Programming with C++

COMP2011: Function II — Recursion

Cecia Chan
Brian Mak
Dimitris Papadopoulos
Pedro Sander
Charles Zhang

Department of Computer Science & Engineering

The Hong Kong University of Science and Technology
Hong Kong SAR, China

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.1
 Part I

Think Recursively!
“God Help Those Who Help
Themselves.”

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.2
Example: Tower of Hanoi Game

It consists of 3 pegs, and a stack of discs of different sizes.

It starts with all discs stacked up on one peg with smaller
discs sitting on top of bigger discs.
The goal is to move the entire stack of discs to another peg,
making use of the remaining peg.
Rules:
only one disc may be moved at a time
no disc may be placed on top of a smaller disc

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.3
Example: Factorial Function

Definition of the factorial function you learn in high school:

n! = n × (n − 1) × (n − 2) × · · · 2 × 1

Recursive Definition of Factorial Function

0! = 1
n! = n × (n − 1)! if n > 0

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.4
Example: Polyline

+ =

A polyline with n line segments is equivalent to adding another line

segment to a polyline with n − 1 line segments.

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.5
Example: Fibonacci Numbers

0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, . . .

Fibonacci (1202) investigated how fast rabbits could breed:

A newly-born pair of rabbits, one male, one female, are put in
a field.
Rabbits mate at the age of one month so that at the end of
its 2nd month, a female can produce another pair of rabbits.
Suppose that our rabbits never die.
Suppose the female always produces one new pair (one male,
one female) every month from the 2nd month on.
How many pairs will there be in one year?

Question : What is special with the above numbers?

Answer : Except for the first 2 numbers, each number is the
sum of the last 2 numbers in the sequence.
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.6
Thinking Recursively

In some problems, it may be natural to define the problem in

terms of the problem itself!
Recursion is useful for problems that can be represented by a
simpler version of the same problem.
1 To solve the Tower of Hanoi game that uses n discs, one first
solves the tower of hanoi game that uses n − 1 discs.
2 To find the value of n!, first find the value of (n − 1)! and then
multiply the result with n.
3 To draw a polyline of n segments, first draw a polyline of n − 1
segments, and then add the last line segment.
4 To find the nth Fibonacci number, first find the (n − 1)th and
(n − 2)th Fibonacci numbers, and then add them up.

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.7
 Part II

Programming Recursion in C++

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.8
Implement Recursion by Recursive Function

In programming, recursion means that a function calls itself!

Although it looks strange in the beginning, solving a
programming task by recursion renders the program
easier to write
easier to read (understand)
shorter (in codes).

Implement a Recursive Solution

1 Decompose the problem into sub-problems — which are
smaller examples of the same problem — plus some additional
work that “glues” the solutions of the sub-problems together.
2 The smallest sub-problem has a non-recursive solution.

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.9
Example: Factorial Function
int factorial(int n) /* factorial.cpp */
{
if (n < 0) // Error checking
return -1;
else if (n == 0) // Base case; ending case too!
return 1;
else // Recursive case
return n * factorial(n-1);
}

Or, equivalently,

int factorial(int n) /* factorial2.cpp */

{
if (n < 0) // Error checking
return -1;
if (n == 0) // Base case; ending case too!
return 1;
return n * factorial(n-1); // Recursive case
}
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.10
How the Recursive Factorial Function Works?
factorial(3) :
3<0 false
3 == 0 false
3 * factorial(2)
factorial(2) :
2<0 false
2 == 0 false
2 * factorial(1)

factorial(1) :
1<0 false
1 == 0 false
1 * factorial(0)

factorial(0) :
0<0 false
0 == 0 true
return 1

return 1*1 = 1

return 2*1 = 2
return 3*2 = 6
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.11
Factorial Function: Recursive vs. Non-Recursive
int factorial(int n) /* factorial.cpp */
{
if (n < 0) // Error checking
return -1;
else if (n == 0) // Base case; ending case too!
return 1;
else // Recursive case
return n * factorial(n-1);
}

int factorial(int n) /* non-recursive-factorial.cpp */

{
int result = 1; // Default value for n = 0 or 1
if (n < 0) // Error checking
return -1;

for (int j = 2; j <= n; j++) // When n >= 2

result *= j;
return result;
}
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.12
Infinite Loop!

When we work with loops, we have to be careful that the

ending condition will be met eventually.
Otherwise, we will get infinite loop!

for (int j = 1; j != 10; j += 2)

{
// Ending condition never met!
cout << j << endl;
...
}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.13
Infinite Recursion!
Just like infinite loops, we have to be careful that a recursion
will eventually end up with a non-recursive base case.
Otherwise, we will get infinite recursion!

int factorial(int n)
{
// Forget the base case, which is the ending case too!
return n * factorial(n-1);
}

int factorial(int n)
{
// Forget checking if n < 0
if (n == 0)
return 1;

// Infinite recursion for negative n

return n * factorial(n-1);
}
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.14
Example: Fibonacci Function as a Recursion

int fibonacci(int n) /* File: fibonacci.cpp */

{
if (n == 0) // Base case #1
return 0;

if (n == 1) // Base case #2
return 1;

return fibonacci(n-1) + fibonacci(n-2);

}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.15
Inefficiency of Recursive Fibonacci Function

f(4)

f(3) f(2)

f(2) f(1) f(1) f(0)

f(1) f(0)

repeated computations

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.16
Example: Non-Recursive Fibonacci Function
int fibonacci(int n) /* non-recursive-fibonacci.cpp */
{
int fn; // keep track of f(n)
int fn_1 = 1; // keep track of f(n-1)
int fn_2 = 0; // keep track of f(n-2)

if (n == 0) return 0; // Base case #1

if (n == 1) return 1; // Base case #2

for (int j = 2; j <= n; j++)

{
fn = fn_1 + fn_2; // f(n) = f(n-1) + f(n-2)

// Prepare for the calculation of the next fibonacci number

fn_2 = fn_1; // f(n-2) = f(n-1)
fn_1 = fn; // f(n-1) = f(n)
}

return fn;
}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.17
Example: Counting Zeros in an Integer

Example: for the integer 120809, there are 2 zeros.

Basic idea:
Break down the number into quotient and remainder.
Count the number of zeros in quotient and remainder.

int num_zeros(int n) /* File: num-zeros.cpp */

{
if (n == 0) // Base case #1
return 1;

else if (n < 10 && n > 0) // Base case #2

return 0;

else
return num_zeros(n/10) + num_zeros(n%10);
}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.18
Example: Factoring

Goal: find how many times factor m appears in the integer n.

Example: if n = 48 and m = 4, since 48 = 4 × 4 × 3, the
answer is 2.
Basic idea:
Divide n by m until the remainder is non-zero.
Increment the count by 1 for every successful division.

int num_factors(int n, int m) /* File: factor.cpp */

{
if (n % m != 0) // Base case
return 0;
else
return 1 + num_factors(n/m, m);
}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.19
Recursive Solution of Tower of Hanoi

A B C
Step 1

Step 2

Step 3

Step 4

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.20
Example: Recursive Solution of Tower of Hanoi

#include <iostream> /* File: toh.cpp */

using namespace std;

void tower_of_hanoi(int num_discs, char pegA, char pegB, char pegC)

{
if (num_discs == 0) // Base case
return;

tower_of_hanoi(num_discs-1, pegA, pegC, pegB);

cout << "move disc " << num_discs

<< " from peg " << pegA << " to peg " << pegC << endl;

tower_of_hanoi(num_discs-1, pegB, pegA, pegC);

}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.21
Binary Search

first mid last

2nd call

3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

1st call

first mid last

binary search for the value 9

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.22
Example: Recursive Solution of Binary Search
const int NOT_FOUND = -1; /* File: bsearch.cpp */
int bsearch(const int data[ ], // sorted in ascending order
int first, // lower bound index
int last, // upper bound index
int value) // value to search
{
if (last < first) // Base case #1
return NOT_FOUND;

int mid = (first + last)/2;

if (data[mid] == value)
return mid; // Base case #2

else if (data[mid] > value) // Search the lower half

return bsearch(data, first, mid-1, value);

else // Search the upper half

return bsearch(data, mid+1, last, value);
}

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.23
Example: Non-Recursive Solution of Binary Search
const int NOT_FOUND = -1; /* File: non-recursive-bsearch.cpp */
int bsearch(const int data[ ], // sorted in ascending order
int size, // number of data in the array
int value) // value to search
{
int first = 0;
int last = size - 1;

while (first <= last)

{
int mid = (first + last)/2;
if (data[mid] == value)
return mid; // Value found!
else if (data[mid] > value)
last = mid - 1; // Set up for searching the lower half
else
first = mid + 1; // Set up for searching the upper half
}

return NOT_FOUND;
}
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.24
Recursion is Natural

Many natural phenomena are recursion: a smaller part of

oneself is embedded in itself!

(b) infinite mirror

(a) tree (c) fractal
images

As a result, it is usually easier for a programmer to write a

solution using recursion ⇒ greater productivity.

{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.25
Disadvantages of Recursion
The greater programming productivity is achieved at the
expenses of the more computing resources. To run recursion,
it usually requires
more memory
more computational time

The reason is that whenever a function is called, the computer

has to memorize its current state, and passes control from the
caller to the callee.
sets up a new data structure (you may think of it as a scratch
paper for rough work) called activation record which contains
information such as
where the caller stops
what actual parameters are passed to the callee
new local variables created by the callee function
the return value of the function at the end
removes the activation record of the callee when it finishes.
passes control back to the caller.
{ kccecia, mak, dipapado, psander, charlesz }@cse.ust.hk COMP2011 (Fall 2022) p.26