MATH2081 | MATH2328: Math for Computing

1
Lecturers: Dr Nguyen Hieu Thao & Dr Jeff Nijsse
Email: [email protected]

Tutorial Solutions: Induction, Recursion

and Recurrence
(2024C, Week 6)1

1. Prove by induction that for every positive integer n, it holds that

n(n + 1)(n + 2)(3n + 5)

1 · 22 + 2 · 32 + · · · + n · (n + 1)2 = . (1)
12

Proof.

• Base case. For n = 1, (1) becomes

1(1 + 1)(1 + 2)(3 + 5)

1 · 22 = ,
12
which is true.
• Induction hypothesis. Assume that (1) is true for all positive integers n ≤ k (k ≥ 1).
• Induction step. It suffices to prove that (1) is true for n = k + 1. That is,

(k + 1)(k + 2)(k + 3)(3k + 8)

1 · 22 + 2 · 32 + · · · + k · (k + 1)2 + (k + 1) · (k + 2)2 = .
12
Indeed,

k(k + 1)(k + 2)(3k + 5)

LHS = + (k + 1) · (k + 2)2 (induction hypothesis for n = k)
12
(k + 1)(k + 2)(k(3k + 5) + 12(k + 2))
=
12
(k + 1)(k + 2)(3k 2 + 17k + 24)
=
12
(k + 1)(k + 2)(k + 3)(3k + 8)
= = RHS.
12
That is, equation (1) is also true for n = k + 1, and hence the proof is complete. 

1 Some of the content of this document is taken from the book [1].
2

2. Write a recursive function finding the minimum of a finite sequence of numbers.

Solution. Recursive function:

min(x1 , x2 , · · · , xn ) :
if n = 1 : return x1
else if n = 2 : if x1 ≤ x2 return x1 , else return x2
else : return min (min(x1 , x2 , . . . , xn−1 ), xn ) .

3. Write a recursive function reversing a finite sequence of numbers.

Solution. Recursive function:

reverse(x1 , x2 , · · · , xn ) :
if n = 1 : return x1
else : return xn , reverse (x1 , x2 , . . . , xn−1 ) .

4. The sequence g1 , g2 , . . . is defined by the recurrence relation

gn = gn−1 + gn−2 + 1, n ≥ 3,

and initial conditions g1 = 1, g2 = 3. By using mathematical induction show that

gn = 2fn+1 − 1, n ≥ 1, (2)

where f1 , f2 , . . . is the Fibonacci sequence.

Proof.
• Base case. For n = 1, (2) becomes g1 = 1 = 2f2 − 1 = 2 − 1 which is true because
g1 = 1 and f2 = 1.

For n = 2, (2) becomes g2 = 3 = 2f3 − 1 = 4 − 1 which is true because g2 = 3 and

f3 = 2.
• Induction hypothesis. Assume that (2) is true for all positive integers n ≤ k (k ≥ 2).
• Induction step. It suffices to prove that (2) is true for n = k + 1. That is,

gk+1 = 2fk+2 − 1.

Indeed, by definition of g, the induction hypothesis and definition of f , we have

gk+1 = gk + gk−1 + 1 (definition of g)

= 2fk+1 − 1 + 2fk − 1 + 1 (induction hypothesis for n = k and n = k − 1)
= 2(fk+1 + fk ) − 1
= 2fk+2 − 1 (definition of f ).

That is, equation (2) is true for n = k + 1, and hence the proof is complete. 
 3

5. The Lucas sequence L1 , L2 , . . . is defined by the recurrence relation

Ln = Ln−1 + Ln−2 , n ≥ 3,
and the initial conditions L1 = 1, L2 = 3.

(a) Find the values of L3 , L4 , and L5 .

Solution. L3 = 4, L4 = 7, L5 = 11.
(b) Show that
Ln+2 = fn+1 + fn+3 , n ≥ 1, (3)
where f1 , f2 , . . . is the Fibonacci sequence.
Proof.
• Base case. For n = 1, (3) becomes L3 = 4 = f2 + f4 = 1 + 3, which is true because
L3 = 4, f2 = 1 and f4 = 3.
For n = 2, (3) becomes L4 = 7 = f3 + f5 = 2 + 5, which is true because L4 = 7,
f3 = 2 and f5 = 5.
• Induction hypothesis. Assume that (3) is true for all positive integers n ≤ k (k ≥ 2).

• Induction step. It suffices to prove that (3) is true for n = k + 1. That is,
Lk+3 = fk+2 + fk+4 .
Indeed, by definition of L, the induction hypothesis and definition of f , we have
Lk+3 = Lk+2 + Lk+1 (definition of L)
= fk+1 + fk+3 + fk + fk+2 (induction hypothesis for n = k and n = k − 1)
= fk+1 + fk + fk+3 + fk+2
= fk+2 + fk+4 (definition of f ).
That is, equation (3) is true for n = k + 1, and hence the proof is complete. 

6. Solve the recurrence relation with the given initial conditions.

(a) 2an = 7an−1 − 3an−2 with a0 = a1 = 1.
Solution. Solve the equation:
7 3 1
t2 − t + = 0 ⇔ t1 = 3 ∨ t2 = .
2 2 2
Then
an = b · 3n + d · 2−n
where b and d are constants.

To find b and d, we use the initial conditions:

( (
1
a0 = b + d = 1 b= 5
⇔
a1 = 3b + 12 d = 1 d= 4
5

Hence,
1 n 4 −n
an = ·4 + ·2 .
5 5
4

√ √ √
(b) an = an−1 + 2 an−2 with a0 = a1 = 1.
√
Solution. Set bn = an . We get b0 = b1 = 1 and bn = bn−1 + 2bn−2 .
Solve the equation:
t2 − t − 2 = 0 ⇔ t1 = −1 ∨ t2 = 2.
Then
bn = b · (−1)n + d · 2n
where b and d are constants.
To find b and d, we use the initial conditions:
( (
1
b0 = b + d = 1 b= 3
⇔ 2
b1 = −b + 2d = 1 d= 3

Then,
1 2
bn = · (−1)n + · 2n .
3 3
Hence,
 2
1 2
an = b2n = · (−1)n + · 2n .
3 3

7. Find a recursive formula of the function f (n) = 2n + 5.

Solution. f (1) = 7, f (n) = f (n − 1) + 2 (∀n ≥ 2) .

8. The recursive formula of f (n) is a linear function f (n) = a · f (n − 1) + b, where a and b

are constants. Find a and b given that f (1) = 2, f (2) = 7 and f (3) = 17.
Solution. Solve the system of equations:
( ( (
f (2) = af (1) + b 7 = 2a + b a=2
⇔ ⇔
f (3) = af (2) + b 17 = 7a + b b=3

Hence, the recursive version of f is: f (1) = 2 and f (n) = 2f (n − 1) + 3 (∀n ≥ 2).

9. Consider the following recursive function f : N+ → N+ :

f (1) = 1, f (n) = 10 · f (n − 1) + 1 (∀n ≥ 2).

(a) Compute f (2), f (3) and f (4).

• f (2) = 10f (1) + 1 = 10 + 1 = 11

• f (3) = 10f (2) + 1 = 102 + 10 + 1 = 111
• f (4) = 10f (3) + 1 = 103 + 102 + 10 + 1 = 1111
 5

(b) Prove by induction that f takes the following explicit form:

10n − 1
f (n) = , for all n ≥ 1. (4)
9
Proof.
• Base case. For n = 1, (4) becomes
10 − 1
f (1) = = 1,
9
which is true.

• Induction hypothesis. Suppose that (4) is true for all positive integers n ≤ k
(k ≥ 1). In particular,
10k − 1
f (k) = .
9

• Induction step. It suffices to prove that (4) is true for n = k + 1. That is,

10k+1 − 1
f (k + 1) = .
9
Indeed,
10k − 1 10k+1 − 1
f (k + 1) = 10f (k) + 1 = 10 +1= .
9 9

That is, equation (4) is true for n = k + 1, and hence the proof is complete. 

References
1. Johnsonbaugh, R.: Discrete Mathematics - Eighth Edition. Pearson Education, New York
(2018).