

tifr2010 searching

Answer

1.15.8 Searching: TIFR2012-B-11 top [Link]

Consider the following three version of the binary search program. Assume that the elements of type
T can be compared with each other; also assume that the array is sorted.
i, j, k : integer;
a : array [1....N] of T;
x : T;

Program 1 : i := 1; j := N;
repeat
k := (i + j) div 2;
if a[k] < x then i := k else j := k
until (a[k] = x) or (i > j)
Program 2 : i := 1; j := N;
repeat
k := (i + j) div 2;
if x < a[k] then j := k - 1;
if a[k] < x then i := k + 1;
until i > j
Program 3 := i := 1; j := N
repeat
k := (i + j) div 2;
if x < a[k] then j := k else i := k + 1
until i > j

A binary search program is called correct provided it terminates with a[k] = x whenever such an
element exists, or it terminates with a [k] ≠ x if there exists no array element with value x. Which
of the following statements is correct?

A. Only Program 1 is correct

B. Only Program 2 is correct
C. Only Program 1 and 2 are correct.
D. Both Program 2 and 3 are correct
E. All the three programs are wrong

tifr2012 algorithms searching

Answer

Answers: Searching

1.15.1 Searching: GATE1996-18 top [Link]


Selected Answer

The code is wrong here

k=(i+j) / 2;
if (a[k] < x) then i = k;
else j = k;

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 163

The (correct) code should be:

k=(i+j) / 2;
if (a[k] < x) then i = k + 1;
else j = k - 1;

We can try an example with the given code in question

Let the array be a[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

Index numbers 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
Let x = 10; now run the code;

Initially i = 1, j = 10 ;
first time k = (i + j)/2 = 11/2 = 5.5 = 5 (because of integer type) =i
second time = k = (i + j)/2 = 15/2 = 7.5 = 7 = i
third time = k = (i + j)/2 = 17/2 = 8.5 = 8 = i
fourth time = k = (i + j)/2 = 18/2 = 9 = i
fifth time = k = (i + j)/2 = 19/2 = 9.5 = 9 = i
sixth time = k = (i + j)/2 = 19/2 = 9.5 = 9 = i
seventh time = k = (i + j)/2 = 19/2 = 9.5 = 9 = i

Going to infinite loop (run time error)

For terminating the loop, it should be i =k+1 instead of i = k and j = k − 1 instead of j = k;

 18 votes -- Mithlesh Upadhyay (5.8k points)

1.15.2 Searching: GATE1996-2.13, ISRO2016-28 top [Link]


Selected Answer

Expected number of comparisons

= 1× Probability of first element being x + 2× Probability of second element being
x + … + n× Probability of last element being x.
1 2 3
= n + n + n +…+ n
n

( )
n×(n+1)
2
= n

= n+1
2
 56 votes -- Arjun Suresh (350k points)

1.15.3 Searching: GATE2002-2.10 top [Link]


Selected Answer

(E) = 1×
© Copyright GATE Overflow. All rights reserved.
164 1 Algorithms (323)

Expected number of comparisons (E) = 1× Probability of find on first comparison +2×Probability

of find on second comparison +. . . +i× Probability of find on ith comparison +. . .

1 (n−1)2
=1× n +2× n−1
n2
+3× n3
+…
1/n (n−1)/n2
= n−1 + 2 (Sum to infinity of aritmetico-geometric series with a = n1 , r = n−1
n and
1− n (1− n−1
)
n

Reference: [Link]

Or we can also do,

1 (n−1)2
E=1× n +2× n−1
n2
+3× n3
+…
(n−1)2 (n−1)3
n =
E n−1 n−1
n2
+2× n3
+3× n4
+…

1 (n−1)2
E − E n−1
n = n + n−1
n2
+ n3
+…
1 (1/n) 1
E. n = = 1 (Sum to infinity of GP with a = n and r = n−1
n ) ⟹ E=n
1− n−1
n

 41 votes -- Arjun Suresh (350k points)

1.15.4 Searching: GATE2008-84 top [Link]


Selected Answer

for Q.84

when it is option C the control will continue to iterate as i = 8 and j = 9 ;

again and again i will be assigned k which itself equals 8 as 8+9
2 being stored in an integer type
variable, will evaluate to 8.

For option A, with x = 9, k will take the following values:

4
6
7
8 − y[8] = 9, x found
For option D, with x = 10, k will take the following values:
4, y[4] = 10, x found

 26 votes -- Amar Vashishth (30.5k points)

1.15.5 Searching: GATE2008-85 top [Link]


Selected Answer

Answer should be A.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 165

if( Y[k] < x) then i = k + 1;

if given element that we are searching is greater then searching will be continued upper half of
array

otherwise
j = k − 1;
lower half.

Take few case in consideration i.e.

1. All elements are same

2. Increasing order with no repeatation
3. Increasing order with repeatation.

 20 votes -- Manoj Kumar (38.6k points)

1.15.6 Searching: GATE2017-1-48 top [Link]


Selected Answer

Here, since 0s are followed by 1s so we have a sorted sequence and we can apply binary search.
(low+high) th
At each stage we compare with 2 element index and if it is 1 we check left and if it is 0
we check right.

Total worst case no. of probes is ⌈log2 31⌉ = 5.

So, answer is 5.
 36 votes -- sriv_shubham (3.3k points)

it should not take more than 5 probes

#include <bits/stdc++.h>
using namespace std;
#define N 100
void display(bool a[]) {
int i=0;
while(i<31) printf("%2d ",a[i++]);
}
void assign(bool a[],int zeros) {
for(int i=0;i<31;i++)
(i<zeros)?a[i] = 0:a[i] = 1;
}
int main() {
srand(time(NULL));
bool a[31];
// header
for(int i=0;i<31;i++) printf("%2d ",i);

© Copyright GATE Overflow. All rights reserved.

166 1 Algorithms (323)

printf("\n\n");
int max_probes = 0;
for(int iteration = 1;iteration <= N;iteration++) {
int zeros = rand()%32;
assign(a,zeros);
sort(a,a+31);
int low,high,mid,ans;
std::vector<int> seq;

low = 0;
high = 31;
while(low < high) {
mid = low + floor((high - low) / 2);
seq.push_back(mid);
if(a[mid] == 0) {
low = mid + 1;
ans = low;
if(a[mid + 1] == 1) break;
} else {
high = mid;
ans = high;
if(mid > 0)
if(a[mid - 1] == 0) break;

}
}

display(a);
printf(" | probes=%d ",[Link]());
for(auto e:seq) printf("%d ",e);
printf(" | at = %dth\n",ans);
//if(ans == 15) printf("\nHHH=-------------\n");
max_probes = max(max_probes,(int)([Link]()));
[Link]();
}
printf("%d\n",max_probes);
}

 20 votes -- Debashish Deka (56.7k points)

1.15.7 Searching: TIFR2010-B-29 top [Link]


Selected Answer

Option D is the correct answer.

We can simply use clever indexing to binary search the element in the odd positions, and in the
even positions separately.

This will take O(log n) time and O(1) space in the worst case.
A: Sorting using Quicksort will take O(n2 ) time.
B: Merging will take O(n) time and O(n) space.
C: Binary search only works on a sorted array.
E: Sequential search will take O(n) time.

 20 votes -- Pragy Agarwal (20.6k points)

1.15.8 Searching: TIFR2012-B-11 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 167


Selected Answer

First program wont work if array has elements same..it may go into infinite loop .To make it work
it properly we have to do following changes j = k − 1 and i = k + 1

For second program a[k] == x condition is missing so it is wrong

Third program is also wrong as j! = k − 1 and condition a[k] == x is missing
So, answer is E.

 12 votes -- Pooja Palod (31.3k points)

1.16 Sorting(55) top

1.16.1 Sorting: CMI2011-B-06a top [Link]

Consider a plate stacked with several disks, each of a different diameter (they could all be, for
instance, dosas or chapatis of different sizes). We want to sort these disks in decreasing order
according to their diameter so that the widest disk is at the bottom of the pile. The only operation
available for manipulating the disks is to pick up a stack of them from the top of the pile and invert
that stack. (This corresponds to lifting up a stack dosas or chapatis between two big spoons and
flipping the stack.)

i. Give an algorithm for sorting the disks using this operation.

cmi2011 descriptive algorithms sorting

Answer

1.16.2 Sorting: CMI2013-A-05 top [Link]

You have n lists, each consisting of m integers sorted in ascending order. Merging these lists into a
single sorted list will take time:

A. O(nm log m)
B. O(mn log n)
C. O(m + n)
D. O(mn)

cmi2013 algorithms sorting

Answer

1.16.3 Sorting: CMI2017-A-08 top [Link]

A stable sort preserves the order of values that are equal with respect to the comparison function.
We have a list of three dimensional points

[(7, 1, 8),(3, 5, 7),(6, 1, 4),(6, 5, 9),(0, 2, 5),(9, 0, 9)].

We sort these in ascending order by the second coordinate. Which of the following corresponds to a
stable sort of this input?

A. [(9, 0, 9),(7, 1, 8),(6, 1, 4),(0, 2, 5),(6, 5, 9),(3, 5, 7)]

© Copyright GATE Overflow. All rights reserved.

168 1 Algorithms (323)

B. [(0, 2, 5),(3, 5, 7),(6, 1, 4),(6, 5, 9),(7, 1, 8),(9, 0, 9)]

C. [(9, 0, 9),(7, 1, 8),(6, 1, 4),(0, 2, 5),(3, 5, 7),(6, 5, 9)]
D. [(9, 0, 9),(6, 1, 4),(7, 1, 8),(0, 2, 5),(3, 5, 7),(6, 5, 9)]

cmi2017 algorithms sorting

Answer

1.16.4 Sorting: GATE1987-1-xviii top [Link]

Let P be a quicksort program to sort numbers in ascending order. Let t1 and t2 be the time taken
by the program for the inputs [1 2 3 4] and [5 4 3 2 1], respectively. Which of the following holds?

A. t1 = t2
B. t1 > t2
C. t1 < t2
D. t1 = t2 + 5 log 5

gate1987 algorithms sorting

Answer

1.16.5 Sorting: GATE1988-1iii top [Link]

Quicksort is ________ efficient than heapsort in the worst case.

gate1988 algorithms sorting

Answer

1.16.6 Sorting: GATE1991-01,vii top [Link]

The minimum number of comparisons required to sort 5 elements is ____

gate1991 normal algorithms sorting

Answer

1.16.7 Sorting: GATE1991-13 top [Link]

Give an optimal algorithm in pseudo-code for sorting a sequence of n numbers which has only k
distinct numbers (k is not known a Priori). Give a brief analysis for the time-complexity of your
algorithm.
gate1991 sorting time-complexity algorithms difficult

Answer

1.16.8 Sorting: GATE1992-02,ix top [Link]

Choose the correct alternatives (more than one may be correct) and write the corresponding letters
only:

Following algorithm(s) can be used to sort n in the range [1 … n3 ] in O(n) time

a. Heap sort
b. Quick sort
c. Merge sort
d. Radix sort

gate1992 easy algorithms sorting

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 169

1.16.9 Sorting: GATE1992-03,iv top [Link]

Assume that the last element of the set is used as partition element in Quicksort. If n distinct
elements from the set [1 … n] are to be sorted, give an input for which Quicksort takes maximum
time.
gate1992 algorithms sorting easy

Answer

1.16.10 Sorting: GATE1994-1.19, ISRO2016-31 top [Link]

Algorithm design technique used in quicksort algorithm is?

A. Dynamic programming

B. Backtracking

C. Divide and conquer

D. Greedy method

gate1994 algorithms sorting easy isro2016

Answer

1.16.11 Sorting: GATE1995-1.16 top [Link]

For merging two sorted lists of sizes m and n into a sorted list of size m + n, we require
comparisons of

A. O(m)
B. O(n)
C. O(m + n)
D. O(log m + log n)

gate1995 algorithms sorting normal

Answer

1.16.12 Sorting: GATE1995-1.5 top [Link]

Merge sort uses:

A. Divide and conquer strategy

B. Backtracking approach

C. Heuristic search

D. Greedy approach

gate1995 algorithms sorting easy

Answer

1.16.13 Sorting: GATE1995-12 top [Link]

© Copyright GATE Overflow. All rights reserved.

170 1 Algorithms (323)

Consider the following sequence of numbers:

92, 37, 52, 12, 11, 25

Use bubblesort to arrange the sequence in ascending order. Give the sequence at the end of each of
the first five passes.
gate1995 algorithms sorting easy

Answer

1.16.14 Sorting: GATE1996-14 top [Link]

A two dimensional array A[1..n][1..n] of integers is partially sorted if

∀i, j ∈ [1..n − 1], A[i][j] < A[i][j + 1] and A[i][j] < A[i + 1][j]
Fill in the blanks:

a. The smallest item in the array is at A[i][j] where i=__ and j=__ .

b. The smallest item is deleted. Complete the following O(n) procedure to insert item x (which is
guaranteed to be smaller than any item in the last row or column) still keeping A partially sorted.

procedure insert (x: integer);

var i,j: integer;
begin
i:=1; j:=1, A[i][j]:=x;
while (x > __ or x > __) do
if A[i+1][j] < A[i][j] ___ then begin
A[i][j]:=A[i+1][j]; i:=i+1;
end
else begin
_____
end
A[i][j]:= ____
end

gate1996 algorithms sorting normal

Answer

1.16.15 Sorting: GATE1996-2.15 top [Link]

Quick-sort is run on two inputs shown below to sort in ascending order taking first element as pivot

i. 1, 2, 3, … n
ii. n, n − 1, n − 2, … , 2, 1
Let C1 and C2 be the number of comparisons made for the inputs (i) and (ii) respectively. Then,

A. C1 < C2
B. C1 > C2
C. C1 = C2
D. we cannot say anything for arbitrary n
gate1996 algorithms sorting normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 171

1.16.16 Sorting: GATE1998-1.22 top [Link]

Give the correct matching for the following pairs:

(A) O
(P) Selection sort
(log n)
(B) O
(n (Q) Insertion sort
)
(C) O
(R) Binary search
(n log n)
(D) O
(S) Merge sort
(n2 )
A. A-R B-P C-Q D-S
B. A-R B-P C-S D-Q
C. A-P B-R C-S D-Q
D. A-P B-S C-R D-Q
gate1998 algorithms sorting easy

Answer

1.16.17 Sorting: GATE1999-1.12 top [Link]

A sorting technique is called stable if

A. it takes O(n log n) time

B. it maintains the relative order of occurrence of non-distinct elements

C. it uses divide and conquer paradigm

D. it takes O(n) space

gate1999 algorithms sorting easy

Answer

1.16.18 Sorting: GATE1999-1.14, ISRO2015-42 top [Link]

If one uses straight two-way merge sort algorithm to sort the following elements in ascending order:

20, 47, 15, 8, 9, 4, 40, 30, 12, 17

then the order of these elements after second pass of the algorithm is:

A. 8, 9, 15, 20, 47, 4, 12, 17, 30, 40

B. 8, 15, 20, 47, 4, 9, 30, 40, 12, 17
C. 15, 20, 47, 4, 8, 9, 12, 30, 40, 17
D. 4, 8, 9, 15, 20, 47, 12, 17, 30, 40

© Copyright GATE Overflow. All rights reserved.

172 1 Algorithms (323)

gate1999 algorithms sorting normal isro2015

Answer

1.16.19 Sorting: GATE1999-8 top [Link]

Let A be an n × n matrix such that the elements in each row and each column are arranged in
ascending order. Draw a decision tree, which finds 1st, 2nd and 3rd smallest elements in minimum
number of comparisons.

gate1999 algorithms sorting normal descriptive

Answer

1.16.20 Sorting: GATE2000-17 top [Link]

An array contains four occurrences of 0, five occurrences of 1, and three occurrences of 2 in any
order. The array is to be sorted using swap operations (elements that are swapped need to be
adjacent).

a. What is the minimum number of swaps needed to sort such an array in the worst case?
b. Give an ordering of elements in the above array so that the minimum number of swaps needed
to sort the array is maximum.

gate2000 algorithms sorting normal descriptive

Answer

1.16.21 Sorting: GATE2001-1.14 top [Link]

Randomized quicksort is an extension of quicksort where the pivot is chosen randomly. What is the
worst case complexity of sorting n numbers using Randomized quicksort?

A. O(n)
B. O(n log n)
C. O(n2 )
D. O(n!)

gate2001 algorithms sorting time-complexity easy

Answer

1.16.22 Sorting: GATE2003-22 top [Link]

The unusual Θ(n2 ) implementation of Insertion Sort to sort an array uses linear search to identify
the position where an element is to be inserted into the already sorted part of the array. If, instead,
we use binary search to identify the position, the worst case running time will

A. remain Θ(n2 )
B. become Θ(n(log n)2 )
C. become Θ(n log n)
D. become Θ(n)

gate2003 algorithms sorting time-complexity normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 173

1.16.23 Sorting: GATE2003-61 top [Link]

In a permutation a1 . . . an , of n distinct integers, an inversion is a pair (ai , aj ) such that i<j and
ai > aj .
If all permutations are equally likely, what is the expected number of inversions in a randomly
chosen permutation of 1.. . n?

n(n−1)
A. 2
n(n−1)
B. 4
n(n+1)
C. 4
D. 2n[log2 n]

gate2003 algorithms sorting normal

Answer

1.16.24 Sorting: GATE2003-62 top [Link]

In a permutation a1 . . . an , of n distinct integers, an inversion is a pair (ai , aj ) such that i<j and
ai > aj .
What would be the worst case time complexity of the Insertion Sort algorithm, if the inputs are
restricted to permutations of 1.. . n with at most n inversions?

A. Θ(n2 )
B. Θ(n log n)
C. Θ(n1.5 )
D. Θ(n)

gate2003 algorithms sorting normal

Answer

1.16.25 Sorting: GATE2004-29 top [Link]

The tightest lower bound on the number of comparisons, in the worst case, for comparison-based
sorting is of the order of

A. n
B. n2
C. n log n
D. n log2 n

gate2004 algorithms sorting asymptotic-notations easy

Answer

1.16.26 Sorting: GATE2005-39 top [Link]

Suppose there are ⌈log n⌉ sorted lists of ⌊n/ log n⌋ elements each. The time complexity of
producing a sorted list of all these elements is: (Hint:Use a heap data structure)

A. O(n log log n)

B. Θ(n log n)
Ω(n log n)
© Copyright GATE Overflow. All rights reserved.
174 1 Algorithms (323)

C. Ω(n log n)
D. Ω (n3/2 )

gate2005 algorithms sorting normal

Answer

1.16.27 Sorting: GATE2005-IT-59 top [Link]

Let a and b be two sorted arrays containing n integers each, in non-decreasing order. Let c be a
sorted array containing 2n integers obtained by merging the two arrays a and b. Assuming the
arrays are indexed starting from 0, consider the following four statements

I. a[i] ≥ b[i] ⇒ c[2i] ≥ a[i]

II. a[i] ≥ b[i] ⇒ c[2i] ≥ b[i]
III. a[i] ≥ b[i] ⇒ c[2i] ≤ a[i]
IV. a[i] ≥ b[i] ⇒ c[2i] ≤ b[i]
Which of the following is TRUE?

A. only I and II
B. only I and IV
C. only II and III
D. only III and IV

gate2005-it algorithms sorting normal

Answer

1.16.28 Sorting: GATE2006-14, ISRO2011-14 top [Link]

Which one of the following in place sorting algorithms needs the minimum number of swaps?

A. Quick sort
B. Insertion sort
C. Selection sort
D. Heap sort

gate2006 algorithms sorting easy isro2011

Answer

1.16.29 Sorting: GATE2006-52 top [Link]

The median of n elements can be found in O(n) time. Which one of the following is correct about
the complexity of quick sort, in which median is selected as pivot?

A. Θ(n)
B. Θ(n log n)
C. Θ(n2 )
D. Θ(n3 )

gate2006 algorithms sorting easy

Answer

1.16.30 Sorting: GATE2007-14 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 175

Which of the following sorting algorithms has the lowest worse-case complexity?

A. Merge sort

B. Bubble sort

C. Quick sort

D. Selection sort

gate2007 algorithms sorting time-complexity easy

Answer

1.16.31 Sorting: GATE2008-43 top [Link]

Consider the Quicksort algorithm. Suppose there is a procedure for finding a pivot element which
splits the list into two sub-lists each of which contains at least one-fifth of the elements. Let T (n) be
the number of comparisons required to sort n elements. Then

A. T (n) ≤ 2T (n/5) + n
B. T (n) ≤ T (n/5) + T (4n/5) + n
C. T (n) ≤ 2T (4n/5) + n
D. T (n) ≤ 2T (n/2) + n

gate2008 algorithms sorting easy

Answer

1.16.32 Sorting: GATE2008-IT-43 top [Link]

If we use Radix Sort to sort n integers in the range (nk/2 , nk ], for some k>0 which is
independent of n, the time taken would be?
A. Θ(n)
B. Θ(kn)
C. Θ(n log n)
D. Θ(n2 )

gate2008-it algorithms sorting normal

Answer

1.16.33 Sorting: GATE2009-11 top [Link]

What is the number of swaps required to sort n elements using selection sort, in the worst case?
A. Θ(n)
B. Θ(n log n)
C. Θ(n2 )
D. Θ(n2 log n)

gate2009 algorithms sorting easy

© Copyright GATE Overflow. All rights reserved.

176 1 Algorithms (323)

Answer

1.16.34 Sorting: GATE2009-39 top [Link]

th
In quick-sort, for sorting n elements, the (n/4) smallest element is selected as pivot using an
O(n) time algorithm. What is the worst case time complexity of the quick sort?
A. Θ(n)
B. Θ(n log n)
C. Θ(n2 )
D. Θ(n2 log n)

gate2009 algorithms sorting normal

Answer

1.16.35 Sorting: GATE2012-39 top [Link]

A list of n strings, each of length n, is sorted into lexicographic order using the merge-sort
algorithm. The worst case running time of this computation is

A. O(n log n)
B. O(n2 log n)
C. O(n2 + log n)
D. O(n2 )

gate2012 algorithms sorting normal

Answer

1.16.36 Sorting: GATE2013-30 top [Link]

The number of elements that can be sorted in Θ(log n) time using heap sort is
A. Θ(1)
−−
−
B. Θ(√logn)
C. Θ( logloglogn n )
D. Θ(log n)

gate2013 algorithms sorting normal

Answer

1.16.37 Sorting: GATE2013-6 top [Link]

Which one of the following is the tightest upper bound that represents the number of swaps required
to sort n numbers using selection sort?

A. O(log n)
B. O(n)
C. O(n log n)
D. O(n2 )

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 177

gate2013 algorithms sorting easy

Answer

1.16.38 Sorting: GATE2014-1-14 top [Link]

Let P be quicksort program to sort numbers in ascending order using the first element as the pivot.
Let t1 and t2 be the number of comparisons made by P for the inputs [1 2 3 4 5] and [4 1 5 3 2]
respectively. Which one of the following holds?

A. t1 =5
B. t1 < t2
C. t1 > t2
D. t1 = t2
gate2014-1 algorithms sorting easy

Answer

1.16.39 Sorting: GATE2014-2-38 top [Link]

Suppose P, Q, R, S, T are sorted sequences having lengths 20, 24, 30, 35, 50 respectively. They
are to be merged into a single sequence by merging together two sequences at a time. The number
of comparisons that will be needed in the worst case by the optimal algorithm for doing this is ____.
gate2014-2 algorithms sorting normal numerical-answers

Answer

1.16.40 Sorting: GATE2014-3-14 top [Link]

You have an array of n elements. Suppose you implement quicksort by always choosing the central
element of the array as the pivot. Then the tightest upper bound for the worst case performance is

A. O(n2 )
B. O(n log n)
C. Θ(n log n)
D. O(n3 )

gate2014-3 algorithms sorting easy

Answer

1.16.41 Sorting: GATE2015-1-2 top [Link]

Which one of the following is the recurrence equation for the worst case time complexity of the quick
sort algorithm for sorting n ( ≥ 2) numbers? In the recurrence equations given in the options below,
c is a constant.
A. T (n) = 2T (n/2) + cn
B. T (n) = T (n − 1) + T (1) + cn
C. T (n) = 2T (n − 1) + cn
D. T (n) = T (n/2) + cn

gate2015-1 algorithms recurrence sorting easy

Answer

1.16.42 Sorting: GATE2015-2-45 top [Link]

© Copyright GATE Overflow. All rights reserved.

178 1 Algorithms (323)

Suppose you are provided with the following function declaration in the C programming language.

int partition(int a[], int n);

The function treats the first element of a[ ] as a pivot and rearranges the array so that all elements
less than or equal to the pivot is in the left part of the array, and all elements greater than the pivot
is in the right part. In addition, it moves the pivot so that the pivot is the last element of the left
part. The return value is the number of elements in the left part.

The following partially given function in the C programming language is used to find the kth smallest
element in an array a[ ] of size n using the partition function. We assume k ≤ n .

int kth_smallest (int a[], int n, int k)

{
int left_end = partition (a, n);
if (left_end+1==k) {
return a[left_end];
}
if (left_end+1 > k) {
return kth_smallest (___________);
} else {
return kth_smallest (___________);
}
}

The missing arguments lists are respectively

A. (a, left_end, k) and (a+left_end+1, n−left_end−1, k−left_end−1)

B. (a, left_end, k) and (a, n−left_end−1, k−left_end−1)
C. (a, left_end+1, n−left_end−1, k−left_end−1) and (a,left_end, k)
D. (a, n−left_end−1, k−left_end−1) and (a,left_end, k)

gate2015-2 algorithms normal sorting

Answer

1.16.43 Sorting: GATE2015-3-27 top [Link]

Assume that a mergesort algorithm in the worst case takes 30 seconds for an input of size 64.
Which of the following most closely approximates the maximum input size of a problem that can be
solved in 6 minutes?

A. 256
B. 512
C. 1024
D. 2018
gate2015-3 algorithms sorting

Answer

1.16.44 Sorting: GATE2016-1-13 top [Link]

The worst case running times of Insertion sort , Merge sort and Quick sort, respectively are:

A. Θ(n log n), Θ(n log n) and Θ(n2 )

B. Θ(n2 ), Θ(n2 ) and Θ(n log n)
C. Θ(n2 ), Θ(n log n) and Θ(n log n)
2) 2)
Θ( Θ(
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 179

D. Θ(n2 ), Θ(n log n) and Θ(n2 )

gate2016-1 algorithms sorting easy

Answer

1.16.45 Sorting: GATE2016-2-13 top [Link]

Assume that the algorithms considered here sort the input sequences in ascending order. If the
input is already in the ascending order, which of the following are TRUE?

I. Quicksort runs in Θ(n2 ) time

II. Bubblesort runs in Θ(n2 ) time
III. Mergesort runs in Θ(n) time
IV. Insertion sort runs in Θ(n) time

A. I and II only
B. I and III only
C. II and IV only
D. I and IV only

gate2016-2 algorithms sorting time-complexity normal ambiguous

Answer

1.16.46 Sorting: ISI2011-A-2a top [Link]

Give a strategy to sort four given distinct integers a, b, c, d in increasing order that minimizes the
number of pairwise comparisons needed to sort any permutation of a, b, c, d.
descriptive isi2011 algorithms sorting

Answer

1.16.47 Sorting: TIFR2010-B-23 top [Link]

Suppose you are given n numbers and you sort them in descending order as follows:
First find the maximum. Remove this element from the list and find the maximum of the remaining
elements, remove this element, and so on, until all elements are exhausted. How many comparisons
does this method require in the worst case?

A. Linear in n.
B. O (n2 ) but not better.
C. O (n log n)
D. Same as heap sort.
E. O (n1.5 ) but not better.

tifr2010 algorithms time-complexity sorting

Answer

1.16.48 Sorting: TIFR2010-B-27 top [Link]

Consider the Insertion Sort procedure given below, which sorts an array L of size n (≥ 2) in
ascending order:

begin
for xindex:= 2 to n do

© Copyright GATE Overflow. All rights reserved.

180 1 Algorithms (323)

x := L [xindex];
j:= xindex - 1;
while j > 0 and L[j] > x do
L[j + 1]:= L[j];
j:= j - 1;
end {while}
L [j + 1]:=X;
end{for}
end

It is known that insertion sort makes at most n(n − 1)/2 comparisons. Which of the following is
true?

A. There is no input on which insertion Sort makes n(n − 1)/2 comparisons.

B. Insertion Sort makes n(n − 1)/2 comparisons when the input is already sorted in ascending
order.
C. Insertion Sort makes n(n − 1)/2 comparisons only when the input is sorted in descending order.
D. There are more than one input orderings where insertion sort makes n(n − 1)/2 comparisons.
E. Insertion Sort makes n(n − 1)/2 comparisons whenever all the elements of L are not distinct.

tifr2010 algorithms sorting

Answer

1.16.49 Sorting: TIFR2011-B-21 top [Link]

LetS = {x1 , . . . . , xn } be a set of n numbers. Consider the problem of storing the elements of S in
an array A [1...n] such that the following min-heap property is maintained for all
2 ≤ i ≤ n : A[⌊i/2⌋] ≤ A[i]. (Note that ⌊x⌋ is the largest integer that is at most x). Which of the
following statements is TRUE?

A. This problem can be solved in O(log n) time.

B. This problem can be solved in O(n) time but not in O(log n) time.
C. This problem can be solved in O(n log n) time but not in O(n) time.
D. This problem can be solved in O (n2 ) time but not in O(n log n) time.
E. None of the above.

tifr2011 algorithms sorting

Answer

1.16.50 Sorting: TIFR2011-B-31 top [Link]

Given a set of n = 2k distinct numbers, we would like to determine the smallest and the second
smallest using comparisons. Which of the following statements is TRUE?

A. Both these elements can be determined using 2k comparisons.

B. Both these elements can be determined using n − 2 comparisons.
C. Both these elements can be determined using n + k − 2 comparisons.
D. 2n − 3 comparisons are necessary to determine these two elements.
E. nk comparisons are necessary to determine these two elements.

tifr2011 algorithms sorting

Answer

1.16.51 Sorting: TIFR2011-B-39 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 181

The first n cells of an array L contain positive integers sorted in decreasing order, and the
remaining m − n cells all contain 0. Then, given an integer x, in how many comparisons can one
find the position of x in L?

A. At least n comparisons are necessary in the worst case.

B. At least log m comparisons are necessary in the worst case.
C. O(log(m − n)) comparisons suffice.
D. O(log n) comparisons suffice.
E. O(log(m/n)) comparisons suffice.

tifr2011 algorithms sorting

Answer

1.16.52 Sorting: TIFR2012-B-13 top [Link]

An array A contains n integers. We wish to sort A in ascending order. We are told that initially no
element of A is more than a distance k away from its final position in the sorted list. Assume that n
and k are large and k is much smaller than n. Which of the following is true for the worst case
complexity of sorting A?

A. A can be sorted with constant . kn comparison but not with fewer comparisons.
B. A cannot be sorted with less than constant . n log n comparisons.
C. A can be sorted with constant . n comparisons.
D. A can be sorted with constant . n log k comparisons but not with fewer comparisons.
E. A can be sorted with constant . k2 n comparisons but not fewer.
tifr2012 algorithms sorting

Answer

1.16.53 Sorting: TIFR2012-B-14 top [Link]

Consider the quick sort algorithm on a set of n numbers, where in every recursive subroutine of the
algorithm, the algorithm chooses the median of that set as the pivot. Then which of the following
statements is TRUE?

A. The running time of the algorithm is Θ(n).

B. The running time of the algorithm is Θ(n log n).
C. The running time of the algorithm is Θ(n1.5 ).
D. The running time of the algorithm is Θ(n2 ).
E. None of the above.

tifr2012 algorithms sorting

Answer

1.16.54 Sorting: TIFR2013-B-20 top [Link]

Suppose n processors are connected in a linear array as shown below. Each processor has a
number. The processors need to exchange numbers so that the numbers eventually appear in
ascending order (the processor P1 should have the minimum value and the the processor Pn should
have the maximum value).

© Copyright GATE Overflow. All rights reserved.

182 1 Algorithms (323)

The algorithm to be employed is the following. Odd numbered processors and even numbered
processors are activated alternate steps; assume that in the first step all the even numbered
processors are activated. When a processor is activated, the number it holds is compared with the
number held by its right-hand neighbour (if one exists) and the smaller of the two numbers is
retained by the activated processor and the bigger stored in its right hand neighbour.
How long does it take for the processors to sort the values?

A. n log n steps
B. n2 steps
C. n steps
D. n1.5 steps
E. The algorithm is not guaranteed to sort

tifr2013 algorithms sorting

Answer

1.16.55 Sorting: TIFR2017-B-7 top [Link]

An array of n distinct elements is said to be un-sorted if for every index i such that 2 ≤ i ≤ n − 1 ,
either A[i] > max{A[i − 1], A[i + 1]}, or A[i] < min{A[i − 1], A[i + 1]}. What is the time-
complexity of the fastest algorithm that takes as input a sorted array A with n distinct elements,
and un-sorts A?

A. O(n log n) but not O(n)

B. O(n) but not O(√− n)
C.
−
O(√n ) but not O(log n)
D. O(log n) but not O(1)
E. O(1)

tifr2017 algorithms sorting

Answer

Answers: Sorting

1.16.1 Sorting: CMI2011-B-06a top [Link]


Selected Answer

Let's say we have disk of radius 0,5,1,4,3,2

We have many disks, one above other, our task is to arrange them from largest disk to smaller
disk.
The only operations available is, take a couple of disks from the top, hold it in your hand, rotate it
& put it back.
Just imagine in a real scenario.
How we will do it.

0,5,1,4,3,2
What I will do is, I will take all plates from 5 to 2 (5, 1, 4, 3, 2) I hold them in my hand, rotated
my hand 180 degree & put it back.
0, 2, 3, 4, 1, 5
Now I will rotate all.
5, 1, 4, 3, 2, 0

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 183
5, 1, 4, 3, 2, 0
Now 5 is at its right place. We will not consider it.

Now find the second largest plate... 4.. rotate from 4 to 0

5 | , 1, 0, 2, 3, 4
rotate from 1 to 4.

5, 4 |, 3, 2, 0, 1
Now 4 is at its right place.

Next maximum is 3, it is at its right place, so we may add one more line in the code to reduce
steps...
Otherwise, the same algo will also keep 3 at it's right place.
The same algo will sort them at last.

Hence, our algo is:

Step 1:
Find the current largest element.
Step 2:
Rotate the elements from largest element to end of the array. (Hold the disks from disk of max size
to disk at top & rotate)
Step 3:
Now rotate the array from current start+1 to top.

Continue this process till it is sorted.

Note that initially current_start = −1 it will be incremented by one every time we get our largest
disk placed at its correct place.

 7 votes -- Ahwan Mishra (10.5k points)

1.16.2 Sorting: CMI2013-A-05 top [Link]


Selected Answer

Since, n lists of each size m.

Since, each list is sorted in ascending order use directly merge procedure of merge sort algo.

Take two list and merge..so one pair will take

2m time.
So, total pairs in first level will be n/2. So total cost for one level is
(n/2) ∗ 2m = nm.
In next level cost for one pair is 4m and no of pairs will be n/4.. so next level cost will be nm.
So, like this each level will have cost nm.
No of levels will be when we have one complete list..

n/2k = 1 ..
k = log2 n .
So, total cost will be log n ∗ (nm)

 17 votes -- sonu (2.3k points)

© Copyright GATE Overflow. All rights reserved.

184 1 Algorithms (323)

1.16.3 Sorting: CMI2017-A-08 top [Link]


Selected Answer

A stable sort preserves the order of values that are equal with respect to the comparison
function.

What does that mean?

That is whenever we compare two keys(values) for sorting and if they are equal, the order in which
they will appear in the sorted list is same as they appeared in the original list. That is, if we have
x1 and x2 in our list, such that x1 = x2 (they are equal when compared) and x1 appeared before
x2 in the list (our list is something like: . . . , x1 , . . . , x2 , . . . ).
Then the sorted list will be like: . . . , x1 , x2 , . . . and NOT like: . . . , x2 , x1 , . . . . That is, in the
sorted list x1 comes before x2 . Then such a sort is known to be stable.

[Link]
it-important

Now the given list:

[(7, 1, 8), (3, 5, 7), (6, 1, 4), (6, 5, 9), (0, 2, 5), (9, 0, 9)]
is sorted in ascending order by the second coordinate. We have to compare second coordinate to
get the ordering.

Answer should be:

(c)
[(9, 0, 9), (7, 1, 8), (6, 1, 4), (0, 2, 5), (3, 5, 7), (6, 5, 9)]
Notice here that when compared, (7, 1, 8) is equal to (6, 1, 4) (compare second coordinates). So,
in the sorted list (7, 1, 8) comes before (6, 1, 4), same order as in original list.

Similarly for (3, 5, 7) and (6, 5, 9).

 1 votes -- Rishabh Gupta (13.9k points)

1.16.4 Sorting: GATE1987-1-xviii top [Link]


Selected Answer

ACTUALLY IN BOTH THE CASES IT WILL TAKE O(n2 ) TIME (O(n) TIME FOR PARTITION
ALGORITHM AND T (n − 1) TIME FOR SUB PROBLEM. AS n IS THE NUMBER OF INPUTS AND IN
THE 2ND CASE INPUTS ARE 5(GREATER THAN 1ST ONE THAT IS 4) I THINK t1 < t2

 18 votes -- Rohan Ghosh (2k points)

1.16.5 Sorting: GATE1988-1iii top [Link]


Selected Answer

Answer is LESS.

As worst case time for Quicksort is O(n2 ) and worst case for heap sort is O(n log n).

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 185

 10 votes -- Pavan Kumar Munnam (10.9k points)

1.16.6 Sorting: GATE1991-01,vii top [Link]


Selected Answer

Answer is 7.

inimum number of comparisons = ⌈log(n!)⌉ = ⌈log(5!)⌉ = ⌈log(120)⌉ = 7.

Reference:
[Link]

 25 votes -- Rajarshi Sarkar (34.1k points)

1.16.7 Sorting: GATE1991-13 top [Link]


Selected Answer

Answer should be counting sort which will take O(n + k) time.

See here:

[Link]

 11 votes -- One (2.3k points)

1.16.8 Sorting: GATE1992-02,ix top [Link]


Selected Answer

Answer is (D) Part.

Although people have provided correct answers but it seems some more explanation is required.
Let there be d digits in max input integer, b is the base for representing input numbers and n is
total numbers then Radix Sort takes O(d ∗ (n + b)) time. Sorting is performed from least
significant digit to most significant digit.

For example, for decimal system, b is 10. What is the value of d? If k is the maximum possible
value, then d would be O(logb (k)). So overall time complexity is O((n + b) ∗ logb (k)). Which
looks more than the time complexity of comparison based sorting algorithms for a large k. Let us
first limit k. Let k ⩽ nc where c is a constant. In that case, the complexity becomes
O(n logb (n)). But it still does not beat comparison based sorting algorithms.
What if we make value of b larger?. What should be the value of b to make the time complexity
linear? If we set b as n then we will get the time complexity as O(n).

In other words, we can sort an array of integers with range from 1 to nc , If the numbers are
represented in base n (or every digit takes log2 (n) bits).

Reference: [Link]

 15 votes -- Chhotu Ram Chauhan (10.5k points)

© Copyright GATE Overflow. All rights reserved.

186 1 Algorithms (323)

1.16.9 Sorting: GATE1992-03,iv top [Link]


Selected Answer

The algorithm will take maximum time when:

1. The array is already sorted in same order.

2. The array is already sorted in reverse order.
3. All elements are same in the array.

 25 votes -- Rajarshi Sarkar (34.1k points)

1.16.10 Sorting: GATE1994-1.19, ISRO2016-31 top [Link]


Selected Answer

Answer: Option C.

It is one of the efficient algorithms in Divide and Conquer strategy.

 22 votes -- Gate Keeda (19.6k points)

1.16.11 Sorting: GATE1995-1.16 top [Link]


Selected Answer

Answer: Option C.

The number of moves are however always m + n so that we can term it as Θ(m + n). But the
number of comparisons vary as per the input. In the best case the comparisons are Min(m, n)
and in worst case they are m + n − 1 .

 26 votes -- Gate Keeda (19.6k points)

1.16.12 Sorting: GATE1995-1.5 top [Link]


Selected Answer

Answer: Option C.

One of the best examples of Divide and Conquer strategy.

 16 votes -- Gate Keeda (19.6k points)

1.16.13 Sorting: GATE1995-12 top [Link]


Selected Answer

1st Pass: 37 52 12 11 25 92

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 187

2nd Pass: 37 12 11 25 52 92
3rd Pass: 12 11 25 37 52 92
4th Pass: 11 12 25 37 52 92
5th Pass: 11 12 25 37 52 92

 17 votes -- Gate Keeda (19.6k points)

1.16.14 Sorting: GATE1996-14 top [Link]


Selected Answer

a. The smallest element is at index 1,1.

b. So we have to give an array which is partially sorted. Definition of partially sorted is given in the
question.
We will give the value of x which is less than last row & column value.
At last, 1,1 should be deleted & x should be at its correct place.

i=1;j=1; a[i][j]=x;
while ((x>a[i+1][j]) || (x>a[i][j+1]))
{
if((a[i+1][j] < x) && (a[i+1][j] <a[i][j+1]))
{
a[i][j]=a[i+1][j];
i=i+1;
}
else
{
a[i][j]=a[i][j+1];
j=j+1;
}
}
a[i][j]=x;

Enter the dimension of n × n array. Give the value of n

3
Enter the array elements in partially sorted order
239
5 6 10
8 11 15
Enter the value of x
7
The final output.
3 6 9
5 7 10
8 11 15

 12 votes -- Ahwan Mishra (10.5k points)

1.16.15 Sorting: GATE1996-2.15 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

188 1 Algorithms (323)

C.
both are the worst cases of quick sort. (assuming pivot is either first or last element)

i. is sorted in ascending order.

ii. is sorted in descending order.

 20 votes -- Gate Keeda (19.6k points)

1.16.16 Sorting: GATE1998-1.22 top [Link]


Selected Answer

Selection sort : O(n2 )

Merge sort : O(n log n)
Binary search : (log n)
Insertion sort : O(n)

Note: if we use O(n2 ) for Insertion sort, we will not be having any suitable choice to fill selection
sort. So, we can assume that the question is asking for best case time complexities.

 13 votes -- Bhagirathi Nayak (14.1k points)

1.16.17 Sorting: GATE1999-1.12 top [Link]


Selected Answer

(B) If it maintains the relative order of occurrence of non-distinct elements.

(from definition of stable sorting)

 16 votes -- Arjun Suresh (350k points)

1.16.18 Sorting: GATE1999-1.14, ISRO2015-42 top [Link]


Selected Answer

20 47 15 8 9 4 40 30 12 17
\ / \ / \ / \ / \ /
20 47 8 15 49 30 40 12 17 after 1st pass
\ / \ / \ /
\ / \ / \/
8, 15, 20, 47 4, 9, 30, 40 12, 17 after 2nd pass

Answer is B.

 54 votes -- Vikrant Singh (13.5k points)

1.16.19 Sorting: GATE1999-8 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 189

This is a two dimensional array of size 4 × 4.

2____5____8____9
4____11___19___21
6____13___24___27
8____15___25___29
You can see the elements in each row and each column are arranged in ascending order.
Smallest element: A[0][0] = 2
2nd Smallest element: min(A[0][1], A[1][0]) = min(5, 4) = 4
3rd smallest element: Just exclude the element you got as 2nd smallest(4). Here, we can compare
A[2][0], A[0][1] no need to compare with A[0][2]. S,o it depends upon from where you got 2nd
element. You can draw a decision tree. If you got 2nd best from A[0][1] then what to do & if you
get from A[1][0] then what to do.

Any way, time complexity is simply

O(1).
The elements are in ascending order. Not in non decreasing order. Clearly they are all distinct in a
particular row or column.

 14 votes -- Ahwan Mishra (10.5k points)

1.16.20 Sorting: GATE2000-17 top [Link]


Selected Answer

Since swaps are needed to be of adjacent elements only, the algorithm is actually Bubble sort.

In bubble sort, all smaller elements to right of an element are required to be swapped. So, if have
ordering

[2, 2, 2, 1, 1, 1, 1, 1, 0, 0, 0, 0], then we need total 47 swaps, and this will be the worst case.
So, it answers actually both parts.

 23 votes -- Happy Mittal (11.2k points)

1.16.21 Sorting: GATE2001-1.14 top [Link]


Selected Answer

Answer will be (C).

There are following two cases, when Randomized Quick Sort will result into worstcase of time
complexity O(n2 )

1. When all elements are same in the input array, Partition algo will divide input array in two sub-
array, one with n − 1 elements and second with 0 element. There is an assumption here that,
we are using the same partition algorithm without any modification.

2. If the randomised pivot selector happens to select e.g. the smallest element N times in a row,
we will get the worst possible performance. Though the probability of this particular case is
1
about n! "

PS:- If the partitioning is unbalanced, Quick Sort algorithm runs asymptotically as slow as
Insertion Sort i.e O(n2 )

© Copyright GATE Overflow. All rights reserved.

190 1 Algorithms (323)

 12 votes -- Manu Thakur (39.6k points)

1.16.22 Sorting: GATE2003-22 top [Link]


Selected Answer

In insertion sort, with linear search, it takes

(worst case) n comparisons for searching the right position, and n swaps to make room to place
the element.

Hence for n elements, a total of n × (n + n) ; n for search and n for swaps.

= Θ(2n2 ) = Θ(n2 )
If we replace it with binary search, it takes

(worst case) log n comparisons for searching the right position, and n swaps to make room to
place the element.

Hence for n elements, a total of n × (log n + n) ; n for search and n for swaps.
= Θ(n × log n + n2 ) = Θ(n2 )
Hence, answer is A.

 38 votes -- ryan sequeira (3.4k points)

1.16.23 Sorting: GATE2003-61 top [Link]


Selected Answer

They are asking the average number of inversion. basically what i learned about averages from
dbms indexing is.

Aaverage apart from the standard definition can be calculated as (best case + worst case)/2
and inversion is like 9, 5.
So, best case will be sorted array −1, 2, 3, 4, 5 no inversion .= zero
worst case = 9, 5, 4, 3, 2, 1 . here total number of inversion will be n(n − 1)/2 as . 9 can be
paired with any 5 elements (5, 4, 3, 2, 1) will form a inversion pair. similarly 5 with [Link] .

So, we can say if we have n elements then. it will be (n − 1) + (n − 2) + (n − 3). . . +2 + 1

which is the sum of first n − 1 natural numbers. So, it is n(n − 1)/2

So, expected average number of inversion = (n(n − 1)/2 + zero(best case))/2 = n(n − 1)/4
So, option is B.

Second question.

we all know that insertion sort has complexity due to swapping and movements, if we have n n
inversion pair then the movements and comparison will be restricted to n only . like if inversion is
1 , then array must be sorted and only the inversion should exist at the end, like 1, 2, 3, 5, 4.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 191
1, 2, 3, 5, 4
otherwise more than one inversion pair will form. so to sort this. for two it will be 1, 2, 3, 7, 5, 4. so
to sort this type of array using insertion sort atmost N swaps wiill be required, so D,

 36 votes -- Ravi Singh (15.7k points)

1.16.24 Sorting: GATE2003-62 top [Link]


Selected Answer

ANSWER: D. Θ(n)
REASON:

Count of number of times the inner loop of insertion sort executes is actually equal to number of
inversions in input permutation a1 , a2 , … an . Since for each value of i = 1...n, j take the value
1...i − 1, which means for every j < i it checks if a[j] > a[i].
In any given permutation, maximum number of inversions possible is n(n − 1)/2 which is O(n2 ).
It is the case where the array is sorted in reverse order. Therefore, to resolve all inversions i.e.,
worst case time complexity of insertion sort is Θ(n2 ).

However, as per the question the number of inversion in input array is restricted to n. The worst
case time complexity of insertion sort reduces to Θ(n).

INSERTION SORT ALGORITHM (for reference)

 25 votes -- Prateek Dwivedi (4.4k points)

1.16.25 Sorting: GATE2004-29 top [Link]


Selected Answer

For comparison-based sorting the asymptotically tight bound for worst case is given by
Θ(n log n), which means it is the tightest upper bound (big O) as well as the tightest lower bound
(big omega). So, answer is n log n.

Tightest lower bound of sorting (say S(n)) is n log n means there is no function f which has an
order of growth larger than n log n and f(n) = Ω(S(n)) holds.

A usual mistake is to think worst case changes with lower and upper bounds, but that is not the
case. Worst case is defined for the algorithm and it is always the input which causes the algorithm
the maximum complexity.

 37 votes -- Arjun Suresh (350k points)

1.16.26 Sorting: GATE2005-39 top [Link]


Selected Answer

Since we have log n lists we can make a min-heap of log n elements by taking the first element
from each of the log n sorted lists. Now, we start deleting the min-element from the heap and put
the next element from the sorted list from which that element was added to the heap. (This
identity can be done by making a structure of two values, one for the number and one for
identifying the origin sorted list of that number and storing this structure in the heap). In this way

O(log log n)
© Copyright GATE Overflow. All rights reserved.
192 1 Algorithms (323)

each delete and the corresponding insert will take O(log log n) time as delete in heap of size n is
O(log n) and inserting an element on a heap of size n is also O(log n). (here, heap size is log n).
Now, we have a total of log n × logn n = n elements. So, total time will be O(n log log n).

 42 votes -- gatecse (18k points)

1.16.27 Sorting: GATE2005-IT-59 top [Link]


Selected Answer

a[i] ≥ b[i]
Since both a and b are sorted in the beginning, there are i elements smaller than or equal to a[i](i
starts from 0), and similarly i elements smaller than or equal to b[i]. So, a[i] ≥ b[i] means there
are 2i elements smaller than or equal to a[i], and hence in the merged array a[i] will come after
these 2i elements (its index will be > 2i). So, c[2i] ≤ a[i] (equality takes care of the "equal to"
case which comes when array contains repeated elements).

Similarly, a[i] ≥ b[i] says for b that, there are not more than 2i elements smaller than b[i] in the
sorted array (i elements from b, and maximum another i elements from a). So, b[i] ≤ c[2i]

So, II and III are correct is option (C).

 26 votes -- Arjun Suresh (350k points)

1.16.28 Sorting: GATE2006-14, ISRO2011-14 top [Link]


Selected Answer

Selection sort.

Because in selection the maximum swaps which can take place are O(n)
Because we pick up an element an find the minimum (in case of forward sorting) from the next
index till the end of array and than perform the swap

Hence, O(n) whereas in all other algos the swaps are greater ( considering Worst-Case scenario )
 19 votes -- ANKUR MAHIWAL (411 points)

1.16.29 Sorting: GATE2006-52 top [Link]


Selected Answer

As we choose the pivot a median element ... so, every time we are going to have good splits
guaranteed so the best case O(n log n).

 23 votes -- Bhagirathi Nayak (14.1k points)

1.16.30 Sorting: GATE2007-14 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 193

A.

Irrespective of the input, merge sort always have a time complexity of Θ(n log n).

 17 votes -- Gate Keeda (19.6k points)

1.16.31 Sorting: GATE2008-43 top [Link]


Selected Answer

T (n) ≤ T (n/5) + T (4n/5) + n

One part contains n/5 elements
and the other part contains 4n/5 elements
+n is common to all options, so we need not to worry about it.
Hence, answer is option B.

 22 votes -- Amar Vashishth (30.5k points)

1.16.32 Sorting: GATE2008-IT-43 top [Link]


Selected Answer

Answer: C

The complexity of Radix Sort is O(wn), for n keys which are integers of word size w.
Here, w = log2 (nk ) = k × log2 (n)
So, the complexity is O(wn) = O(k × log2 (n) × n), which leads to option C.

 34 votes -- Rajarshi Sarkar (34.1k points)

1.16.33 Sorting: GATE2009-11 top [Link]


Selected Answer

The answer is A.

we have 1 swap in each loop and hence n swaps at max for 1 to n. Therefore the worst case
number of swaps is Θ(n)

 21 votes -- Gate Keeda (19.6k points)

1.16.34 Sorting: GATE2009-39 top [Link]


Selected Answer

Answer is B.

T (n) = O(n) +T (n/4 − 1) + T (3n/4)

© Copyright GATE Overflow. All rights reserved.
194 1 Algorithms (323)

T (n) = O(n) pivot selection time +T (n/4 − 1) + T (3n/4)

which'll give Θ (n log n).
Pivot selection complexity is given in questions. Pivot being the (n/4)th smallest element, once it
is found, we have two sub arrays- one of size (n/4 − 1) and other of size (3n/4) and for both of
these we solve recursively.

 29 votes -- Gate Keeda (19.6k points)

1.16.35 Sorting: GATE2012-39 top [Link]


Selected Answer

I thought like this:

you are given the first character of each n strings to sort it will take O(n log n) time..in the worst
case we may have to do the above process 2 times,3 times,........,n times so
n ∗ O(n log n) = O(n2 log n) please correct me if my approach is wrong...
 39 votes -- Bhagirathi Nayak (14.1k points)

1.16.36 Sorting: GATE2013-30 top [Link]


Selected Answer

log n
To sort k elements in a heap, complexity is Θ(k log k). Lets assume there are log log n
elements in
the heap.

Complexity = Θ ( logloglogn n log( logloglogn n ))

= Θ ( logloglogn n (log log n − log log log n))

= Θ (log n − log n log log log n

log log n
)

= Θ(log n) (as shown below)

So, (C) is the answer.

log log n > log log log n

log log log n
⟹ log log n
<1
log n log log log n
⟹ log log n
< log n

⟹ Θ (log n − log n log log log n

log log n
) = Θ(log n)

 93 votes -- Arjun Suresh (350k points)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 195

1.16.37 Sorting: GATE2013-6 top [Link]


Selected Answer

In selection max you can do is n swaps..selecting the smallest element from all the elements and
replacing it correct position so O(n)

 23 votes -- Bhagirathi Nayak (14.1k points)

1.16.38 Sorting: GATE2014-1-14 top [Link]


Selected Answer

it would be t1 > t2 , because the first case is the worst case of quicksort i.e. minimum number is
chosen as pivot. Hence in the worst case the comparisons are high.

The splitting occurs as

[1][2345]
[2][345]
[3][45]
[4][5]
and

[123][45]
[1][23][4][5]
[2][3]
Number of recursive calls remain the same, but in second case the number of elements passed for
the recursive call is less and hence the number of comparisons also less.

 29 votes -- Parul Agarwal (809 points)

1.16.39 Sorting: GATE2014-2-38 top [Link]


Selected Answer

The optimal algorithm always chooses the smallest sequences for merging.

20 24 − 44, 43 comparisons
30 35 − 65, 64 comparisons
44 50 − 94, 93 comparisons
65 94 − 159, 158 comparisons
so, totally 43 + 64 + 93 + 158 = 358 comparisons.

PS: In merge operation we do a comparison of two elements and put one element in the sorted
output array. So, every comparison produces one output element. But for the last element we
won't need a comparison and we simply insert it to the output array. So for n output elements we
need (n − 1) comparisons.

 73 votes -- Arjun Suresh (350k points)

1.16.40 Sorting: GATE2014-3-14 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

196 1 Algorithms (323)

(A) O(n2 ) is the answer. When we choose the first element as the pivot, the worst case of quick
sort comes if the input is sorted- either in ascending or descending order. Now, when we choose
the middle element as pivot, sorted input no longer gives worst case behavior. But, there will be
some permutation of the input numbers which will be giving the same worst case behavior. For
example,

1234567
This array gives worst case behavior for quick sort when the first element is pivot.

6421357
This array gives the worst case behavior of O(n2 ) if we take middle element as the pivot- each
split will be 1 element on one side and n − 1 elements on other side. Similarly, for any input, we
can have a permutation where the behavior is like this. So, whichever element we take as pivot it
gives worst case complexity of O(n2 ) as long as pivot is from a fixed position (not random position
as in randomized quick sort).

 45 votes -- Arjun Suresh (350k points)

1.16.41 Sorting: GATE2015-1-2 top [Link]


Selected Answer

B.
Worst case for quick sort happens when 1 element is on one list and n − 1 elements on another
list.

 26 votes -- Arjun Suresh (350k points)

1.16.42 Sorting: GATE2015-2-45 top [Link]


Selected Answer

First of all, here the return value is the number of elements less than the pivot

Pivot is just to minimize searching

So, now we are assuming our array has 10 elements, N = 10, k = 8

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 197

So, in STEP 1 and STEP 2 'else' condition satisfying, and STEP 3 and STEP 4 'if ' condition satisfying

Here, partition is calling and returning left_end value

Answer will be (A).

 21 votes -- srestha (87k points)

© Copyright GATE Overflow. All rights reserved.

198 1 Algorithms (323)

1.16.43 Sorting: GATE2015-3-27 top [Link]


Selected Answer

The worst case time complexity of Mergesort is k × n log n for an input of size n.
For an input of size 64, the algorithm takes 30s. Therefore,
k × 64 log2 64 = 30s

k × 384 = 30s

⟹ k = 0.078125s

Let the size of the problem that can be solved in 6 minutes be x. Then,
k × x log2 x = 360s

From this, we get:

360s
x log2 x =
0.078125s

⟹ x = 512

 45 votes -- Pragy Agarwal (20.6k points)

1.16.44 Sorting: GATE2016-1-13 top [Link]


Selected Answer

Answer is D.

Insertion sort: = Θ(n2 )

Merge sort: = Θ(n log n)
Quick sort: = Θ(n2 )
Note : here Θ is not average case since question asked worst case so Θ represent worst case only

 25 votes -- Abhilash Panicker (9.5k points)

1.16.45 Sorting: GATE2016-2-13 top [Link]


Selected Answer

Q = Q can be used instead of Theta

I. Quicksort takes Q(N 2 ) in case of already sorted input. This is true
II. This is false. If no swap happens then bubble sort can stop in single loop. Q(N) is best case.
This is false !
Q(N log N)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 199

III. Mergesort never takes more than Q(N log N) This is false
IV. This is true. Insertion sort will finish in Q(N) time in case of sorted input.

Answer D. I and IV.

Proof Bubble sort has best case O(N) =>

Ref: [Link] Aduni lecture

Now quicksort taking

O(N) can happen in some cases but not all cases, so that is why I) should be considered
true. Whereas Bubble sort time complexity in best case is always
O(N) . So D is any time stronger answer than C .

 29 votes -- Akash Kanase (42.5k points)

1.16.46 Sorting: ISI2011-A-2a top [Link]

It will be merge sort.

a, b, c, d
1 + 1 comparisons for lowest step
+3 comparisons for upper one
So, total 5 comparisons

 6 votes -- Tanushree (385 points)

1.16.47 Sorting: TIFR2010-B-23 top [Link]


Selected Answer

The given procedure resembles something like Bubble sort or Selection Sort.
Every time, for every input it will take O(n2 ) so, B is the answer.

If you are thinking about heap sort, where in the procedure, it is talked about building a heap?
You can do extract_max only on a heap. Right ? The procedure is nearly same as bubble sort or
selection sort.

Read qsn again: "First find the maximum. Remove this element from the list and find the

© Copyright GATE Overflow. All rights reserved.

200 1 Algorithms (323)

maximum of the remaining elements, remove this element, and so on, until all elements are
exhausted."

Just follow the steps. Find max, take it to right side of array. Shift.
So n − 1 comparisons & n − 1 shifts in the worst case.
Total n − 1 + n − 2 + n − 3+. . . . +1 + n − 1 + n − 2 + n − 3+. . +1 which is O(n2 )
Only number of comparisons is asked... which is also O(n2 )

 7 votes -- Ahwan Mishra (10.5k points)

1.16.48 Sorting: TIFR2010-B-27 top [Link]


Selected Answer

In worst case Insertion sort will have n(n − 1)/2 comparisons i.e. when input is sorted in
descending order.

50 40 30 20 10 … n
pass 1: 50 40 30 20 10 … n 0 comparison
pass 2: 40 50 30 20 10 … n 1 comparison
.
.
.
.
pass n: n … 10 20 30 40 50 n − 1 comparisons
Total 1 + 2 + 3 + … + n − 1 = n(n − 1)/2 comparisons

 7 votes -- Umang Raman (15.7k points)

1.16.49 Sorting: TIFR2011-B-21 top [Link]


Selected Answer

store the elements in an array and then call build_heap(A). the build_heap takes O(n) time.

so, option 'b' is correct.

but, if we try building heap by inserting each element one by one, the total complexity will be then
O(n log n). cause insertion takes O(log n) and inserting 'n' elements will take O(n log n).

 17 votes -- Sujit Kumar Muduli (209 points)

1.16.50 Sorting: TIFR2011-B-31 top [Link]


Selected Answer

Option (c) n+k−2

Here is a nice explanation of the algorithm: [Link]
smallest

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 201

it is solution to the problem known for ages, and it has to do with tennis tournaments. The
question was, knowing the outcome of the tennis tournament, how can we tell which player
was the second best? The defeated finalist is a good candidate, but there are other players
that were defeated directly by the tournament winner and any of them could also be a good
candidate for the second best. So the solution to the problem is quite simple: Once the
tournament finishes, pick up the log N competitors that were beaten by the tournament
winner and hold a mini-tournament to find which one is the best among them. If we imagine
that better players correspond with smaller numbers, the algorithm now goes like this. Hold
the tournament to find the smallest number (requires N − 1 comparisons). During this step,
for each number construct the list of numbers it was smaller than. Finally, pick the list of
numbers associated with the smallest number and find their minimum in log N − 1 steps.
This algorithm requires N + log N − 2 comparisons to complete, but unfortunately it
requires additional space proportional to N (each element except the winner will ultimately
be added to someone’s list); it also requires more time per step because of the relatively
complex enlisting logic involved in each comparison. When this optimized algorithm is
applied to example array, we get the following figure.

Tournament held among numbers promotes value 1 as the smallest number. That operation,
performed on an array with nine numbers, requires exactly eight comparisons. While
promoting the smallest number, this operation has also flagged four numbers that were
removed from competition by direct comparison with the future winner: 6, 2, 3 and 8 in that
order. Another sequence of three comparisons is required to promote number 2 as the
second-smallest number in the array. This totals 11 comparisons, while naive algorithm
requires 17 comparisons to come up with the same result.

All in all, this algorithm that minimizes number of comparisons looks to be good only for real
tournaments, while number cracking algorithms should keep with the simple logic explained
above. Implementation of simple algorithm may look like this:

a – array containing n elements

min1 = a[0] – candidate for the smallest value
min2 = a[1] – candidate for the second smallest value
if min2 < min1
min1 = a[1]
min2 = a[0]
for i = 2 to n – 1
if a[i] < min1
min2 = min1
min1 = a[i]
else if a[i] < min2
min2 = a[i]

by Zoran Horvat @zoranh75

 24 votes -- Pragy Agarwal (20.6k points)

1.16.51 Sorting: TIFR2011-B-39 top [Link]


Selected Answer

(d)
O(log n) comparisons suffice.

O(n)
© Copyright GATE Overflow. All rights reserved.
202 1 Algorithms (323)

Since it is possible that m ⋙ n, we need to restrict ourselves to the first O(n) elements to
perform the binary search.

We start with the first element (index i = 1), and check if it is equal to 0. If not, we double the
value of i, and check again. We repeat this process until we hit a 0.

i = 1;
while(arr[i] != 0)
i *= 2;

Once we hit a 0, the largest possible value (worst case) of i can be 2n − 2. This will happen if
n = 2k + 1 for some k. Then, our 2nd last value of i will be 2k , and then we get 2k+1 , which is
equal to 2n − 2.

Now that we've hit a 0, and the array contains positive numbers in decreasing order, if x is present
in L, it must be in the first i elements.

We can binary search the first i elements in O(log i) comparisons.

Since the largest possible value of i = 2n − 2 , our algorithm takes O(log(2n − 2)) = O(log n)
comparisons.

 15 votes -- Pragy Agarwal (20.6k points)

1.16.52 Sorting: TIFR2012-B-13 top [Link]


Selected Answer

Let Array element be {4, 3, 2, 1, 7, 5, 6, 9, 10, 8} and K be 3 here no element is more than 3
distance away from its final position
So if we take
arr(1 to 6) and sort then surely first three element will be sorted in its final position
{12345769108} O(6 log 6)
then sort arr(3 to 9) then 3 to 6 will be sorted {12345679108} O(6 log 6)
then at last arr(6 to 9) less than O(6 log 6) {12345678910}

in general
Sort arr(0 to 2k)
Now we know that arr[0 to k) are in their final sorted positions
and arr(k to 2k) may be not sorted.

Sort arr(k to 3k)

Now we know that arr[k to 2k) are in their final sorted positions
and arr(2k to 3k) may be not sorted.
.
.
.
.

sort till arr(ik. . N)

in final sorting there will be less than 2k element.
in each step it will take O(2k log 2k)
and there will n steps so O(n log k)
k
option D.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 203

 10 votes -- Umang Raman (15.7k points)

1.16.53 Sorting: TIFR2012-B-14 top [Link]


Selected Answer

Algorithm is choosing median = n/2 smallest element as pivot.

Hence, the array is divided as:

Median at
( n2 −1) (n− n2 )
2 th
n
elements elements
location

Therefore quick sort recurrence relation is given by:

n n
T (n) = T ( − 1) + T (n − ) + Θ(n)
2 2

= Θ(n log n)

Hence, Option B is the correct answer.

 15 votes -- Umang Raman (15.7k points)

1.16.54 Sorting: TIFR2013-B-20 top [Link]

Answer: C.

© Copyright GATE Overflow. All rights reserved.

204 1 Algorithms (323)

 10 votes -- Vidhi Sethi (8.7k points)

1.16.55 Sorting: TIFR2017-B-7 top [Link]


Selected Answer

A pairwise swap will make the sorted array unsorted. Hence, the option (B) is correct.

For eg - if an array is 12345678

The array will become after a pair wise swap to 2 1 4 3 6 5 8 7. For all i between 2 and n − 1,
a[i] is either lower, or either greater than their adjacent elements.
Since, each element is being swapped exactly once. The operation has O(n) time complexity.

 9 votes -- tarun_svbk (1.5k points)

1.17 Spanning Tree(31) top

1.17.1 Spanning Tree: GATE1991-03,vi top [Link]

Choose the correct alternatives (more than one may be correct) and write the corresponding letters
only:

Kruskal’s algorithm for finding a minimum spanning tree of a weighted graph G with n vertices and
m edges has the time complexity of:

A. O(n2 )
B. O(mn)
C. O(m + n)
D. O(m log n)
E. O(m2 )

gate1991 algorithms spanning-tree

Answer

1.17.2 Spanning Tree: GATE1992-01,ix top [Link]

Complexity of Kruskal’s algorithm for finding the minimum spanning tree of an undirected graph
containing n vertices and m edges if the edges are sorted is _______
gate1992 spanning-tree algorithms time-complexity easy

Answer

1.17.3 Spanning Tree: GATE1995-22 top [Link]

How many minimum spanning trees does the following graph have? Draw them. (Weights are
assigned to edges).

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 205

gate1995 algorithms graph-algorithms spanning-tree easy

Answer

1.17.4 Spanning Tree: GATE1996-16 top [Link]

A complete, undirected, weighted graph G is given on the vertex {0, 1, … , n − 1} for any fixed
‘n’. Draw the minimum spanning tree of G if

A. the weight of the edge (u, v) is ∣u − v∣

B. the weight of the edge (u, v) is u + v

gate1996 algorithms graph-algorithms spanning-tree normal

Answer

1.17.5 Spanning Tree: GATE1997-9 top [Link]

Consider a graph whose vertices are points in the plane with integer co-ordinates (x, y) such that
1 ≤ x ≤ n and 1 ≤ y ≤ n, where n ≥ 2 is an integer. Two vertices (x1 , y1 ) and (x2 , y2 ) are
adjacent iff ∣x1 − x2 ∣≤ 1 and ∣ y1 – y2 ∣≤ 1 . The weight of an edge
−−−−−−−−−−−−−−−−
{(x1 , y1 ), (x2 , y2 )} is √(x1 – x2 )2 + (y1 – y2 )2
A. What is the weight of a minimum weight-spanning tree in this graph? Write only the answer
without any explanations.

B. What is the weight of a maximum weight-spanning tree in this graph? Write only the answer
without any explanations.

gate1997 algorithms spanning-tree normal

Answer

1.17.6 Spanning Tree: GATE2000-2.18 top [Link]

Let G be an undirected connected graph with distinct edge weights. Let emax be the edge with
maximum weight and emin the edge with minimum weight. Which of the following statements is
false?

A. Every minimum spanning tree of G must contain emin

B. If emax is in a minimum spanning tree, then its removal must disconnect G
C. No minimum spanning tree contains emax
D. G has a unique minimum spanning tree

gate2000 algorithms spanning-tree normal

Answer

1.17.7 Spanning Tree: GATE2001-15 top [Link]

Consider a weighted undirected graph with vertex set V = {n1, n2, n3, n4, n5, n6} and edge set

E = {(n1, n2, 2), (n1, n3, 8), (n1, n6, 3), (n2, n4, 4), (n2, n5, 12), (n3, n4, 7), (n4, n5, 9), (n4, n
© Copyright GATE Overflow. All rights reserved.
206 1 Algorithms (323)

E = {(n1, n2, 2), (n1, n3, 8), (n1, n6, 3), (n2, n4, 4), (n2, n5, 12), (n3, n4, 7), (n4, n5, 9), (n4, n
The third value in each tuple represents the weight of the edge specified in the tuple.

A. List the edges of a minimum spanning tree of the graph.

B. How many distinct minimum spanning trees does this graph have?
C. Is the minimum among the edge weights of a minimum spanning tree unique over all possible
minimum spanning trees of a graph?
D. Is the maximum among the edge weights of a minimum spanning tree unique over all possible
minimum spanning tree of a graph?

gate2001 algorithms spanning-tree normal descriptive

Answer

1.17.8 Spanning Tree: GATE2003-68 top [Link]

What is the weight of a minimum spanning tree of the following graph?

A. 29
B. 31
C. 38
D. 41
gate2003 algorithms spanning-tree normal

Answer

1.17.9 Spanning Tree: GATE2005-6 top [Link]

An undirected graph G has n nodes. its adjacency matrix is given by an n × n square matrix whose
(i) diagonal elements are 0’s and (ii) non-diagonal elements are 1’s. Which one of the following is
TRUE?

A. Graph G has no minimum spanning tree (MST)

B. Graph G has unique MST of cost n − 1
C. Graph G has multiple distinct MSTs, each of cost n − 1
D. Graph G has multiple spanning trees of different costs
gate2005 algorithms spanning-tree normal

Answer

1.17.10 Spanning Tree: GATE2005-IT-52 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 207

Let G be a weighted undirected graph and e be an edge with maximum weight in G. Suppose there
is a minimum weight spanning tree in G containing the edge e. Which of the following statements is
always TRUE?

A. There exists a cutset in G having all edges of maximum weight.

B. There exists a cycle in G having all edges of maximum weight.
C. Edge e cannot be contained in a cycle.
D. All edges in G have the same weight.

gate2005-it algorithms spanning-tree normal

Answer

1.17.11 Spanning Tree: GATE2006-11 top [Link]

Consider a weighted complete graph G on the vertex set {v1 , v2 , . . . . . vn } such that the weight of
the edge (vi , vj ) is 2|i − j| . The weight of a minimum spanning tree of G is:

A. n−1
B. 2n − 2
( )
n
C.
2
2
D. n
gate2006 algorithms spanning-tree normal

Answer

1.17.12 Spanning Tree: GATE2006-47 top [Link]

Consider the following graph:

Which one of the following cannot be the sequence of edges added, in that order, to a minimum
spanning tree using Kruskal’s algorithm?

A. (a − b), (d − f), (b − f), (d − c), (d − e)

B. (a − b), (d − f), (d − c), (b − f), (d − e)
C. (d − f), (a − b), (d − c), (b − f), (d − e)
D. (d − f), (a − b), (b − f), (d − e), (d − c)

gate2006 algorithms graph-algorithms spanning-tree normal

Answer

1.17.13 Spanning Tree: GATE2007-49 top [Link]

Let w be the minimum weight among all edge weights in an undirected connected graph. Let e be a
specific edge of weight w. Which of the following is FALSE?

© Copyright GATE Overflow. All rights reserved.

208 1 Algorithms (323)

A. There is a minimum spanning tree containing e

B. If e is not in a minimum spanning tree T, then in the cycle formed by adding e to T, all edges
have the same weight.

C. Every minimum spanning tree has an edge of weight w

D. e is present in every minimum spanning tree
gate2007 algorithms spanning-tree normal

Answer

1.17.14 Spanning Tree: GATE2008-IT-45 top [Link]

For the undirected, weighted graph given below, which of the following sequences of edges
represents a correct execution of Prim's algorithm to construct a Minimum Span​ning Tree?

A. (a, b), (d, f), (f, c), (g, i), (d, a), (g, h), (c, e), (f, h)
B. (c, e), (c, f), (f, d), (d, a), (a, b), (g, h), (h, f), (g, i)
C. (d, f), (f, c), (d, a), (a, b), (c, e), (f, h), (g, h), (g, i)
D. (h, g), (g, i), (h, f), (f, c), (f, d), (d, a), (a, b), (c, e)

gate2008-it algorithms graph-algorithms spanning-tree normal

Answer

1.17.15 Spanning Tree: GATE2009-38 top [Link]

Consider the following graph:

Which one of the following is NOT the sequence of edges added to the minimum spanning tree using
Kruskal’s algorithm?

A. (b, e) (e, f) (a, c) (b, c) (f, g) (c, d)

B. (b, e) (e, f) (a, c) (f, g) (b, c) (c, d)
C. (b, e) (a, c) (e, f) (b, c) (f, g) (c, d)
D. (b, e) (e, f) (b, c) (a, c) (f, g) (c, d)

gate2009 algorithms spanning-tree normal

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 209

Answer

1.17.16 Spanning Tree: GATE2010-50 top [Link]

Consider a complete undirected graph with vertex set {0, 1, 2, 3, 4}. Entry Wij in the matrix W
below is the weight of the edge {i, j}

⎛0 1 8 1 4⎞
⎜1 0 9⎟
⎜ ⎟
12 4
W =⎜
⎜ 3⎟
⎟
⎜ ⎟
8 12 0 7
⎜1 4 7 0 2⎟
⎝4 9 3 2 0⎠
What is the minimum possible weight of a spanning tree T in this graph such that vertex 0 is a leaf
node in the tree T ?

A. 7
B. 8
C. 9
D. 10
gate2010 algorithms spanning-tree normal

Answer

1.17.17 Spanning Tree: GATE2010-51 top [Link]

Consider a complete undirected graph with vertex set {0, 1, 2, 3, 4}. Entry Wij in the matrix W
below is the weight of the edge {i, j}

⎛0 1 8 1 4⎞
⎜1 0 9⎟
⎜ ⎟
12 4
W =⎜
⎜ 3⎟
⎟
⎜ ⎟
8 12 0 7
⎜1 4 7 0 2⎟
⎝4 9 3 2 0⎠
What is the minimum possible weight of a path P from vertex 1 to vertex 2 in this graph such that
P contains at most 3 edges?
A. 7
B. 8
C. 9
D. 10
gate2010 normal algorithms spanning-tree

Answer

1.17.18 Spanning Tree: GATE2011-54 top [Link]

An undirected graph G(V , E) contains n (n > 2) nodes named v1 , v2 , … , vn . Two nodes vi , vj

are connected if and only if 0 <∣ i − j ∣≤ 2 . Each edge (vi , vj ) is assigned a weight i + j. A
sample graph with n = 4 is shown below.

© Copyright GATE Overflow. All rights reserved.

210 1 Algorithms (323)

What will be the cost of the minimum spanning tree (MST) of such a graph with n nodes?
1
A. 12 (11n2 − 5n)
B. n2 − n + 1
C. 6n − 11
D. 2n + 1

gate2011 algorithms graph-algorithms spanning-tree normal

Answer

1.17.19 Spanning Tree: GATE2011-55 top [Link]

An undirected graph G(V , E) contains n (n > 2) nodes named v1 , v2 , … , vn . Two nodes vi , vj

are connected if and only if 0 <∣ i − j ∣≤ 2 . Each edge (vi , vj ) is assigned a weight i + j. A
sample graph with n = 4 is shown below.

The length of the path from v5 to v6 in the MST of previous question with n = 10 is
A. 11
B. 25
C. 31
D. 41
gate2011 algorithms graph-algorithms spanning-tree normal

Answer

1.17.20 Spanning Tree: GATE2012-29 top [Link]

Let G be a weighted graph with edge weights greater than one and G′ be the graph constructed
by squaring the weights of edges in G. Let T and T ′ be the minimum spanning trees of G and
G′ , respectively, with total weights t and t′ . Which of the following statements is TRUE?
A. T ′ = T with total weight t′ = t2
B. T ′ = T with total weight t′ < t2
C. T ′ ≠ T but total weight t′ = t2
D. None of the above

gate2012 algorithms spanning-tree normal marks-to-all

Answer

1.17.21 Spanning Tree: GATE2014-2-52 top [Link]

The number of distinct minimum spanning trees for the weighted graph below is _____

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 211

gate2014-2 algorithms spanning-tree numerical-answers normal

Answer

1.17.22 Spanning Tree: GATE2015-1-43 top [Link]

The graph shown below has 8 edges with distinct integer edge weights. The minimum spanning tree
(MST) is of weight 36 and contains the edges: {(A, C), (B, C), (B, E), (E, F), (D, F)}. The
edge weights of only those edges which are in the MST are given in the figure shown below. The
minimum possible sum of weights of all 8 edges of this graph is_______________.

gate2015-1 algorithms spanning-tree normal numerical-answers

Answer

1.17.23 Spanning Tree: GATE2015-3-40 top [Link]

Let G be a connected undirected graph of 100 vertices and 300 edges. The weight of a minimum
spanning tree of G is 500. When the weight of each edge of G is increased by five, the weight of a
minimum spanning tree becomes ______.
gate2015-3 algorithms spanning-tree easy numerical-answers

Answer

1.17.24 Spanning Tree: GATE2016-1-14 top [Link]

Let G be a weighted connected undirected graph with distinct positive edge weights. If every edge
weight is increased by the same value, then which of the following statements is/are TRUE?

P : Minimum spanning tree of G does not change.

Q: Shortest path between any pair of vertices does not change.
A. P only
B. Q only
C. Neither P nor Q
D. Both P and Q

gate2016-1 algorithms spanning-tree normal

Answer

1.17.25 Spanning Tree: GATE2016-1-39 top [Link]

© Copyright GATE Overflow. All rights reserved.

212 1 Algorithms (323)

Let G be a complete undirected graph on 4 vertices, having 6 edges with weights being 1, 2, 3, 4, 5,
a n d 6. The maximum possible weight that a minimum weight spanning tree of G can have is
__________
gate2016-1 algorithms spanning-tree normal numerical-answers

Answer

1.17.26 Spanning Tree: GATE2016-1-40 top [Link]

G = (V , E) is an undirected simple graph in which each edge has a distinct weight, and e is a
particular edge of G. Which of the following statements about the minimum spanning trees
(MST s) of G is/are TRUE?
I. If e is the lightest edge of some cycle in G, then every MST of G includes e.
II. If e is the heaviest edge of some cycle in G, then every MST of G excludes e.
A. I only.
B. II only.
C. Both I and II.
D. Neither I nor II.

gate2016-1 algorithms spanning-tree normal

Answer

1.17.27 Spanning Tree: TIFR2011-B-35 top [Link]

Let G be a connected simple graph (no self-loops or parallel edges) on n ≥ 3 vertices, with distinct
edge weights. Let e1 , e2 , . . . , em be an ordering of the edges in decreasing order of weight. Which
of the following statements is FALSE?

A. The edge e1 has to be present in every maximum weight spanning tree.

B. Both e1 and e2 have to be present in every maximum weight spanning tree.
C. The edge em has to be present in every minimum weight spanning tree.
D. The edge em is never present in any maximum weight spanning tree.
E. G has a unique maximum weight spanning tree.
tifr2011 algorithms graph-algorithms spanning-tree

Answer

1.17.28 Spanning Tree: TIFR2013-B-17 top [Link]

In a connected weighted graph with n vertices, all the edges have distinct positive integer
weights. Then, the maximum number of minimum weight spanning trees in the graph is

a. 1
b. n
c. equal to number of edges in the graph.
d. equal to maximum weight of an edge of the graph.
e. nn−2

tifr2013 spanning-tree

Answer

1.17.29 Spanning Tree: TIFR2014-B-4 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 213

Consider the following undirected graph with some edge costs missing.

Suppose the wavy edges form a Minimum Cost Spanning Tree for G. Then, which of the following
inequalities NEED NOT hold?

a. cost(a, b) ≥ 6.
b. cost(b, e) ≥ 5.
c. cost(e, f) ≥ 5.
d. cost(a, d) ≥ 4.
e. cost(b, c) ≥ 4.

tifr2014 algorithms graph-algorithms spanning-tree

Answer

1.17.30 Spanning Tree: TIFR2014-B-5 top [Link]

Let G = (V , E) be an undirected connected simple (i.e., no parallel edges or self-loops) graph with
the weight function w : E → R on its edge set. Let w(e1 ) < w(e2 ) < ⋅ ⋅ ⋅ < w(em ) , where
E = {e1 , e2 , . . . , em } . Suppose T is a minimum spanning tree of G. Which of the following
statements is FALSE?

A. The tree T has to contain the edge e1 .

B. The tree T has to contain the edge e2 .
C. The minimum weight edge incident on each vertex has to be present in T .
D. T is the unique minimum spanning tree in G.
E. If we replace each edge weight wi = w(ei ) by its square w2i , then T must still be a minimum
spanning tree of this new instance.

tifr2014 algorithms spanning-tree

Answer

1.17.31 Spanning Tree: TIFR2015-B-2 top [Link]

Consider the following undirected connected graph G with weights on its edges as given in the figure
below. A minimum spanning tree is a spanning tree of least weight and a maximum spanning tree is
one with largest weight. A second best minimum spanning tree whose weight is the smallest among
all spanning trees that are not minimum spanning trees in G.

© Copyright GATE Overflow. All rights reserved.

214 1 Algorithms (323)

Which of the following statements is TRUE in the above graph? (Note that all the edge weights are
distinct in the above graph)

A. There is more than one minimum spanning tree and similarly, there is more than one maximum
spanning tree here.
B. There is a unique minimum spanning tree, however there is more than one maximum spanning
tree here.
C. There is more than one minimum spanning tree, however there is a unique maximum spanning
tree here.
D. There is more than one minimum spanning tree and similarly, there is more than one second-
best minimum spanning tree here.
E. There is unique minimum spanning tree, however there is more than one second-best minimum
spanning tree here.

tifr2015 spanning-tree algorithms graph-algorithms

Answer

Answers: Spanning Tree

1.17.1 Spanning Tree: GATE1991-03,vi top [Link]


Selected Answer

Answer: D, B, E.

When Union- Find algorithm is used to detect cycle while constructing the MST time complexity is
O(m log n) where m is the number of edges, and n is the number of vertices. Since n = O (m2 )
in a graph, options B and E are also correct as big-O specifies asymptotic upper bound only.

Reference: [Link]
tree-mst/

 13 votes -- Rajarshi Sarkar (34.1k points)

1.17.2 Spanning Tree: GATE1992-01,ix top [Link]


Selected Answer

If all edges are already sorted then this problem will reduced to union-find problem on a graph with
E edges and V vertices.
for each edge (u,v) in E

if(FIND-SET(u) != FIND-SET(v))
UNION(u,v)

FIND-SET(v) and UNION(u, v) runs in α(|V |)

where α(n) is inverse ackermann function i.e log∗ (n)
So, overall complexity becomes O(|E|. α(|V |))

 21 votes -- Vikrant Singh (13.5k points)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 215

1.17.3 Spanning Tree: GATE1995-22 top [Link]


Selected Answer

2 only.
{AB, BC, AE, BD} and {AB, BC, AE, CD}.

 21 votes -- Gate Keeda (19.6k points)

1.17.4 Spanning Tree: GATE1996-16 top [Link]


Selected Answer

(A).

(B).

 19 votes -- Anu (5.9k points)

1.17.5 Spanning Tree: GATE1997-9 top [Link]


Selected Answer

Consider n = 3, we have 9 vertices in the plane. The vertices are:

(1, 1), (1, 2), (1, 3), (2, 1), (2, 2), (2, 3), (3, 1), (3, 2), (3, 3)

© Copyright GATE Overflow. All rights reserved.

216 1 Algorithms (323)

So, the corresponding graph will be:

The cost of the edge between the vertices (2, 3) and (2, 2) = 1 and the cost of the edge
between the vertices (2, 2) and (2, 1) = 1.

Minimum spanning tree:

(One of the possibilities)

There is no problem with minimum spanning tree. We have n2 vertices so, the minimum spanning
tree contains n2 − 1 edges of cost 1. So, the cost of the minimum spanning tree is n2 − 1 .

Maximum spanning tree:

–
In this maximum spanning tree of n2 −1 edges n2 −2 edges is of cost √2 and 1 edge is of cost

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 217

1.
–
So, the answer is √2 (n2 − 2) + 1 .
This pattern continues for high values of n.
For n = 4, the maximum spanning tree is:
(one of possibilities)

For n = 4, 16 vertices are there and the maximum spanning tree requires 15 edges. Out of 15
– –
edges, 14 edges are of cost √2 and 1 edge is of cost 1. Thus satisfying the formula √2 (n2 − 2)
+1.
The cost of Minimum spanning tree is: n2 − 1 .
–
The cost of Maximum spanning tree is: (n2 − 1) √2 + 1 .

 16 votes -- balaeinstein (1.3k points)

1.17.6 Spanning Tree: GATE2000-2.18 top [Link]


Selected Answer

C the case should be written as "may or may not", to be true.

D will always be true as per the question saying that the graph has distinct weights.

 20 votes -- Gate Keeda (19.6k points)

1.17.7 Spanning Tree: GATE2001-15 top [Link]


Selected Answer

a) edges with weight: 2,3,4,7,9

© Copyright GATE Overflow. All rights reserved.

218 1 Algorithms (323)

b) no of distinct minimum spanning tree: 2 (2nd with the different edge of weight 4)
c) yes

d) yes

Edit:- Combining both answers into 1.

 14 votes -- jayendra (8.2k points)

1.17.8 Spanning Tree: GATE2003-68 top [Link]


Selected Answer

Apply Prim's algorithm, start from A as shown in figure below.

add all the weights in the given figure which would be equal to 31.

 8 votes -- Monanshi Jain (9.3k points)

1.17.9 Spanning Tree: GATE2005-6 top [Link]


Selected Answer

Graph G has multiple distinct MSTs, each of cost n − 1

From the given data given graph is a complete graph with all edge weights 1. A MST will contain
n − 1 edges . Hence weight of MST is n − 1.
The graph will have multiple MST. In fact all spanning trees of the given graph wll be MSTs also
since all edge weights are equal.

 17 votes -- Sankaranarayanan P.N (11.5k points)

1.17.10 Spanning Tree: GATE2005-IT-52 top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 219


Selected Answer

Option A is correct.
Questions says the MST of graph G contain an edge e which is a maximum weight edge in G.
Need to choose the answer which is always true to follow the above constraint.

Case 1:

Option B says tht if edge e ia in MST then for sure there is a cycle having all edges of
maximum weight. But it is not true always because when there is only n-1 edges( but no
cycle) in graph then also maximum edge has to be taken for MST.

Case 2:

Option C says otherwise. That if e is in MST then it cannot be in any cycle that is wrong as if there
is a cycle with all maximum edges then also e will be in MST

Option D says all edges should be of same weight same explanation if there are n − 1 distinct
edges( but no cycle) in G then have to take all edges including maximum weight edge.

And at last option A says if e is in MST then for sure there is a cut-set ( minimum edge set
whose removal disconnects the graph) in G having all edges of maximum weight. And it
is true.

Because then only we maximum weight edges has to be taken in MST.

For eg. If there are n − 1 edges (but no cycle) then if edge e is not taken in the MST then MST will
not be connected.

 12 votes -- khush tak (7.6k points)

1.17.11 Spanning Tree: GATE2006-11 top [Link]


Selected Answer

2(n − 1) the spanning tree will traverse adjacent edges since they contain the least weight.
 22 votes -- anshu (3.3k points)

1.17.12 Spanning Tree: GATE2006-47 top [Link]


Selected Answer

In kruskal's algo the edges are added in non decreasing order of their weight. But in Option D edge
d − e with weight 3 is added before edge d − c with weight [Link], option D is wrong option.

 14 votes -- Sankaranarayanan P.N (11.5k points)

1.17.13 Spanning Tree: GATE2007-49 top [Link]


Selected Answer

D is the false statement.

© Copyright GATE Overflow. All rights reserved.

220 1 Algorithms (323)

A minimum spanning tree must have the edge with the smallest weight (In Kruskal's algorithm we
start from the smallest weight edge). So, C is TRUE.

If e is not part of a minimum spanning tree, then all edges which are part of a cycle with e, must
have weight ≤ e, as otherwise we can interchange that edge with e and get another minimum
spanning tree of lower weight. So, B and A are also TRUE.

 16 votes -- Arjun Suresh (350k points)

1.17.14 Spanning Tree: GATE2008-IT-45 top [Link]


Selected Answer

Prim's algorithm starts with from any vertex and expands the MST by adding one vertex in each
step which is close to the Intermediate MST(made till previous step).

Therefore, correct answer would be (C).

(A): (d, f) is chosen but neither d nor f vertices are part of the previous MST(MST made till
previous step).

(B): (g, h) is chosen but neither g or h vertices are part of the previous MST(MST made till
previous step).

(D): (f, c) is chosen but at that point (f, d) is close to the intermediate MST.

 16 votes -- suraj (5.6k points)

1.17.15 Spanning Tree: GATE2009-38 top [Link]


Selected Answer

In Option D b-c with weight, 4 is added before a-c with weight 3 is added. In Kruskal's algorithm,
edges should be added in non-decreasing order of weight.

So, Option D may be correct.

 14 votes -- Sankaranarayanan P.N (11.5k points)

1.17.16 Spanning Tree: GATE2010-50 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 221

Answer is (D) 10. The edges of the spanning tree are: 0 − 1, 1 − 3, 3 − 4, 4 − 2 . Total Weight
= 10

 26 votes -- Ashis Kumar Sahoo (865 points)

1.17.17 Spanning Tree: GATE2010-51 top [Link]


Selected Answer

Answer is (B) 8. The possible path is: 1 − 0, 0 − 4, 4 − 2.

 20 votes -- Ashis Kumar Sahoo (865 points)

1.17.18 Spanning Tree: GATE2011-54 top [Link]


Selected Answer

Q 54. Answer is B.

We observe a pattern in the weight of MST being formed

For n=3 (1 + 2 + 3) + (1)
For n=4 (1 + 2 + 3 + 4) + (1 + 2)
© Copyright GATE Overflow. All rights reserved.
222 1 Algorithms (323)

For n=4 (1 + 2 + 3 + 4) + (1 + 2)
For n=5 (1 + 2 + 3 + 4 + 5) + (1 + 2 + 3)
These can be obtained by drawing graphs for these graphs.
2
∴ Total weight of MST is ∑ni=1 i + ∑n−2
i=1 i = n − n + 1

 21 votes -- Sona Praneeth Akula (4.2k points)

1.17.19 Spanning Tree: GATE2011-55 top [Link]


Selected Answer

Above is the graph....

Below is the MST.

Length of the path from v5 to v6 = 8 + 4 + 3 + 6 + 10 = 31 (Answer)

 26 votes -- Ahwan Mishra (10.5k points)

1.17.20 Spanning Tree: GATE2012-29 top [Link]


Selected Answer

When the edge weights are squared the minimum spanning tree won't change.

t′ < t2 , because sum of squares is always less than the square of the sums except for a single
element case.

Hence, B is the general answer and A is also true for a single edge graph. Hence, in GATE 2012,
marks were given to all.

 37 votes -- gatecse (18k points)

1.17.21 Spanning Tree: GATE2014-2-52 top [Link]


Selected Answer

6 is the answer.
2 × 3 = 6 possibilities

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 223

 31 votes -- Arjun Suresh (350k points)

1.17.22 Spanning Tree: GATE2015-1-43 top [Link]


Selected Answer

Consider the cycle ABC. AC and AB are part of minimum spanning tree. So, AB should be
greater than max(AC, BC) (greater and not equal as edge weights are given to be distinct), as
otherwise we could add AB to the minimum spanning tree and removed the greater of AC, BC
and we could have got another minimum spanning tree. So, AB > 9.

Similarly, for the cycle DEF, ED > 6 .

And for the cycle BCDE, CD > 15.
So, minimum possible sum of these will be 10 + 7 + 16 = 33. Adding the weight of spanning tree,
we get the total sum of edge weights

= 33 + 36 = 69
 51 votes -- Arjun Suresh (350k points)

1.17.23 Spanning Tree: GATE2015-3-40 top [Link]


Selected Answer

First find no of edges in mst.

st has n − 1 edges where n is no of vertices. 100 − 1 = 99 edges
Each 99 edges in mst increases by 5 so weight in mst increased 99 ∗ 5 = 495
Now total weight of mst = 500 + 495 = 995

 38 votes -- Anoop Sonkar (5k points)

1.17.24 Spanning Tree: GATE2016-1-14 top [Link]


Selected Answer

Statement P is true.

For statement Q consider a simple graph with 3 nodes.

A BC
A0 1 100
B1 0 2
C 1002 0
Shortest path from A to C is A-B-C =1+2 =3

© Copyright GATE Overflow. All rights reserved.

224 1 Algorithms (323)

Now if the value of each edge is increased by 100,

A B C
A 0 101200
B 1010 102
C 2001020

The shortest path from A to C is A-C = 200, (A-B-C = 101 + 102 = 203)
Hence, option A is correct.

 51 votes -- ryan sequeira (3.4k points)

1.17.25 Spanning Tree: GATE2016-1-39 top [Link]


Selected Answer

Graph G can be like this:

 45 votes -- shaiklam09 (237 points)

1.17.26 Spanning Tree: GATE2016-1-40 top [Link]


Selected Answer

Statement 1:- False by [Cut Property of MST]

See counter example :- Here in below Graph G in (cycle 3, 4, 5) 3 is the lightest edge and still it is
not included in MST.

G 1−2−3
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 225

Statement 2: True by[ Cycle property of MST] : (in above Graph G 1 − 2 − 3 is a cycle and 3 is
the heaviest edge) If heaviest edge is in cycle then we will always exclude that bcz Cycle is their
means we can have other choice of low cost edges.

So, option B is answer.

Must visit Links:

[Link]

 52 votes -- Rajesh Pradhan (22.6k points)

1. For an edge to be INCLUDED in the MST,it must satisfy the CUT PROPERTY which states that

Assume that all edge costs are distinct. Let S be any subset of nodes that is neither empty nor
equal to all of V, and let edge e = (v, w) be the minimum cost edge with one end in S and the
other in V − S. Then every minimum spanning tree contains the edge e.

Now,in the given problem, e is the lightest edge of some cycle in G,but it may not be the lightest
edge connecting S and V - S. See the below image for clarification.

Here, e is the lightest edge of cycle [Link] are three edges e,e',e" connecting S and V - S.

e is obviously minimal weighted than e' since e' is part of the cycle C,but e MAY NOT be minimal
weighted than e" .

So,every MST of G MAY NOT include e.

2. For an edge to be EXCLUDED from the MST, it must satisfy the CYCLE PROPERTY which states
that

Assume that all edge costs are distinct. Let C be any cycle in G, and let edge e = (v, w) be the
most expensive edge belonging to C. Then e does not belong to the minimum spanning tree of G.

Now,in the given problem, e is the heaviest edge of some cycle C .

So,every MST of G MUST exclude e.

ANSWER is B.

NOTE: For proofs of the above two properties, one can refer to the source of this answer:
Algorithm Design by Jon Kleinberg and Eva Tardos

 29 votes -- Tanaya Pradhan (625 points)

1.17.27 Spanning Tree: TIFR2011-B-35 top [Link]

© Copyright GATE Overflow. All rights reserved.

226 1 Algorithms (323)


Selected Answer

a) & c) are trivially true. Edge with max value e1 must be present in Maximum spanning tree &
same with minimum.

e) This is true, because all edge weights are distinct. maximum spanning tree is unique.

b) e1 & e2 must be present in Maximum spanning tree. I'll prove it using Kruskal Algorithm.

We will first insert weight with biggest value, e1 . Then we insert e2 (second highest) . 2 edges do
not create cycle. Then we can go on from there inserting edges according to edge weights. As they
have just asked for top 2 edges, using Kruskal Algo we can say that top 2 edges must be in
Maximum spanning tree.

d) This is false. There are chances that this em weight edge is cut edge(Bridge) Then it must be
inserted to from any spanning tree.

D is answer.

We can not say the same for Top 3 as they can create cycle & They we can not take a3 to make
spanning tree.

Kruskal Algo Reference: [Link]

spanning-tree

 7 votes -- Akash Kanase (42.5k points)

1.17.28 Spanning Tree: TIFR2013-B-17 top [Link]


Selected Answer

There will be unique min weight spanning tree since all weights are distinct.
Option is A.

 10 votes -- Umang Raman (15.7k points)

1.17.29 Spanning Tree: TIFR2014-B-4 top [Link]


Selected Answer

Now check this diagram, this is forest obtained from above given graph using Kruskal's algorithm
for MST.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 227

So, according to the question edge d − e has weight 5 and it is included in the formation of MST.
Now if edges b − e and e − f has weight greater than 5 than it is not a problem for our MST
because still we will get the given tree as Kruskal's algorithm takes the smallest weighted edge
without forming a cycle.

Cost of edge b − c ≥ 4 may also lead us to the same tree as above though Kruskal's algorithm will
have choice between c − f and b − c .

Now if the edge weight of a − d becomes 4, it is guaranteed that Kruskal's algorithm will not select
edge d − e because its edge cost is 5, and hence the tree structure will change. But there can be
the case where edge weight is greater than 4 and we still get the same tree (happens when
a − d ≥ 5 ). Because in the question they asked to point out an unnecessary condition this case is
not the answer as we need a − d ≥ 5 which implies a − d ≥ 4.

Now notice option A. Put a − b = 5. The given MST would not change. So, this condition is not
always necessary and hence is the answer.

Therefore, option A is the answer.

 14 votes -- Riya Roy(Arayana) (7.2k points)

1.17.30 Spanning Tree: TIFR2014-B-5 top [Link]


Selected Answer

Answer is E. The catch here is edge-weights belongs to real number. Therefore, edge weight can
be negative. In that case the minimum spanning tree may be different.

e1= −5
e2 = 1
e3 = 2

T is:

New edge weights:

e1 = 25
e2 = 1
e3 = 4

EDIT:

© Copyright GATE Overflow. All rights reserved.

228 1 Algorithms (323)

(Here every edge weight is distinct, therefore MST is unique.)

Option A is True. If we apply Kruskal's algorithm, it will choose e1

Option B is True. If we apply Kruskal's algorithm then it will also choose e2 , and 2 edges can not
form a cycle. (e3 is not guaranteed in MST, as it may form a cycle)

Option C is also true. If we apply Prims algorithm also on any vertex (say u), it chooses minimum

weight edge incident on vertex u.

Option D is true. Because every edge weight is distinct.

 23 votes -- Riya Roy(Arayana) (7.2k points)

1.17.31 Spanning Tree: TIFR2015-B-2 top [Link]


Selected Answer

In the graph we have all edge weights are distinct so we will get unique minimum and maximum
spanning tree.
Each Cycle must exclude maximum weight edge in minimum spanning tree.
Here, we have two cycle of 3 edges, a − d − e and c − g − k.
For second best minimum spanning tree, exclude a − e edge and include d − e edge

Other way for second best minimum spanning tree: exclude c − g edge and include g − k edge.
So, e should be the answer.

 12 votes -- papesh (25.7k points)

1.18 Time Complexity(34) top

1.18.1 Time Complexity: CMI2013-A-10 top [Link]

The below question is based on following program:

procedure mystery (A : array [1..100] of int)

int i,j,position,tmp;
begin
for j := 1 to 100 do
position := j;
for i := j to 100 do
if (A[i] > A[position]) then
position := i;
endfor
tmp := A[j];
A[j] := A[position];
A[position] := tmp;
endfor
end

The number of times the test A[i] > A[position] is executed is:
A. 100
B. 5050
C. 10000
A
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 229

D. Depends on contents of A
cmi2013 algorithms time-complexity

Answer

1.18.2 Time Complexity: CMI2015-A-08 top [Link]

How many times is the comparison i >= n performed in the following program?
int i=85, n=5;
main() {
while (i >= n) {
i=i-1;
n=n+1;
}
}

A. 40
B. 41
C. 42
D. 43
cmi2015 algorithms time-complexity

Answer

1.18.3 Time Complexity: CMI2017-B-8 top [Link]

Consider the following function that takes as input a sequence A of integers with n elements,
A[A1], [A2] , … , [An]and an integer k and returns an integer value. The function length(S)
returns the length of sequence S. Comments start with //.

function mystery(A, k){

n = length(A);
if (k > n) return A[n];
v = A[1];
AL = [ A[j] : 1 <= j <= n, A[j] < v ]; // AL has elements < v in A
Av = [ A[j] : 1 <= j <= n, A[j] == v ]; // Av has elements = v in A
AR = [ A[j] : 1 <= j <= n, A[j] > v ]; // AR has elements > v in A
if (length(AL) >= k) return mystery(AL,k);
if (length(AL) + length(Av) >= k) return v;
return mystery(AR, k - (length(AL) + length(Av)));
}

A. Explain what the function computes.

B. What is the worst-case complexity of this algorithm in terms of the length of the input sequence
A?
C. Give an example of a worst-case input for this algorithm.

cmi2017 algorithms time-complexity descriptive

Answer

1.18.4 Time Complexity: GATE1989-2-iii top [Link]

Match the pairs in the following questions:

(A) O
(p) Heapsort
(log n)

© Copyright GATE Overflow. All rights reserved.

230 1 Algorithms (323)

(B) O
(n (q) Depth-first search
)
(C) O
(r) Binary search
(n log n)
(D) O
(s) Selection of the kth smallest element in a set of n elements.
(n2 )
gate1989 match-the-following algorithms time-complexity

Answer

1.18.5 Time Complexity: GATE1993-8.7 top [Link]

∑ O(n), where O(n) stands for order n is:

1≤k≤n

A. O(n)
B. O(n2 )
C. O(n3 )
D. O(3n2 )
E. O(1.5n2 )

gate1993 algorithms time-complexity easy

Answer

1.18.6 Time Complexity: GATE1999-1.13 top [Link]

Suppose we want to arrange the n numbers stored in any array such that all negative values occur
before all positive ones. Minimum number of exchanges required in the worst case is

A. n − 1
B. n
C. n + 1
D. None of the above

gate1999 algorithms time-complexity normal

Answer

1.18.7 Time Complexity: GATE1999-1.16 top [Link]

If n is a power of 2, then the minimum number of multiplications needed to compute an is

A. log2 n
−
B. √n
C. n−1
D. n
gate1999 algorithms time-complexity normal

Answer

1.18.8 Time Complexity: GATE1999-11a top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 231

Consider the following algorithms. Assume, procedure A and procedure B take O(1) and O(1/n)
unit of time respectively. Derive the time complexity of the algorithm in O-notation.

algorithm what (n)

begin
if n = 1 then call A
else
begin
what (n-1);
call B(n)
end
end.

gate1999 algorithms time-complexity normal

Answer

1.18.9 Time Complexity: GATE2000-1.15 top [Link]

L e t S be a sorted array of n integers. Let T (n) denote the time taken for the most efficient
algorithm to determined if there are two elements with sum less than 1000 in S. Which of the
following statement is true?

A. T (n) is O(1)
B. n ≤ T (n) ≤ n log2 n
C. n log2 n ≤ T (n) < n2
D. T (n) = ( n2 )

gate2000 easy algorithms time-complexity

Answer

1.18.10 Time Complexity: GATE2003-66 top [Link]

The cube root of a natural number n is defined as the largest natural number m such that
(m3 ≤ n) . The complexity of computing the cube root of n (n is represented by binary notation)
is

A. O(n) but not O(n0.5 )

B. O(n0.5 ) but not O((log n)k ) for any constant k > 0
C. O((log n)k ) for some constant k > 0, but not O((log log n)m ) for any constant m > 0
D. O((log log n)k ) for some constant k > 0.5, but not O((log log n)0.5 )

gate2003 algorithms time-complexity normal

Answer

1.18.11 Time Complexity: GATE2004-39 top [Link]

Two matrices M1 and M2 are to be stored in arrays A and B respectively. Each array can be stored
either in row-major or column-major order in contiguous memory locations. The time complexity of
an algorithm to compute M1 × M2 will be

A. best if A is in row-major, and B is in column-major order

B. best if both are in row-major order

C. best if both are in column-major order

D. independent of the storage scheme

© Copyright GATE Overflow. All rights reserved.

232 1 Algorithms (323)

gate2004 algorithms time-complexity easy

Answer

1.18.12 Time Complexity: GATE2004-82 top [Link]

Let A[1, … n] be an array storing a bit (1 or 0) at each location, and f(m) is a function whose time
complexity is Θ(m). Consider the following program fragment written in a C like language:

counter = 0;
for (i=1; i<=n; i++)
{ if a[i] == 1) counter++;
else {f (counter); counter = 0;}
}

The complexity of this program fragment is

A. Ω(n2 )
B. Ω(n log n) and O(n2 )
C. Θ(n)
D. o(n)

gate2004 algorithms time-complexity normal

Answer

1.18.13 Time Complexity: GATE2006-15 top [Link]

Consider the following C-program fragment in which i, j and n are integer variables.
for( i = n, j = 0; i > 0; i /= 2, j +=i );

Let val(j) denote the value stored in the variable j after termination of the for loop. Which one of
the following is true?

A. val(j) = Θ(log n)
B. val(j) = Θ(√−n)
C. val(j) = Θ(n)
D. val(j) = Θ(n log n)

gate2006 algorithms normal time-complexity

Answer

1.18.14 Time Complexity: GATE2007-15,ISRO2016-26 top

[Link]

Consider the following segment of C-code:

int j, n;
j = 1;
while (j <= n)
j = j * 2;

The number of comparisons made in the execution of the loop for any n > 0 is:
A. ⌈log2 n⌉ + 1

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 233

B. n
C. ⌈log2 n⌉
D. ⌊log2 n⌋ + 1

gate2007 algorithms time-complexity normal isro2016

Answer

1.18.15 Time Complexity: GATE2007-44 top [Link]

In the following C function, let n ≥ m.

int gcd(n,m) {
if (n%m == 0) return m;
n = n%m;
return gcd(m,n);
}

How many recursive calls are made by this function?

A. Θ(log2 n)
B. Ω(n)
C. Θ(log2 log2 n)
D. Θ(√−n)

gate2007 algorithms time-complexity normal

Answer

1.18.16 Time Complexity: GATE2007-45 top [Link]

What is the time complexity of the following recursive function?

int DoSomething (int n) {
if (n <= 2)
return 1;
else
return (DoSomething (floor (sqrt(n))) + n);
}

A. Θ(n2 )
B. Θ(n log2 n)
C. Θ(log2 n)
D. Θ(log2 log2 n)

gate2007 algorithms time-complexity normal

Answer

1.18.17 Time Complexity: GATE2007-50 top [Link]

An array of n numbers is given, where n is an even number. The maximum as well as the minimum

© Copyright GATE Overflow. All rights reserved.

234 1 Algorithms (323)

of these n numbers needs to be determined. Which of the following is TRUE about the number of
comparisons needed?

A. At least 2n − c comparisons, for some constant c are needed.

B. At most 1.5n − 2 comparisons are needed.
C. At least n log2 n comparisons are needed
D. None of the above

gate2007 algorithms time-complexity easy

Answer

1.18.18 Time Complexity: GATE2007-51 top [Link]

Consider the following C program segment:

int IsPrime (n)

{
int i, n;
for (i=2; i<=sqrt(n);i++)
if(n%i == 0)
{printf("Not Prime \n"); return 0;}
return 1;
}

Let T (n) denote number of times the for loop is executed by the program on input n. Which of the
following is TRUE?

A. T (n) = O(√−n ) and T (n) = Ω(√−n)

B. T (n) = O(√−n ) and T (n) = Ω(1)
C. T (n) = O(n) and T (n) = Ω(√− n)
D. None of the above

gate2007 algorithms time-complexity normal

Answer

1.18.19 Time Complexity: GATE2007-IT-17 top [Link]

Exponentiation is a heavily used operation in public key cryptography. Which of the following options
is the tightest upper bound on the number of multiplications required to compute
bn mod m, 0 ≤ b, n ≤ m ?
A. O(log n)
B. O(√−n)

C. O( logn n )

D. O(n)

gate2007-it algorithms time-complexity normal

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 235

1.18.20 Time Complexity: GATE2007-IT-81 top [Link]

Let P1 , P2 , … , Pn be n points in the xy-plane such that no three of them are collinear. For every
pair of points Pi and Pj , let Lij be the line passing through them. Let Lab be the line with the
steepest gradient among all n(n − 1)/2 lines.

The time complexity of the best algorithm for finding Pa and Pb is

A. Θ (n)
B. Θ (n log n)
C. Θ (n log2 n)
D. Θ (n2 )

gate2007-it algorithms time-complexity normal

Answer

1.18.21 Time Complexity: GATE2008-40 top [Link]

The minimum number of comparisons required to determine if an integer appears more than n
2
times in a sorted array of n integers is

A. Θ(n)
B. Θ(log n)
C. Θ(log∗ n)
D. Θ(1)

gate2008 normal algorithms time-complexity

Answer

1.18.22 Time Complexity: GATE2008-47 top [Link]

We have a binary heap on n elements and wish to insert n more elements (not necessarily one after
another) into this heap. The total time required for this is

A. Θ(log n)
B. Θ(n)
C. Θ(n log n)
D. Θ(n2 )

gate2008 algorithms time-complexity normal

Answer

1.18.23 Time Complexity: GATE2008-74 top [Link]

Consider the following C functions:

int f1 (int n)
{

© Copyright GATE Overflow. All rights reserved.

236 1 Algorithms (323)

if(n == 0 || n == 1)
return n;
else
return (2 * f1(n-1) + 3 * f1(n-2));
}
int f2(int n)
{
int i;
int X[N], Y[N], Z[N];
X[0] = Y[0] = Z[0] = 0;
X[1] = 1; Y[1] = 2; Z[1] = 3;
for(i = 2; i <= n; i++){
X[i] = Y[i-1] + Z[i-2];
Y[i] = 2 * X[i];
Z[i] = 3 * X[i];
}
return X[n];
}

The running time of f1(n) and f2(n) are

A. Θ(n) and Θ(n)
B. Θ(2n ) and Θ(n)
C. Θ(n) and Θ(2n )
D. Θ(2n ) and Θ(2n )

gate2008 algorithms time-complexity normal

Answer

1.18.24 Time Complexity: GATE2008-75 top [Link]

Consider the following C functions:

int f1 (int n)
{
if(n == 0 || n == 1)
return n;
else
return (2 * f1(n-1) + 3 * f1(n-2));
}
int f2(int n)
{
int i;
int X[N], Y[N], Z[N];
X[0] = Y[0] = Z[0] = 0;
X[1] = 1; Y[1] = 2; Z[1] = 3;
for(i = 2; i <= n; i++){
X[i] = Y[i-1] + Z[i-2];
Y[i] = 2 * X[i];
Z[i] = 3 * X[i];
}
return X[n];
}

f1(8) and f2(8) return the values

A. 1661 and 1640
B. 59 and 59
C. 1640 and 1640
D. 1640 and 1661

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 237

gate2008 normal algorithms time-complexity

Answer

1.18.25 Time Complexity: GATE2010-12 top [Link]

Two alternative packages A and B are available for processing a database having 10k records.
Package A requires 0.0001n2 time units and package B requires 10n log10 n time units to process
n records. What is the smallest value of k for which package B will be preferred over A?
A. 12
B. 10
C. 6
D. 5
gate2010 algorithms time-complexity easy

Answer

1.18.26 Time Complexity: GATE2014-1-42 top [Link]

Consider the following pseudo code. What is the total number of multiplications to be performed?

D = 2
for i = 1 to n do
for j = i to n do
for k = j + 1 to n do
D = D * 3

A. Half of the product of the 3 consecutive integers.

B. One-third of the product of the 3 consecutive integers.
C. One-sixth of the product of the 3 consecutive integers.
D. None of the above.

gate2014-1 algorithms time-complexity normal

Answer

1.18.27 Time Complexity: GATE2015-1-40 top [Link]

1 1
An algorithm performs (log N) 2 find operations , N insert operations, (log N) 2 delete operations,
1
a n d (log N) decrease-key operations on a set of data items with keys drawn from a linearly
2

ordered set . For a delete operation, a pointer is provided to the record that must be deleted . For
the decrease-key operation, a pointer is provided to the record that has its key decreased. Which
one of the following data structures is the most suited for the algorithm to use, if the goal is to
achieve the best total asymptotic complexity considering all the operations?

A. Unsorted array
B. Min - heap
C. Sorted array
D. Sorted doubly linked list

gate2015-1 algorithms data-structure normal time-complexity

Answer

1.18.28 Time Complexity: GATE2015-2-22 top [Link]

© Copyright GATE Overflow. All rights reserved.

238 1 Algorithms (323)

An unordered list contains n distinct elements. The number of comparisons to find an element in this
list that is neither maximum nor minimum is

A. Θ(n log n)
B. Θ(n)
C. Θ(log n)
D. Θ(1)

gate2015-2 algorithms time-complexity easy

Answer

1.18.29 Time Complexity: GATE2017-2-03 top [Link]

Match the algorithms with their time complexities:

Algorithms Time Complexity

P. Tower of Hanoi with n disks i. Θ(n2 )
ii. Θ(n log
Q. Binary Search given n numbers n sorted numbers
n)
R. Heap sort given n numbers at the worst case iii. Θ(2n )
S. Addition of two n × n matrices iv. Θ(log n)

A. P→ (iii) Q →(iv) r →(i) S →(ii)

B. P→ (iv) Q →(iii) r →(i) S→(ii)
C. P→ (iii) Q →(iv) r →(ii) S→(i)
D. P→ (iv) Q →(iii) r →(ii) S→(i)

gate2017-2 algorithms time-complexity

Answer

1.18.30 Time Complexity: GATE2017-2-38 top [Link]

Consider the following C function

int fun(int n) {
int I, j;
for(i=1; i<=n; i++) {
for (j=1; j<n; j+=i) {
printf("%d %d", I, j);
}
}
}

Time complexity of fun in terms of Θ notation is

A. Θ(n√− n)
2
B. Θ(n )
C. Θ(n log n)
D. Θ(n2 log n)

gate2017-2 algorithms time-complexity

Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 239

1.18.31 Time Complexity: TIFR2013-B-12 top [Link]

It takes O(n) time to find the median in a list of n elements, which are not necessarily in sorted
order while it takes only O(1) time to find the median in a list of n sorted elements. How much time
does it take to find the median of 2n elements which are given as two lists of n sorted elements
each?

A. O(1)
B. O (log n) but not O(1)
C. O(√− n ) but not O (log n)
D. O(n) but not O(√− n)
E. O (n log n) but not O(n)

tifr2013 algorithms time-complexity

Answer

1.18.32 Time Complexity: TIFR2013-B-18 top [Link]

Let S be a set of numbers. For x ∈ S , the rank of x is the number of elements in S that are
less than or equal to x. The procedure Select(S, r) takes a set S of numbers and a rank
r (1 ≤ r ≤ |S|) and returns the element in S of rank r. The procedure MultiSelect(S, R) takes a
set of numbers S and a list of ranks R = {r1 < r2 < … < rk } , and returns the list
{x1 < x2 <. . . < xk } of elements of S, such that the rank of xi is ri . Suppose there is an
implementation for Select(S, r) that uses at most ‘‘ constant ·|S| " binary comparisons between
elements of S. The minimum number of comparisons needed to implement MultiSelect(S, R) is

a. constant · |S| log |S|

b. constant · |S|
c. constant · |S||R|
d. constant · |R| log |S|
e. constant · |S|(1 + log |R|)

tifr2013 algorithms time-complexity

Answer

1.18.33 Time Complexity: TIFR2014-B-7 top [Link]

Which of the following statements is TRUE for all sufficiently large n?

A. (log n)log log n < 2√log n < n1/4

B. 2√log n < n1/4 < (log n)log log n

C. n1/4 < (log n)log log n < 2√log n

D. (log n)log log n < n1/4 < 2√log n

E. 2√log n < (log n)log log n < n1/4

tifr2014 algorithms time-complexity

Answer

© Copyright GATE Overflow. All rights reserved.

240 1 Algorithms (323)

1.18.34 Time Complexity: TIFR2015-B-3 top [Link]

Consider the following code fragment in the C programming language when run on a non-negative
integer n.

int f (int n)
{
if (n==0 || n==1)
return 1;
else
return f (n - 1) + f(n - 2);
}

Assuming a typical implementation of the language, what is the running time of this algorithm and
how does it compare to the optimal running time for this problem?

A. This algorithm runs in polynomial time in n but the optimal running time is exponential in n.
B. This algorithm runs in exponential time in n and the optimal running time is exponential in n.
C. This algorithm runs in exponential time in n but the optimal running time is polynomial in n.
D. This algorithm runs in polynomial time in n and the optimal running time is polynomial in n.
E. The algorithm does not terminate.

tifr2015 time-complexity

Answer

Answers: Time Complexity

1.18.1 Time Complexity: CMI2013-A-10 top [Link]


Selected Answer

Answer: 5050 (100 + 99 + 98+. . . . . +1 = (100 ∗ 101)/2)

 7 votes -- Dhananjay Kumar Sharma (25.7k points)

1.18.2 Time Complexity: CMI2015-A-08 top [Link]


Selected Answer

Correct answer is (C)

42
It will start comparison from (85, 5), (84, 6), (83, 7), … , (45, 45), (44, 46)
Hence, total number of comparison will be (85 − 44) + 1 = 41 + 1 = 42

 16 votes -- Muktinath Vishwakarma (35.4k points)

1.18.3 Time Complexity: CMI2017-B-8 top [Link]


Selected Answer

a. Function computes kth smallest element (if k<number of elements in the array, else it returns
max element).

O(k n)
© Copyright GATE Overflow. All rights reserved.
1 Algorithms (323) 241

b. Time complexity = O(k*n), where n is the number of elements in the array.

c. Worst case input: Array consists of a distinct increasing sequence of numbers and from this
sequence, if we need to find nth smallest (i.e. max element) then it will take O(n*n) time.

 1 votes -- Digvijay (54.9k points)

1.18.4 Time Complexity: GATE1989-2-iii top [Link]


Selected Answer

(A)
O (r) Binary Search
(log n)
(B)
O
(n (s) Selection of the kth smallest element in a set of n elements.(Worst case)
)
(C)
O (p) Heap sort
(n log n)
(q) Depth-first search
(D) (It will beO(n if the graph is given in the form of adjacency list but if the
O
+ m)
(n2 graph is in the form of adjacency matrix then the complexity is O(n , as we
)
× n)
have to traverse through the whole row until we find an edge.)

PS: kth smallest element can be found in O(n) time using partition algorithm.
T (n) = T (n/2) + n (For finding kth smallest element)

By Masters Theorem it will be done in O(n).

 11 votes -- Prajwal Bhat (11.3k points)

1.18.5 Time Complexity: GATE1993-8.7 top [Link]


Selected Answer

N added itself N times. So it is N 2 . Even if you consider as sum of

This is
O(1) + O(2)+. . O(n − 1) + O(N). it will add up to N 2
So answer is

A) O(N) this is false.

B,C,D,E ) All of this are true. We have N2 here, so all options apart from A are correct.

In fact B = D = E this three options are same. and N 3 is always upper bound of N2. So O(N 3 ) is

© Copyright GATE Overflow. All rights reserved.

242 1 Algorithms (323)

also true.

-------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------

PS: ∑(K = 1 to n) O(K) is never equal to n functions. It is always equal to one single
function.

It can be written as c.1 + c.2 + c.3 and so on which results in O(N 2 ) (source Cormen)

 24 votes -- Akash Kanase (42.5k points)

1.18.6 Time Complexity: GATE1999-1.13 top [Link]


Selected Answer

Answer is (D) None of these.

We just require n/2 swaps in the worst case. The algorithm is as given below:
Find positive number from left side and negative number from right side and do exchange. Since,
at least one of them must be less than or equal to n/2, there cannot be more than n/2
exchanges. An implementation is given below:

[Link]

 36 votes -- Arjun Suresh (350k points)

1.18.7 Time Complexity: GATE1999-1.16 top [Link]


Selected Answer

a. log n
n
an = (a2 ) 2 .

One multiplication and recurrence on n

2 . So, we get the recurrence relation for the number of
multiplications as

T (n) = T (n/2) + 1.
This gives T (n) = log2 n
For n = 8, we can do

b =a×a
b =b×b
b =b×b and we get b = a8
 24 votes -- Arjun Suresh (350k points)

1.18.8 Time Complexity: GATE1999-11a top [Link]

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 243


Selected Answer

The recurrence relation for time complexity is

1
T (n) = T (n − 1) + n + c (O(1/n) replaced with 1/n and so our answer will also be in O only
and not Θ)

1
T (n) = T (n − 2) + n−1 + n1 + 2c
= T (1) + 12 + 13 + ⋯ + n1 + (n − 1)c
= A(1) + 12 + 13 ⋯ + n1 + nc
= 1 + 12 + 13 … + n1 + nc
= log n + nc
(Sum of the first n terms in harmonic series is Θ(log n))
So, our time complexity will be O(n).
 24 votes -- Arjun Suresh (350k points)

1.18.9 Time Complexity: GATE2000-1.15 top [Link]


Selected Answer

Answer: Option A. Because array is always sorted just check the 1st two elements.

 29 votes -- anshu (3.3k points)

1.18.10 Time Complexity: GATE2003-66 top [Link]


Selected Answer

We can simply do a binary search in the array of natural numbers from 1..n and check if the cube
of the number matches n (i.e., check if a[i] ∗ a[i] ∗ a[i] == n). This check takes O(log n) time
and in the worst case we need to do the search O(log n) times. So, in this way we can find the
cube root in O(log2 n). So, options (A) and (B) are wrong.
Now, a number is represented in binary using log n bit. Since each bit is important in finding the
cube root, any cube root finding algorithm must examine each bit at least once. This ensures that
complexity of cube root finding algorithm cannot be lower than log n. (It must be Ω (log n)). So,
(D) is also false and (C) is the correct answer.

 52 votes -- gatecse (18k points)

1.18.11 Time Complexity: GATE2004-39 top [Link]


Selected Answer

D is correct

Here time complexity is asked, for each access of array element it will be constant,

So the time complexity will not depend upon storage. If at all program execution time is asked

© Copyright GATE Overflow. All rights reserved.

244 1 Algorithms (323)

a is true

 34 votes -- Anurag Semwal (8k points)

1.18.12 Time Complexity: GATE2004-82 top [Link]


Selected Answer

The key part in the code is "counter = 0" in the else part as we can see below.
Lets take the best case. This happens when a[i] = 1 for all i, and then the loop executes with time
complexity Θ(1) for each iteration and hence overall time complexity of Θ(n) and we can say
time complexity of the code fragment is Ω(n) and hence options A and B are false.

Now, consider the worst case. This happens when a[i] = 0 or when else part is executed. Here,
the time complexity of each iteration will be Θ(counter) and after each else, counter is reset to 0.
Let k iterations go to the else part during the worst case. Then the worst case time complexity will
be Θ(x1 ) + Θ(x2 ) + ⋯ + Θ(xk ) + Θ(n − k) , where xi is the value of the counter when,
A[i] = 0 and f(counter) is called. But due to counter = 0 after each call to f(), we have,
x1 + x2 + ⋯ + xk = n . So,
Θ(x1 ) + Θ(x2 ) + ⋯ + Θ(xk ) + Θ(n − k) = Θ(n) + Θ(n − k) = Θ(n) .
Since the time complexity is Ω(n) and Θ(n) we can say it is Θ(n) - Option (C). (Option D is false
because the small o needs the growth rate to be STRICTLY lower and not equal to or lower as the
case for big 0)

If counter = 0 was not there in else part, then time complexity would be Ω(n) and O(n2 ) as in
worst case we can have equal number of 0's and 1's in array a giving time complexity
Θ(1) + Θ(2) + ⋯ + Θ(n/2) + Θ(n/2) would give O(n2 ).

 30 votes -- Arjun Suresh (350k points)

1.18.13 Time Complexity: GATE2006-15 top [Link]


Selected Answer

Answer will be Θ(n)

j = n/2 + n/4 + n/8 + … + 1

= n [1/21 + 1/22 + 1/23 + … + 1/2lg n ]

1−rn
(Sum of first n terms of GP is [a. 1−r ] , where a is the first term, r is the common ratio < 1, and
n is the number of terms)
1−(1/2)lg n
= n [1/2 1−1/2
]

= n [ n−1
n ]

= n − 1 = Θ(n)
 26 votes -- rahulkr (805 points)

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 245

1.18.14 Time Complexity: GATE2007-15,ISRO2016-26 top

[Link]

n no. of comparisons Ceil(log2 n) + 1

2(j = 1,
1 1
2)
3(j = 1,
2 2
2, 4)
3(j = 1,
3 2
2, 4)
4(j = 1,
4 3
2, 4, 8)
4(j = 1,
5 4
2, 4, 8)
may be we have to count those comparisons which results in the execution of loop.

Answer should be Ceil (log2 n) + 1

EDIT: but answer could be: floor(log2 n) + 2

 40 votes -- Vikrant Singh (13.5k points)

1.18.15 Time Complexity: GATE2007-44 top [Link]


Selected Answer

Worst case will arise when both n and m are consecutive Fibonacci numbers.
gcd(Fn , Fn−1 ) = gcd(Fn−1 , Fn−2 ) = ⋯ = gcd(F1 , F0 ) = 1
and nth Fibonacci number is 1.618n , where 1.618 is the Golden ratio.
So, to find gcd(n, m), number of recursive calls will be Θ(log n).

 32 votes -- Vikrant Singh (13.5k points)

1.18.16 Time Complexity: GATE2007-45 top [Link]


Selected Answer

We are asked the time complexity which will be the number of recursive calls in the function as in
each call we perform a constant no. of operations and a recursive call. The recurrence relation for
this is (considering constant time "c" as 1)

T (n) = T (√−
n) + 1
= T (n1/4 ) + 2
= T (n1/8 ) + 3
Going like this we will eventually reach T (3) or T (2). For asymptotic case this doesn't matter and
T (2) T (1)
© Copyright GATE Overflow. All rights reserved.
246 1 Algorithms (323)

we can assume we reach T (2) and in next step reach T (1). So, all we want to know is how many
steps it takes to reach T (1) which will be 1+no. of steps to reach T (2).

( 1
)
From the recurrence relation we know that T (2) happens when n 2k = 2.
Taking log and equating,
1
log n = 1
2k
⟹ 2k = log n
⟹ k = log log n.
So, T (1) happens in log log n + 1 calls, but for asymptotic complexity we can write as
Θ (log log n)

Alternatively,

Substituting values

T (1) = 1
T (2) = 1
T (3) = T (1) + 1 = 2
⋮
T (8) = T (2) + 1 = 2
T (9) = T (3) + 1 = 3
⋮
2 2
T (((22 ) ) ) = T ((22 ) ) + 1
2

= T (22 ) + 2
= T (2) + 3 = 1 + 3 = 4,
log log n = 3 as n = 256.

⎛ ⎞
2
2 2
T ((((2 ) ) ) )
2 2
⎝ ⎠
= 6,

log log n = 5 as n = 65536 × 65536 = 232

T (2(2 ) ) = T (2512 ) + 1
10

= T (2256 ) + 2
= T (2128 ) + 3
= T (264 ) + 4
= T (232 ) + 5
= T (216 ) + 6
= T (28 ) + 7
= T (24 ) + 8
= T (22 ) + 9
= T (2) + 10 = 11,
log log n = 10
So, answer is D.

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 247

[Link]
n-complexity

 36 votes -- Arjun Suresh (350k points)

1.18.17 Time Complexity: GATE2007-50 top [Link]


Selected Answer

An easier way to find it is by using Tournament Method Technique -

1. To find the smallest element in the array will take n − 1 comparisions.

2. To find the largest element -
a. After the first round of Tournament , there will be exactly n/2 numbers that will loose the
round.
b. So the biggest looser (the largest number) should be among these 50 [Link] find the
largest number will take n/2 − 1 .

Total Comparisons = (n − 1) + (n/2 − 1) = 1.5n − 2 .

 41 votes -- Harsh181996 (4k points)

1.18.18 Time Complexity: GATE2007-51 top [Link]


Selected Answer

Answer is option B.

Worst Case :

Best Case : When is an even number body of loop is executed only time (due to "
" inside if) which is irrespective of . ∴

 24 votes -- Gate Keeda (19.6k points)

1.18.19 Time Complexity: GATE2007-IT-17 top [Link]


Selected Answer

Answer is (A)

We need to divide
recursively and compute like following:

. In this, we need to calculate only once.

© Copyright GATE Overflow. All rights reserved.

248 1 Algorithms (323)

Recurrence relation:

 34 votes -- Sandeep_Uniyal (7.5k points)

1.18.20 Time Complexity: GATE2007-IT-81 top [Link]


Selected Answer

Answer: B

Gradient

For gradient to be maximum should be minimum. So, sort the points (in time)
according to coordinate and find the minimum difference between them (in time).

Best complexity: which leads to B.

[Link]

[Link]

 17 votes -- Rajarshi Sarkar (34.1k points)

1.18.21 Time Complexity: GATE2008-40 top [Link]


Selected Answer

Answer is option B.

whenever there exists an element which is present in the array : more than times, then
definitely it will be present at the middle index position; in addition to that it will also be present at
anyone of the neighbourhood indices namely and
No matter how we push that stream of More than times of elements of same value around the
Sorted Array, it is bound to be present at the middle index atleast anyone of its neighbourhood

once we got the element which should have occurred more that times we count its total
occurrences in time.

 35 votes -- Amar Vashishth (30.5k points)

1.18.22 Time Complexity: GATE2008-47 top [Link]


Selected Answer

An insert operation on a binary heap takes time, but an alternative approach we can use.
which requires us to insert elements in heap without any computation i.e. in constant time. after
which we can apply Heapify operation(this operation creates heap in linear time) on the array of
those element and Hence obtain a Heap in time.

Here "not necessarily one after another" should mean that we can insert elements at once and

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 249

not necesaairly have to wait for first insert to be completed before doing second.

 31 votes -- Amar Vashishth (30.5k points)

1.18.23 Time Complexity: GATE2008-74 top [Link]


Selected Answer

Q. = option B
Q. = option C

Time complexity of is given by

, (multiplication by and won't affect complexity as it is a
constant time operation)

The solution to this (fibonacci series) is given by Golden

ratio. [Link] which is . (Using theta in question must be
a mistake)

Time complexity of is as here all recursive calls are avoided by saving the results in an
array (dynamic programming).

So, answer to is (B).

. Both and are calculating the same function. So,

 31 votes -- Arjun Suresh (350k points)

1.18.24 Time Complexity: GATE2008-75 top [Link]


Selected Answer

Both and are calculating the same function in recursive and iterative fashion respectively.

So. lets solve the recurrence relation.

Its characteristic equation will be

Now, we have two roots. So the equation will be

© Copyright GATE Overflow. All rights reserved.

250 1 Algorithms (323)

Now from the function

and

So,

and

After putting the values in the equation will become

Putting it will become

Or we can do it manually

 1 votes -- Puja Mishra (10.6k points)

1.18.25 Time Complexity: GATE2010-12 top [Link]


Selected Answer

Trying the values, doesn't satisfy this but satisfies.

 47 votes -- Arjun Suresh (350k points)

1.18.26 Time Complexity: GATE2014-1-42 top [Link]


Selected Answer

Total number of multiplications

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 251

Therefore, correct answer would be (C).

 54 votes -- suraj (5.6k points)

1.18.27 Time Complexity: GATE2015-1-40 top [Link]


Selected Answer

find decrease-
delete
insert
key

Unsorted
Array

Min-heap

Sorted Array

© Copyright GATE Overflow. All rights reserved.

252 1 Algorithms (323)

Sorted doubly
linked-list

So, Unsorted array is the answer.

The operations given can be performed in any order. So, for Min-heap we cannot do the usual
BuildHeap method.

Delete in unsorted array is as we can just swap the deleted element with the last element in
the array and delete the last element.

For sorted-doubly linked-list we cannot do binary search as this would require another array to
maintain the pointers to the nodes.

 61 votes -- Arjun Suresh (350k points)

1.18.28 Time Complexity: GATE2015-2-22 top [Link]


Selected Answer

Ans: , because all elements are distinct, select any three numbers and output nd largest
from them.

 60 votes -- Vikrant Singh (13.5k points)

1.18.29 Time Complexity: GATE2017-2-03 top [Link]


Selected Answer

According to the recurrrence relation

Tower of hanoi we get it is

Now, heap sort worst case

Binary Search given n numbers n sorted numbers

Addition of two nxn matrices

So, C is correct answer here.

 22 votes -- Aboveallplayer (18.4k points)

1.18.30 Time Complexity: GATE2017-2-38 top [Link]


Selected Answer

© Copyright GATE Overflow. All rights reserved.

1 Algorithms (323) 253

Inner for loop is dependent on , so for each we have to check no of times inner loop operating..

It ll be something like

 26 votes -- 2018 (6.7k points)

1.18.31 Time Complexity: TIFR2013-B-12 top [Link]


Selected Answer

1. Calculate the medians and of the input arrays

and respectively.
2. If and both are equal.
return (or )
3. If is greater than , then median is present in one
of the below two subarrays.
a) From first element of to to
b) From to last element of to
4. If is greater than , then median is present in one
of the below two subarrays.
a) From to last element of to
b) From first element of to to
5. Repeat the above process until size of both the subarrays
becomes .
6. If size of the two arrays is then
the median.
Median
Time complexity

[Link]

 16 votes -- Umang Raman (15.7k points)

1.18.32 Time Complexity: TIFR2013-B-18 top [Link]

a) should be the answer

Reason: Sort the set. Initialize a counter to 1. Keep a parallel pointer to the sorted list of needed
ranks. Keep incrementing the counter as you traverse the sorted set. As soon as counter matches
the first needed rank, add this to the required output. Increment the parallel pointer.

 2 votes -- Akshay Arora (2.5k points)

© Copyright GATE Overflow. All rights reserved.

254 1 Algorithms (323)

1.18.33 Time Complexity: TIFR2014-B-7 top [Link]


Selected Answer

Let us take for each function.

1.
2.
3.

Here, If we consider as a term (which is common in all 3), first 1 is a log function, second
one is sqrt function and third one is linear function of . Order of growth of these functions are
well known and is the slowest growing followed by sqrt and then linear. So, option A is the
correct answer here.

PS: After taking is we arrive at functions distinguished by some constant terms only, then we
can not conclude the order of grpwth of the original functions using the function. Examples are
.

 10 votes -- srestha (87k points)

1.18.34 Time Complexity: TIFR2015-B-3 top [Link]


Selected Answer

Answer: C.

It is fibanacci series generation. it takes exponential time if we won't use dynamic programming.

If we use dynamic programming then it takes

 12 votes -- pramod (3.3k points)

© Copyright GATE Overflow. All rights reserved.

390 3 Programming & DS: DS (208)

The tester now tests the program on all input strings of length five consisting of characters ‘ ’, ‘ ’, ‘
’, ‘ ’ and ‘ ’ with duplicates allowed. If the tester carries out this testing with the four test cases
given above, how many test cases will be able to capture the flaw?

A. Only one
B. Only two
C. Only three
D. All four

gate2013 data-structure arrays normal

Answer

3.2.10 Arrays: GATE2013-51 top [Link]

The procedure given below is required to find and replace certain characters inside an input
character string supplied in array . The characters to be replaced are supplied in array , while
their respective replacement characters are supplied in array . Array has a fixed length of
five characters, while arrays and contain three characters each. However, the procedure
is flawed.

void find_and_replace (char *A, char *oldc, char *newc) {

for (int i=0; i<5; i++)
for (int j=0; j<3; j++)
if (A[i] == oldc[j])
A[i] = newc[j];
}

The procedure is tested with the following four test cases.

1.
2.
3.
4.

If array is made to hold the string “ ”, which of the above four test cases will be successful
in exposing the flaw in this procedure?

A. None
B. only
C. and only
D. only

gate2013 data-structure arrays normal

Answer

3.2.11 Arrays: GATE2014-3-42 top [Link]

Consider the C function given below. Assume that the array contains elements, sorted
in ascending order.

int ProcessArray(int *listA, int x, int n)

{
int i, j, k;
i = 0; j = n-1;
do {
k = (i+j)/2;
if (x <= listA[k]) j = k-1;

© Copyright GATE Overflow. All rights reserved.