MATH 300 Problems with solutions

F. Baudier (Texas A&M University)

October 20, 2023

Contents

1 Logical connectives and equivalences, Boolean Calculus 2

2 Quantifiers 8

3 Proofs 12

4 Applications of the Principle of Mathematical Induction 18

5 Applications of the Principle of Strong Mathematical Induction 25

6 Sequences defined by a recurrence relation 26

7 Set Theory 27
7.1 Subsets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
7.2 Complements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.3 Arbitrary unions and intersections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.4 More problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

8 Functions 35
8.1 Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8.2 Injectivity, surjectivity, bijectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
8.3 Composition and injectivity/surjectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

9 Injectivity, surjectivity, and one-sided invertibility 37

10 Functions and sets 38

1
11 Supplementary problems 39

1 Logical connectives and equivalences, Boolean Calculus

Problem 1.1. Show that P ∨ P is logically equivalent to P.

Solution of Problem 1.1. We write the truth table for P ∨ P.

P P P∨P
T T T
F F F

Problem 1.2. Show that P ∧ P is logically equivalent to P.

Solution of Problem 1.2. We write the truth table for P ∧ P.

P P P∧P
T T T
F F F

Problem 1.3. Are the statement forms (P ∧ Q) ∧ R and P ∧ (Q ∧ R) logically equivalent?

Solution of Problem 1.3. We write a truth table for (P ∧ Q) ∧ R and P ∧ (Q ∧ R):

P Q R P∧Q Q∧R (P ∧ Q) ∧ R P ∧ (Q ∧ R)
T T T T T T T
T T F T F F F
T F T F F F F
T F F F F F F
F T T F T F F
F T F F F F F
F F T F F F F
F F F F F F F
Since the columns corresponding to (P ∧ Q) ∧ R and P ∧ (Q ∧ R) are identical, the two statement forms are
equivalent.

2
Problem 1.4. Are the statement forms (P ∨ Q) ∨ R and P ∨ (Q ∨ R) logically equivalent?

Solution of Problem 1.4. We write a truth table for (P ∨ Q) ∨ R and P ∨ (Q ∨ R):

P Q R P∨Q Q∨R (P ∨ Q) ∨ R P ∨ (Q ∨ R)
T T T T T T T
T T F T T T T
T F T T T T T
T F F T F T T
F T T T T T T
F T F T T T T
F F T F T T T
F F F F F F F
Since the columns corresponding to (P ∨ Q) ∨ R and P ∨ (Q ∨ R) are identical, the two statement forms are
equivalent.

Problem 1.5. Is the statement form (P ∧ Q) ∨ ((¬P) ∧ ¬Q) a tautology, a contradiction, or neither?

Solution of Problem 1.5. We write a truth table for (P ∧ Q) ∨ ((¬P) ∧ ¬Q):

P Q P∧Q (¬P) ∧ ¬Q (P ∧ Q) ∨ ((¬P) ∧ ¬Q)

T T T F T
T F F F F
F T F F F
F F F T T
Since (P∧Q)∨((¬P)∧¬Q) is neither always true nor always false, the statement form is neither a tautology
nor a contradiction.

Problem 1.6. Are the statement forms (P ∨ Q) ∧ R and P ∨ (Q ∧ R) logically equivalent?

Solution of Problem 1.6. We write a truth table for (P ∨ Q) ∧ R and P ∨ (Q ∧ R):

3
 P Q R P∨Q Q∧R (P ∨ Q) ∧ R P ∨ (Q ∧ R)
T T T T T T T
T T F T F F T
T F T T F T T
T F F T F F T
F T T T T T T
F T F T F F F
F F T F F F F
F F F F F F F
Since the columns corresponding to (P ∨ Q) ∧ R and P ∨ (Q ∧ R) differ (in the second and fourth rows), the
two statements are not logically equivalent.

Problem 1.7. Show that P ∨ (Q ∧ R) is logically equivalent to (P ∨ Q) ∧ (P ∨ R).

Solution of Problem 1.7. We write a truth table for P ∨ (Q ∧ R) and (P ∨ Q) ∧ (P ∨ R):

P Q R P∨Q P∨R (P ∨ Q) ∧ (P ∨ R) P ∨ (Q ∧ R)
T T T T T T T
T T F T T T T
T F T T T T T
T F F T T T T
F T T T T T T
F T F T F F F
F F T F T F F
F F F F F F F

Since the columns corresponding to P ∨ (Q ∧ R) and (P ∨ Q) ∧ (P ∨ R) are identical, the two statement
forms are equivalent.

Problem 1.8. Show that P ∧ (Q ∨ R) is logically equivalent to (P ∧ Q) ∨ (P ∧ R).

Solution of Problem 1.8. We write a truth table for P ∧ (Q ∨ R) and (P ∧ Q) ∨ (P ∧ R):

4
 P Q R P∧Q P∧R (P ∧ Q) ∨ (P ∧ R) P ∧ (Q ∨ R)
T T T T T T T
T T F T F T T
T F T F T T T
T F F F F F F
F T T F F F F
F T F F F F F
F F T F F F F
F F F F F F F

Since the columns corresponding to P ∧ (Q ∨ R) and (P ∧ Q) ∨ (P ∧ R) are identical, the two statement
forms are equivalent.

Problem 1.9. Are the statement forms P =⇒ (Q ∨ R) and (P =⇒ Q) ∨ (P =⇒ R) logically equivalent?

Solution of Problem 1.9. We could write a truth table for P =⇒ (Q ∨ R) and (P =⇒ Q) ∨ (P =⇒ R):

P Q R Q∨R P =⇒ Q P =⇒ R P =⇒ (Q ∨ R) (P =⇒ Q) ∨ (P =⇒ R)
T T T T T T T T
T T F T T F T T
T F T T F T T T
T F F F F F F F
F T T T T T T T
F T F T T T T T
F F T T T T T T
F F F F T T T T
Since the columns corresponding to P =⇒ (Q ∨ R) and (P =⇒ Q) ∨ (P =⇒ R) are identical, the two
statements are logically equivalent.
An alternate proof would go as follows:

P =⇒ (Q ∨ R) ≡ (¬P) ∨ (Q ∨ R) ≡ (¬P) ∨ Q ∨ R

while

5
 (P =⇒ Q) ∨ (P =⇒ R) ≡ ((¬P) ∨ Q) ∨ ((¬P) ∨ R)

≡ (¬P) ∨ (¬P) ∨ Q ∨ R

≡ (¬P) ∨ Q ∨ R

and thus P =⇒ (Q ∨ R) ≡ (P =⇒ Q) ∨ (P =⇒ R).

Problem 1.10. Are the statement forms P =⇒ (Q ∧ R) and (P =⇒ Q) ∧ (P =⇒ R) logically equivalent?

Hint. Try to use Problem 1.7.

Solution of Problem 1.10. We could write a truth table for P =⇒ (Q ∧ R) and (P =⇒ Q) ∧ (P =⇒ R):

P Q R Q∧R P =⇒ Q P =⇒ R P =⇒ (Q ∧ R) (P =⇒ Q) ∧ (P =⇒ R)
T T T T T T T T
T T F F T F F F
T F T F F T F F
T F F F F F F F
F T T T T T T T
F T F F T T T T
F F T F T T T T
F F F F T T T T
Since the columns corresponding to P =⇒ (Q ∧ R) and (P =⇒ Q) ∧ (P =⇒ R) are identical, the two
statements are logically equivalent.
An alternate proof (assuming we know the result from Problem 1.7 ) would go as follows:

P =⇒ (Q ∧ R) ≡ (¬P) ∨ (Q ∧ R) ≡ ((¬P) ∨ Q) ∧ ((¬P) ∨ R)

while

(P =⇒ Q) ∧ (P =⇒ R) ≡ ((¬P) ∨ Q) ∧ ((¬P) ∨ R)

and thus P =⇒ (Q ∧ R) ≡ (P =⇒ Q) ∧ (P =⇒ R).

6
Problem 1.11. Show that the statement forms (P ∨ Q) =⇒ R and (P =⇒ R) ∧ (Q =⇒ R) are logically
equivalent.

Hint. Try to use DeMorgan’s Laws and Problem 1.7.

Solution of Problem 1.11. We could write a truth table for (P ∨ Q) =⇒ R and (P =⇒ R) ∧ (Q =⇒ R):

P Q R P∨Q Q =⇒ R P =⇒ R (P ∨ Q) =⇒ R (P =⇒ R) ∧ (Q =⇒ R)
T T T T T T T T
T T F T F F F F
T F T T T T T T
T F F T T F F F
F T T T T T T T
F T F T F T F F
F F T F T T T T
F F F F T T T T
Since the columns corresponding to (P ∨ Q) =⇒ R and (P =⇒ R) ∧ (Q =⇒ R) are identical, the two
statements are logically equivalent.
An alternate proof (assuming we know DeMorgan’s laws and the result from Problem 1.7) would go as
follows:

(P ∨ Q) =⇒ R ≡ (¬(P ∨ Q) ∨ R) ≡ ((¬P) ∧ ¬Q) ∨ R

≡ ((¬P) ∨ R) ∧ ((¬Q) ∨ R)

while

(P =⇒ R) ∧ (Q =⇒ R) ≡ ((¬P) ∨ R) ∧ ((¬Q) ∨ R)

and thus (P ∨ Q) =⇒ R ≡ (P =⇒ Q) ∧ (P =⇒ R).

Problem 1.12. For all the statement forms below write a logically equivalent statement form that involves
only the logical connective ¬ and ∨.

1. P ∨ (Q ∧ R)

7
 2. (P ∨ Q) ∧ (P ∨ R)

3. P ⇐⇒ Q

Hint. Try to use DeMorgan’s laws.

Solution of Problem 1.12. 1. P ∨ (Q ∧ R) ≡ P ∨ ¬((¬Q) ∨ (¬R))

2. (P ∨ Q) ∧ (P ∨ R) ≡ ¬((¬(P ∨ Q)) ∨ ¬(P ∨ R))

3. P ⇐⇒ Q ≡ (P =⇒ Q) ∧ (Q =⇒ P) ≡ ((¬P) ∨ Q) ∧ ((¬Q) ∨ P) ≡ ¬((¬((¬P) ∨ Q)) ∨ ¬((¬Q) ∨ P))

Problem 1.13. For all the statement forms below write a logically equivalent statement form that involves
only the logical connective ¬ and ∧.

1. P ∧ (Q ∨ R)

2. (P ∧ Q) ∨ (P ∧ R)

3. P ⇐⇒ Q

Hint. Try to use DeMorgan’s laws.

Solution of Problem 1.13. 1. P ∧ (Q ∨ R) ≡ P ∧ ¬((¬Q) ∧ (¬R))

2. (P ∧ Q) ∨ (P ∧ R) ≡ ¬((¬(P ∧ Q)) ∧ ¬(P ∧ R))

3. P ⇐⇒ Q ≡ (P =⇒ Q) ∧ (Q =⇒ P) ≡ ((¬P) ∨ Q) ∧ ((¬Q) ∨ P) ≡ (¬((¬¬P) ∧ ¬Q)) ∧ ¬((¬¬Q) ∧

¬P) ≡ (¬(P ∧ ¬Q)) ∧ ¬(Q ∧ ¬P)

Problem 1.14. Are the statement forms [(¬P) =⇒ [Q ∧ ¬Q]] and P logically equivalent?

2 Quantifiers

Problem 2.1. Let x0 ∈ (a, b), ℓ ∈ R and f : (a, x0 ) ∪ (x0 , b) → R. We say that ℓ is the limit of f at x0 , and
we write limx→x0 f (x) = ℓ, if for all ε > 0 there exists δ > 0 such that if x satisfies 0 < |x − x0 | < δ then
| f (x) − ℓ| < ε. Formally,

lim f (x) = ℓ ⇐⇒ (∀ε > 0)(∃δ > 0)(∀x)[0 < |x − x0 | < δ =⇒ | f (x) − ℓ| < ε].
x→x0

Negate the statement (∀ε > 0)(∃δ > 0)(∀x)[0 < |x − x0 | < δ =⇒ | f (x) − ℓ| < ε].

8
Hint. Remember that ¬(P =⇒ Q) ≡ P ∧ ¬Q.

Solution of Problem 2.1.

(∃ε > 0)(∀δ > 0)(∃x)[(0 < |x − x0 | < δ ) ∧ (| f (x) − ℓ| ⩾ ε)].

Problem 2.2.

1. Give a possible definition of even numbers using logical symbols, quantifiers, and only the multipli-
cation operation.

2. Negate the definition you gave above.

Solution of Problem 2.2.

1.
x is even ⇐⇒ (x ∈ Z) ∧ (∃k ∈ Z)(x = 2k)

2.
x is not even ⇐⇒ (x ∈
/ Z) ∨ (∀k ∈ Z)(x ̸= 2k)

Problem 2.3.

1. Give a possible definition of a prime number using logical symbols, quantifiers, and only the multipli-
cation operation.

2. Negate the definition you gave above.

Solution of Problem 2.3.

1.

x is a prime number ⇐⇒ [(x ∈ N) ∧ (x ̸= 1) ∧ [(∀m ∈ N)(∀n ∈ N)[x = mn =⇒ ((m = 1) ∨ (n = 1))]].

2.

x is not a prime number ⇐⇒ [(x ∈

/ N) ∨ (x = 1) ∨ [(∃m ∈ N)(∃n ∈ N)(x = mn) ∧ (m > 1) ∧ (n > 1)].

9
Problem 2.4. Write a formal mathematical expression that expresses the fact that a given sequence (xn )n∈N
does not have a real limit.

Solution of Problem 2.4.

(∀ℓ ∈ R)(∃ε > 0)(∀N ∈ N)(∃n ⩾ N)(|xn − ℓ| ⩾ ε)

Problem 2.5. Negate the statement P : (∀n ∈ Z)(∃k ∈ Z)(n2 + n + 1 = 2k). Try to explain what P and ¬P
mean.

Solution of Problem 2.5. ¬P : (∃n ∈ Z)(∀k ∈ Z)(n2 + n + 1 ̸= 2k). P means that for every integer n the
integer n2 + n + 1 is even. ¬P means that there exists an integer n such that the integer n2 + n + 1 is not
even.

Problem 2.6. Let f be a function from R to R. We say that f is strictly increasing if

(∀x ∈ R)(∀y ∈ R)[(x < y) =⇒ ( f (x) < f (y))].

Negate the statement above.

Solution of Problem 2.6.

(∃x ∈ R)(∃y ∈ R)[(x < y) ∧ ( f (x) ⩾ f (y))].

Problem 2.7. Let f be a function from R to R. Define what it means for f to be strictly decreasing

Solution of Problem 2.7.

f is strictly decreasing ⇐⇒ (∀x ∈ R)(∀y ∈ R)[(x < y) =⇒ ( f (x) > f (y))].

Problem 2.8. Let f be a function from R to R. Write a formal mathematical expression which expresses the
fact that it is not true that f is strictly decreasing or strictly increasing.

10
Solution of Problem 2.8. If it is not true that f is strictly decreasing or strictly increasing then f is not strictly
decreasing and not strictly decreasing. Formally, it can be expressed with the following statement:

[(∃x ∈ R)(∃y ∈ R)[(x < y) ∧ ( f (x) ⩽ f (y))]] ∧ [(∃w ∈ R)(∃z ∈ R)[(w < z) ∧ ( f (w) ⩾ f (z))]].

Problem 2.9. Define formally what it means that an integer k divides an integer n.

Solution of Problem 2.9.

k divides n ⇐⇒ (∃m ∈ Z)(n = km)

Problem 2.10. Give a formal definition of what it means for a number x to be a rational number.

Solution of Problem 2.11.

p
x is rational ⇐⇒ (x ∈ R) ∧ [(∃p ∈ Z)(∃q ∈ N) x = ]
q

Problem 2.11. Give a formal definition of what it means for a number x to be a irrational number.

Solution of Problem 2.11. Since x is irrational if and only if it is not rational,

p
x is irrational ⇐⇒ (x ∈
/ R) ∨ [(∀p ∈ Z)(∀q ∈ N)(x ̸= )].
q

Problem 2.12. What is the truth value of the statement (∀x ∈ R)(∃y ∈ R)(∀z ∈ R)[xy = xz]?

Solution of Problem 2.12. The statement is false, i.e., (∃x ∈ R)(∀y ∈ R)(∃z ∈ R)[xy ̸= xz] is true. To see
this, let x = 1, and let y ∈ R be given. If y ̸= 0, put z = 2y. Then xy = y ̸= 2y = xz. If now y = 0, put z = 1.
Then xy = 0 ̸= 1 = xz.

Problem 2.13. What is the truth value of the statement (∃y ∈ R)(∀x ∈ R)(∃z ∈ R)[xy = xz]?

Solution of Problem 2.13. The statement is true. To see this, put y = 1, and let x ∈ R be given. Put z = 1.
Then we have xy = x = xz.

11
3 Proofs

Problem 3.1. Prove that the equation (E) : 7x − 2 = 0 has a unique solution in R.

Solution of Problem 3.1. Let x0 = 72 . Then x0 ∈ R and 7x0 − 2 = 7 27 − 2 = 2 − 2 = 0, so the equation (E)
has a solution. Assume now that y is a real solution to the equation (E), then 7y − 2 = 0 and thus y = 27 = x0 .
Therefore, the equation (E) has a unique solution.

Problem 3.2. Prove that the equation (E) : − 3x + 8 = 0 has a unique solution in R.

Solution of Problem 3.2. Assume that y and z are real solutions to the equation (E), then −3y + 8 = −3z + 8
and thus −3y = −3z. Since −3 ̸= 0 it follows that y = z. Therefore, the equation (E) has at most one
solution. Now, let x = 38 . Then x ∈ R and −3x + 8 = −3 38 + 8 = −8 + 8 = 0, so the equation (E) has at least
solution. We can thus conclude that the equation (E) at a unique solution.

Problem 3.3. Let a, b, c ∈ R with a ̸= 0. Prove that the equation (E) : ax + b = c has a unique solution in
R.

c−b
Solution of Problem 3.3. Let x0 = a . Then x0 ∈ R and ax0 + b = a c−b
a + b = c − b + b = c, and hence the

equation (E) has a solution. Assume now that y and z are real solutions to the equation (E), then ay + b = c
and az + b = c. Therefore,
0 = c − c = (ay + b) − (az + b) = a(y − z).

Since a ̸= 0, one has y − z = 0, and thus y = z. The equation (E) has a unique solution.

Problem 3.4. Let a, b, and c be integers. Prove that for all integers m and n, if a divides b and a divides c,
then a divides (bm + cn).

Solution of Problem 3.4. Let m and n be fixed integers. Assume that a divides b and that a divides c. Then
there exist integers j and k such that b = a j and c = ak. (We must show that there exists an integer l such
that bm + cn = al.) Observe that

bm + cn = a jm + akn = a( jm + kn).

Put l = jm + kn. Then l is an integer such that bm + cn = al, and therefore a | (bm + cn). Since m and n
were fixed but arbitrary, we proved that for all integers m and n, if a divides b and a divides c, then a divides
(bm + cn).

Problem 3.5. Prove that if m and n are even, then m + n is even.

12
Solution of Problem 3.5. Let m and n be fixed integer and assume that m and n are fixed even numbers.
Then there exist integers j and k such that m = 2 j and n = 2k. Let l = j + k, then l is an integer, and we have

m + n = 2 j + 2k = 2( j + k) = 2l.

Therefore, m + n is even. Since m and n were fixed but arbitrary even numbers then the proof is complete.

Problem 3.6. Prove that if m is even and n is odd, then m + n is odd.

Solution of Problem 3.6. Assume that m is a fixed even number and n is a fixed odd number. Then there
exist integers j and k such that m = 2 j + 1 and n = 2k. Let l = j + k, then l is an integer, and we have

m + n = 2 j + 1 + 2k = 2( j + k) + 1 = 2l + 1.

Therefore, m + n is odd. The proof is complete since m and n were fixed but arbitrary.

Problem 3.7. Prove that for all m, n ∈ Z, if m is even, then mn is even.

Solution of Problem 3.7. Let m and n be integers and assume that m is even. Then there exists and integer k
such that m = 2k. Put l = kn. Then l is an integer, and we have mn = 2kn = 2l. Therefore, mn is even.

Problem 3.8. Show that for all n ∈ Z, 4n + 7 is odd.

Solution of Problem 3.8. Fix n ∈ Z. Let k = 2n + 3, then k is an integer, and we have

4n + 7 = 2(2n) + 2(3) + 1 = 2(2n + 3) + 1 = 2k + 1.

Therefore, 4n + 7 is odd. Since n was fixed but arbitrary the conclusion follows.

Problem 3.9. Let n be an integer. Prove that if n2 is even, then n is even.

Solution of Problem 3.9. Let n ∈ Z and assume n is not even. Then n is odd, hence there is some k ∈ Z such
that n = 2k + 1. Thus n2 = (2k + 1)2 = 2(2k2 + 2k) + 1 = 2r + 1, with r = 2k2 + 2k ∈ Z, and thus n2 is odd.
Therefore, by contraposition, if n is even, then n2 is even.

Problem 3.10. Let n be an integer. Prove that if n3 is even, then n is even.

Solution of Problem 3.10. Let us prove the contrapositive “if n is not even then n3 is not even”, or equiv-
alently “if n is odd then n3 is odd”. Assume that n is odd, then there exists k ∈ Z such that n = 2k + 1,
and hence n3 = (2k + 1)3 = (2k)3 + 3 · (2k)2 + 3 · 2k + 1 = 2(4k3 + 6 · k2 + 3k) + 1 = 2r + 1, where r =
4k3 + 6 · k2 + 3k is an integer. Therefore n3 is odd.

13
Problem 3.11. For this problem you can use the following fact that will be proven later: 3 does not divides
n if and only if there exists an integer k and an integer i ∈ {1, 2} such that n = 3k + i.
Prove that for every integer n, if 3 divides n2 then 3 divides n.

Solution of Problem 3.12. Let n be an integer and assume that 3 does not divide n. Then there exists an
integer k and an integer i ∈ {1, 2} such that n = 3k + i. Therefore,

n2 = (3k + i)2 = 9k2 + 6ki + i2 = 3(3k2 + 2ki) + i2 .

If i = 1 then i2 = 12 = 1, and n2 = 3r + 1 with r = 3k2 + 2ki ∈ Z. Otherwise, if i = 2 then n2 = 3(3k2 +

2ki) + 22 = 3(3k2 + 2ki) + 3 + 1 = 3(3k2 + 2ki + 1) + 1 = 3s + 1, with s = 3k2 + 2ki + 1 ∈ Z. In any case,
n2 = 3t + 1 for some integer t and thus 3 does not divide n2 .

Problem 3.12. Prove that there are no integers m and n such that 8m + 26n = 1.

Solution of Problem 3.12. Assume, towards a contradiction, that there exist integers m and n such that 8m +
26n = 1. Then 1 = 2(4m + 13n) = 2r, with r = 4m + 13n ∈ Z, and 1 would be even. But 1 is odd, a
contradiction. Thus, there are no integers m and n such that 8m + 26n = 1.

Problem 3.13. Are there integers m and n such that m2 = 4n + 3?

Solution of Problem 3.13. Assume, towards a contradiction, that there exist integers m and n such that m2 =
4n + 3. Then m2 = 2(2n + 1) + 1, so m2 is odd, and therefore m is odd. Thus there is some k ∈ Z such that
m = 2k + 1, and we have
4n + 3 = (2k + 1)2 = 4k2 + 4k + 1,

so 4n + 2 = 4k2 + 4k, and thus 2n + 1 = 2k2 + 2k = 2(k2 + k). Hence 2n + 1 is both even and odd, a
contradiction. Therefore, there do not exist integers m and n such that m2 = 4n + 3.

Problem 3.14. Let x ∈ R. Show that if for all ε > 0, |x| < 2ε, then x = 0.

Solution of Problem 3.14. Let x ∈ R. Assume that for all ε > 0, |x| < 2ε and for the sake of a contradiction
|x|
assume that x ̸= 0. If we put ε0 = 4 then ε0 > 0. Therefore, |x| < 2ε0 = 2 |x|
4 . Since |x| =
̸ 0, it follows that
|x|
|x| < 2 and thus 1 < 12 , a contradiction.
√
3
Problem 3.15. Prove that 2 is irrational.

Hint. Use Problem 3.10.

14
 √
3 √
3 p
Solution of Problem 3.15. Assume by contradiction that 2 is rational and write, as we may, 2 = q with
p ∈ Z, q ∈ Z, q > 0, where p and q have no common factors. Thus, 2 = ( qp )3 and 2q3 = p3 , which means
that p3 is even. By Problem 3.10 above p is even, and there exists k ∈ Z such that p = 2k. It follows that
p3 8k3
q3 = 2 = 2 = 2 · 2k3 and hence q3 is even. By Problem 3.10, q is even, a contradiction. Indeed, p and q
being even they have 2 as a common factor which contradicts our assumption.
√
Problem 3.16. Show that 3 is irrational.

Hint. Use Problem 3.11.

√
Solution of Problem 3.16. Assume, towards a contradiction, that 3 is rational. Then there exist nonnega-
√
tive integers m and n, with n ̸= 0, such that 3 = mn . Without loss of generality, we may assume that m and
n have no common factors. Then m2 = 3n2 , so 3 divides m2 , and by the result of Problem 3.11 it follows
that 3 divides m. Thus there is some k ∈ Z such that m = 3k. Hence we have

9k2 = m2 = 3n2 ,

so n2 = 3k2 . Thus 3 divides n2 , and applying the results of Problem 3.11 again, it follows that 3 divides
n. But then both m and n are divisible by 3, so they share a common factor, contradicting our assumption.
√
Therefore, 3 is irrational.

Problem 3.17. Show that log(3) is irrational.

Hint. You can use the the following property of the log function: x = log(3) ⇐⇒ 2x = 3 (no proof needed)
You can also use the binomial formula (no proof needed). Everything else that you might need needs to be
proven.

√ √
Problem 3.18. Prove that for all real numbers x and y with y ⩾ 0, if x2 ⩾ 4y, then x ⩾ 2 y or x ⩽ −2 y.

√
Solution of Problem 3.18. Let x ∈ R and y > 0 and assume that x2 ⩾ 4y. Either x ⩾ 2 y and the conclusion
√ √ √ √ √
follows, or x < 2 y but then 0 ⩽ x2 − 4y = (x − 2 y)(x + 2 y) and hence x + 2 y ⩽ 0, i.e. x ⩽ −2 y, and
the conclusion follows.

Problem 3.19. Prove that for all integers k, k(k + 3) is even.

Solution of Problem 3.19. Let k ∈ Z.

Case 1 k is even and there exists n ∈ Z such that k = 2n. So, k(k + 3) = 2n(2n + 3) which is even.

15
Case 2 k is odd and there exists n ∈ Z such that k = 2n + 1. So, k(k + 3) = (2n + 1)(2n + 4) = 2(2n + 1)(n + 2)
which is even.

Problem 3.20. Prove that for all integers k, (k + 1)(k + 6) is even.

Solution of Problem 3.20. Let k ∈ Z. Either k is even and there exists n ∈ Z such that k = 2n, and (k +
1)(k + 6) = (2n + 1)(2n + 6) = 2(2n + 1)(n + 3) which is even, or k is odd and there exists n ∈ Z such that
k = 2n + 1, and thus (k + 1)(k + 6) = (2n + 2)(2n + 7) = 2(n + 1)(2n + 7) is even.

For the following problems we recall the definition of the absolute value function

x if x ⩾ 0

|x| :=
−x if x < 0


Problem 3.21. Show that for all x ∈ R, |x| ⩾ 0 with |x| = 0 if and only if x = 0.

Proof of Problem 3.21. We prove the first part of the statement. Let x ∈ R. Then, either x ⩾ 0 or x < 0. If
x ⩾ 0 then by definition |x| = x ⩾ 0. Otherwise, if x < 0 then by definition |x| = −x > 0. For the equivalence
in the second part, if x = 0 then by definition |x| = 0. If |x| = 0 then by definition |x| = x and thus x = 0.

Problem 3.22. Prove that for all real numbers x and y, |x − y| = |y − x|.

Proof of Problem 3.22. Let x, y ∈ R. In the case x − y ⩾ 0 then y − x ⩽ 0 and |x − y| = x − y, but |y − x| =

−(y − x) = x − y and thus |x − y| = |y − x|. In the case x − y < 0 then y − x > 0 and |x − y| = −(x − y) = y − x,
but |y − x| = y − x and thus |x − y| = |y − x|. Therefore in all cases |x − y| = |y − x|.

Problem 3.23. Prove that for all real numbers x and y, |xy| = |x||y|.

Problem 3.24. Let x ∈ R and M ⩾ 0. Show that |x| ⩽ M ⇐⇒ −M ⩽ x ⩽ M.

Proof of Problem 3.24. Let x ∈ R and M ⩾ 0.

Proof of =⇒ : Assume that |x| ⩽ M, then if x ⩾ 0, then x = |x| and −M ⩽ 0 ⩽ x = |x| ⩽ M. Otherwise,
if x < 0 then |x| = −x and −M ⩽ 0 < −x = |x| < M.
Proof of ⇐=: Assume that −M ⩽ x ⩽ M. In the case x ⩾ 0 then |x| = x, but x ⩽ M and it follows that
|x| ⩽ M. In the case x < 0 then |x| = −x, but since −M ⩽ x then −x ⩽ M and hence |x| ⩽ M

Problem 3.25. Prove that for all real numbers x and y, |x + y| ⩽ |x| + |y|.

16
Hint. You could use Problem 3.24.

Proof of Problem 3.25. Let x, y ∈ R. Since −|x| ⩽ x ⩽ |x| and −|y| ⩽ y ⩽ |y| then by adding up theses two
inequalities −(|y| + |x|) ⩽ x + y ⩽ |x| + |y| and by Problem 3.24 |x + y| ⩽ |x| + |y|.

Problem 3.26. Prove that for all x, y, z ∈ R, |x − y| ⩽ |x − z| + |y − z|.

Hint. You could use Problem 3.25.

Proof of Problem 3.26. Let x, y, z ∈ R and set a = x − z and b = z − y. It follows from Problem 3.25 that
|x − y| = |a + b| ⩽ |a| + |b| = |x − z| + |y − z|.

Problem 3.27. Prove that for all real numbers x and y, |x| − |y| ⩽ |x − y|.

Hint. You could use Problem 3.25.

Proof of Problem 3.27. Let x, y ∈ R, then by Problem 3.25 |x| = |x − y + y| ⩽ |x − y| + |y|, and |y| = |y − x +
x| ⩽ |y − x| + |x|. Thus |x| − |y| ⩽ |x − y| and |y| − |x| ⩽ |x − y| and the conclusion follows.

Problem 3.28. Let x, y be real numbers. Show that

∀ε > 0, x < y + ε ⇐⇒ x ⩽ y.

Proof of Problem 3.28. Proof of ⇐=: Assume that x ⩽ y, then if ε > 0 it follows that x < y + ε. Proof of
=⇒ : Assume that x < y + ε for all ε > 0. Assume by contradiction that x > y and let ε0 = x − y > 0. By
our assumption, x < y + ε0 = y + (x − y) = x; a contradiction.

Problem 3.29. Let x, y be real numbers. Show that x > y − ε for all ε > 0 if and only if x ⩾ y.

Proof of Problem 3.29. Assume that x < y + ε for all ε > 0. Assume by contradiction that x > y and let
ε0 = x − y > 0. By our assumption, x < y + ε0 = y + (x − y) = x; a contradiction. For the other direction, if
x ⩾ y and ε > 0 then x > y − ε.

x+y
Problem 3.30. Prove that for all real numbers x and y, if x < y, then x < 2 < y.

Solution of Problem 3.30. Let x and y be real numbers such that x < y. Then 2x = x + x < x + y, and thus
x+y x+y
x< 2 . Similarly, x + y < y + y = 2y, and thus 2 < y. Combining our results, we obtain
x+y
x< <y
2
as desired.

17
Problem 3.31. Prove that for all positive real numbers x, the sum of x and its reciprocal is greater than 2.

Solution of Problem 3.31. Let x be a real number such that x > 0. Observe that 0 ⩽ (x − 1)2 = x2 − 2x + 1,
so 2x ⩽ x2 + 1 and thus since x > 0 one has

1 1 1
2= · 2x ⩽ · (x2 + 1) = x +
x x x

as desired.

√ x+y
Problem 3.32. 1. Prove that for all x, y ∈ R+ , xy ⩽ 2 .

√ x+y
2. Show that that for all x, y ∈ R+ , xy = 2 if and only if x = y

Solution of Problem 3.32. 1. Let x, y ∈ R+ be given. Observe that 0 ⩽ (x − y)2 = x2 − 2xy + y2 , and so
√ √
4xy ⩽ x2 + 2xy + y2 = (x + y)2 . Thus 2 xy ⩽ x + y, and therefore xy ⩽ x+y
2 .

2. Assuming x = y, we obtain
√ √ 2x x + y
xy = x2 = x = = .
2 2
√ x+y
Conversely, assume xy = 2 . Then 4xy = (x + y)2 = x2 + 2xy + y2 , and rearranging we obtain
0 = x2 − 2xy + y2 = (x − y)2 , so 0 = x − y and therefore x = y.

4 Applications of the Principle of Mathematical Induction

Problem 4.1. Prove that for all integers n ⩾ 1,

n
n(n + 1)(2n + 1)
∑ k2 = 6
.
k=1

Solution of Problem 4.1. First observe that for n = 1 we have

n(n + 1)(2n + 1) 1 · 2 · 3
= = 1,
6 6

but ∑1k=1 k2 = 12 = 1 and thus the equality holds if n = 1. Now let n ⩾ 1 and assume that assume ∑nk=1 k2 =

18
n(n+1)(2n+1)
6 . Then we have
n+1 n
∑ k2 = (n + 1)2 + ∑ k2
k=1 k=1
n(n + 1)(2n + 1)
= (n + 1)2 +
6
6(n + 1)(n + 1) n(n + 1)(2n + 1)
= +
6 6
(n + 1)[6(n + 1) + n(2n + 1)]
=
6
(n + 1)(2n2 + 7n + 6)
=
6
(n + 1)(n + 2)(2n + 3)
=
6
(n + 1)((n + 1) + 1)(2(n + 1) + 1)
= ,
6

so the equality holds for n + 1. Therefore, it follows by the Principle of Mathematical Induction that for all
n(n+1)(2n+1)
integers n ⩾ 1, ∑nk=1 k2 = 6 .

Problem 4.2. Prove that for all integers n ⩾ 0,

n
∑ 2k = 2n+1 − 1.
k=0

Solution of Problem 4.2. First observe that for n = 0, we have 20 = 1 = 20+1 − 1, so the equality holds in
this case. Let n ⩾ 0 and assume that the equality holds for n, i.e., ∑nk=0 2k = 2n+1 − 1. Then we have
n+1 n
∑ 2k = 2n+1 + ∑ 2k = 2n+1 + 2n+1 − 1 = 2 · 2n+1 − 1 = 2(n+1)+1 − 1
k=0 k=0

so the equality holds for n + 1. Therefore, it follows by the Principle of Mathematical Induction that for all
integers n ⩾ 0, ∑nk=0 2k = 2n+1 − 1.

Problem 4.3. Prove that for all integers n ⩾ 1,

n
∑ (2k − 1) = n2 .
k=1

Solution of Problem 4.3. First observe that for n = 1, we have 12 = 1 = 2 · 1 − 1, so the result is true in this
case. Now assume that the result holds for some positive integer m. Then we have
m+1 m
∑ (2k − 1) = 2(m + 1) − 1 + ∑ (2k − 1) = m2 + 2m + 1 = (m + 1)2 ,
k=1 k=1

so the result is true for m + 1. Therefore, it follows by induction that the result holds for all positive integers.

19
Problem 4.4. Prove that for all integers n ⩾ 1,
n
1 n
∑ k(k + 1) = n + 1 .
k=1

Solution of Problem 4.4. For n = 1, ∑1k=1 k(k+1)

1
= 1
1(1+1) = 1
2 = 1
1+1 , and the equality holds. Assume that
the result holds for some positive integer m. Then we have
m+1 m
1 1 1
∑ = +∑
k=1 k(k + 1) (m + 1)(m + 2) k=1 k(k + 1)
1 m
= +
(m + 1)(m + 2) m + 1
1 m2 + 2m
= +
(m + 1)(m + 2) (m + 1)(m + 2)
(m + 1)2
=
(m + 1)(m + 2)
(m + 1)
=
(m + 1) + 1

so the result is true for m + 1. Therefore, it follows by induction that the result holds for all positive integers.

Problem 4.5. Prove that for all integers n ⩾ 1,

n
4n3 − n
∑ (2k − 1)2 = 3
.
k+1

Solution of Problem 4.5. First observe that

4 · 13 − 1
= 1 = (2 · 1 − 1)2 ,
3

so the result is true for n = 1. Now assume that the result holds for some positive integer m. Then we have
m+1
4m3 − m
∑ (2k − 1)2 = (2m + 1)2 + 3
k=1
12m2 + 12m + 3 4m3 − m
= +
3 3
4m3 + 12m2 + 12m + 3 − m
=
3
4(m3 + 3m2 + 3m + 1) − (m + 1)
=
3
4(m + 1)3 − (m + 1)
= ,
3

so the result is true for m + 1. Therefore, it follows by induction that the result holds for all positive integers.

20
Problem 4.6. Conjecture a formula for ∑nk=1 (−1)k k2 , for all n ⩾ 1 and then prove the formula is correct
using induction.

Solution of Problem 4.6. After calculating a few iterates, you should see that for all n ⩾ 1 the formula you
are looking for is
n  
k 2 n n(n + 1)
∑ (−1) k = (−1)
2
.
k=1
Indeed, this holds for n = 1. Now assume the formula holds for some positive integer m. Then we have
m+1  
k 2 m+1 2 m m(m + 1)
∑ (−1) k = (−1) (m + 1) + (−1)
2
k=1
(−1)m+1
2m2 + 4m + 2 − m2 − m


=
2
(−1)m+1 2

= m + 3m + 2
2  
m+1 (m + 1)((m + 1) + 1)
= (−1) ,
2

so the formula holds for m + 1. Therefore, by induction, the result holds for all positive integers.

Problem 4.7. Prove that for all integers n ⩾ 1, n < 10n .

Solution of Problem 4.7. The result is true for n = 1. Now assume the result holds for some positive integer
m. Then we have
m + 1 < 10m + 1 < 9 · 10m + 10m = 10 · 10m = 10m+1 ,

so the result is true for m + 1. Therefore, by induction it follows that the result is true for all positive
integers.

4 n


Problem 4.8. Prove that for all integers n ⩾ 7, 3 > n.

7
Solution of Problem 4.8. Direct calculation show that 43 > 7. Now assume that for some m ⩾ 7 we have
4 m

m
m
> m. Then 3 < 7 ⩽ m < 43 , so 1 < 13 43 and thus


3
 m
1 4 m
   m+1
4 4
m+1 < + = ,
3 3 3 3

so the result holds for m + 1. Therefore, by induction it follows that the result holds for all positive integers
n ⩾ 7.

Problem 4.9. Prove that for all integers n ⩾ 1, n3 + 8n + 9 is divisible by 3.

21
Solution of Problem 4.9. First observe that 13 + 8 · 1 + 9 = 18 = 3 · 6, so the result is true for n = 1. Now
assume that the result holds for some positive integer m. Then there is some integer k such that m3 +8m+9 =
3k, and we have

(m + 1)3 + 8(m + 1) + 9 = (m3 + 3m2 + 3m + 1) + (8m + 8) + 9

= (m3 + 8m + 9) + 3m2 + 3m + 9

= 3(k + m2 + m + 3),

so the result is true for m + 1. Therefore, by induction it follows that the result holds for all positive integers.

Problem 4.10. Prove that for all integers n ⩾ 1, 32n − 1 is divisible by 8.

Solution of Problem 4.10. First observe that 32·1 − 1 = 8 · 1, so the result is true for n = 1. Now assume the
result is true for some positive integer m. Then there is some integer k such that 32m − 1 = 8k, and we have

32(m+1) − 1 = 32m+2 − 1

= 9 · 32m − 1

= 9 · (32m − 1) + 8

= 8 · (9k + 1),

so the result is true for m1 . Therefore, by induction it follows that the result is true for all positive integers.

Problem 4.11. Prove that for all integers n ⩾ 5, n2 < 2n .

Solution of Problem 4.11. Let P(n) be the statement n2 < 2n . Observe that 52 = 25 < 32 = 25 , so P(5) is
true. Now assume that P(m) is true for some m ⩾ 5. Since m ⩾ 5, we have 2m + 1 < 3m < m2 , so that

(m + 1)2 = m2 + 2m + 1 < 2m2 < 2 · 2m = 2m+1 .

and thus P(m + 1) is true. Therefore, by induction it follows that P(n) is true for all n ⩾ 5.

Problem 4.12. Prove that for all integers n ⩾ 4, 2n < n!.

Solution of Problem 4.12. Clearly 24 = 16 < 24 = 4!. Now assume that 2m < m! for some positive integer
m ⩾ 4. Then we have
2m+1 = 2 · 2m < 2m! < (m + 1) · m! = (m + 1)!

Therefore, it follows by induction that 2n < n! for all positive integers n.

22
Problem 4.13. Assuming that (1 + 1n )n < e, for all n ⩾ 1, prove that for all n ⩾ 1, n! > ( ne )n .

Problem 4.14. Show that for all n ⩾ 12 there exist xn ∈ Z and yn ∈ Z such that n = 3xn + 7yn

Problem 4.15. Prove that for all positive integers n, 4n − 1 is divisible by 3.

Solution of Problem 4.15. First note that 41 − 1 = 3, and thus 3 | 41 − 1. Now assume that 4m − 1 is divisible
by 3 for some positive integer m. Then 4m − 1 = 3k for some k ∈ Z, hence we have

4m+1 − 1 = 4(4m − 1) + 3 = 3(4k + 1).

Thus 4m+1 − 1 is divisible by 3, and it follows by induction that 4n − 1 is divisible by 3 for all positive
integers n.

Problem 4.16. Let a1 = 2, and let an+1 = 21 (an + 3) for all n ⩾ 1.

(a) Prove that for all positive integers n, an < an+1 .

(b) Prove that for all positive integers n, an < 3.

1
(c) Prove that for all positive integers n, an = 3 − 2n−1 .

Solution of Problem 4.16. (a) First note that a2 = 12 (a1 + 3) = 12 (2 + 3) = 25 > a1 . Now assume that am+1 >
am for some positive integer m. Then we have

1 1
am+2 = (am+1 + 3) > (am + 3) = am+1 .
2 2

By induction, it follows that an+1 > an for all positive integers n.

(b) Clearly a1 = 2 < 3. Now assume that am < 3 for some positive integer m. Then we have

1 1
am+1 = (am + 3) < (3 + 3) = 3.
2 2

By induction, it follows that an < 3 for all positive integers n.

1 1
(c) Clearly a1 = 2 = 3 − 21−1 . Now assume that am = 3 − 2m−1 for some positive integer m. Then we have
 
1 1 1 1
am+1 = (am + 3) = 6 − m−1 = 3 − m+1−1 .
2 2 2 2
1
By induction, it follows that an = 3 − 2n−1 for all positive integers n.

23
Problem 4.17. Let r ∈ R with r ̸= 1. Prove that
n−1
1 − rn
∑ rk = 1−r
.
k=0

1−r1
Solution of Problem 4.17. Fix r ∈ R with r ̸= 1. Note that ∑1−1 k
k=0 r = 1 = 1−r . Now assume that ∑m−1 k
k=0 r =
1−rm
1−r for some positive integer n. Then we have

m m−1
1−r 1 − rm 1 − rm+1
∑ rk = rm + ∑ rk = rm 1 − r + 1−r
=
1−r
.
k=0 k=0

1−rn
Thus, by induction it follows that ∑n−1 k
k=0 r = 1−r is true for all positive integers n.

Problem 4.18. Prove Bernoulli’s Inequality: Let x > −1. Then for all n ∈ N, (1 + x)n ⩾ 1 + nx.

Solution of Problem 4.18. For n = 1 equality (hence the inequality) holds. Now suppose that (1 + x)m ⩾
1 + mx for some m ∈ N. Then we have

(1 + x)m+1 ⩾ (1 + x)(1 + mx) = 1 + mx + x + mx2 ⩾ 1 + (m + 1)x.

(Note that we used the assumption that x > 1 in the first inequality.) Thus by induction, the inequality holds
for all n ∈ N.

Problem 4.19. Let x, y ∈ R. Prove the binomial theorem: for all integers n ⩾ 1,
n  
n n n−k k
(x + y) = ∑ x y.
k=0 k

Solution of Problem 4.19. First note that

   
1 1 0 1 0 1
(x + y)1 = x + y = x y + x y .
0 1

Now assume that for some m ∈ N we have

  m
m m m−k k
(x + y) = ∑ x y.
k=0 k

24
Then we have

(x + y)m+1 = (x + y)(x + y)m

m  
m m+1−k k m m m−k k+1
 
=∑ x y +∑ x y
k=0 k k=0 k
m m+1 0 m m m+1−k k m−1 m m−k k+1
       
m 0 m+1
= x y +∑ x y +∑ x y + x y
0 k=1 k k=0 k m
m m+1 0 m
       
m m m 0 m+1
= x y +∑ + xm+1−k yk + x y
0 k=1 k k − 1 m
m + 1 m+1 0 m m + 1 m+1−k k
     
m + 1 0 m+1
= x y +∑ x y + x y
0 k=1 k m+1
m+1  
m + 1 m+1−k k
= ∑ x y
k=0 k

as desired.

Problem 4.20. Let n be an integer. Show that if n is even then nk is even for all k ∈ N.

5 Applications of the Principle of Strong Mathematical Induction

Problem 5.1. For i ∈ N, let pi denote the ith prime number, so that

p1 = 2, p2 = 3, p3 = 5, . . . .

n−1
Prove that for all n ∈ N, pn ⩽ 22 .

Hint. For the induction step, given m ∈ N, show that pm+1 ⩽ p1 p2 · · · pm + 1.

1−1 k−1
Solution of Problem 5.1. First observe that p1 = 2 = 22 . Now fix m ∈ N, and assume that pk ⩽ 22 for
1 ⩽ k ⩽ m. Note that pm+1 ⩽ p1 p2 · · · pm + 1, since pk does not divide p1 p2 · · · pm + 1 for 1 ⩽ k ⩽ m. Thus,
we have
m−1 k m −1 m −1 m
pm+1 ⩽ p1 p2 · · · pm + 1 ⩽ 2∑k=0 2 + 1 = 22 + 1 < 2 · 22 = 22 .

Problem 5.2. Show that the principle of strong mathematical induction implies the principle of mathemati-
cal induction.

Solution of Problem 5.2. Assume the principle of strong mathematical induction, and let P(n) be a statement
about the positive integer n. Assume that P(1) is true, and that for all m ∈ N, if P(m) is true then P(m + 1) is

25
true. Let m ∈ N be given, and assume that P(k) is true for 1 ⩽ k ⩽ m. Then P(m) is true, so P(m + 1) is true.
Thus, by the principle of strong mathematical induction, P(n) is true for all n ∈ N. Therefore, the principle
of mathematical induction is true.

Problem 5.3. Show that the principle of mathematical induction implies the principle of strong mathemati-
cal induction.

Solution of Problem 5.3. Assume the principle of mathematical induction, and let P(n) be a statement about
the positive integer n. Assume that P(1) is true, and that for all m ∈ N, if P(k) is true for 1 ⩽ k ⩽ m, then
P(m + 1) is true. For n ∈ N, let Q(n) be the statement ”P(k) is true for all k ⩽ n”. Clearly Q(1) is true.
If m ∈ N and Q(m) is true, then P(k) is true for all 1 ⩽ k ⩽ m. By assumption P(m + 1) is true, and thus
Q(m + 1) is true. By the principle of mathematical induction, Q(n) is true for all n ∈ N. But if Q(n) is
true, then P(n) is true, and thus P(n) is true for all n ∈ N. Therefore, the principle of strong mathematical
induction is true.

6 Sequences defined by a recurrence relation

Problem 6.1. Let a1 = 2, a2 = 4, and an+1 = 7an − 10an−1 for all n ⩾ 2. Conjecture a closed formula for
an and prove your result.

Solution of Problem 6.1. We will show that an = 2n for each n ∈ N. Indeed, a1 = 2 = 21 and a2 = 4 = 22 .
Now assume that for some m ⩾ 2 we have ak = 2k for 1 ⩽ k ⩽ m. Then we have

am+1 = 7am − 10am−1 = 7 · 2m − 10 · 2m−1 = 14 · 2m−1 − 10 · 2m−1 = 4 · 2m−1 = 2m+1 .

Therefore, by the principle of strong mathematical induction it follows that an = 2n for all n ∈ N.

Problem 6.2. Let a1 = 3, a2 = 4, and an+1 = 31 (2an + an−1 ) for all n ⩾ 2. Prove that for all positive integers
n, 3 ⩽ an ⩽ 4.

Solution of Problem 6.2. Clearly 3 ⩽ a1 ⩽ 4 and 3 ⩽ a2 ⩽ 4. Now fix m ⩾ 2, and assume that 3 ⩽ ak ⩽ 4
for all 1 ⩽ k ⩽ m. Then we have
1 1
am+1 = (2am + am−1 ) ⩾ (2 · 3 + 3) = 3
3 3
and
1 1
am+1 = (2am + am−1 ) ⩽ (2 · 4 + 4) = 4.
3 3
Therefore, by the principle of strong mathematical induction it follows that 3 ⩽ an ⩽ 4 for all n ∈ N.

26
Problem 6.3. Consider the sequence (an )∞
n=1 recursively defined as a1 = 1, a2 = 8 and for all n ⩾ 3,

an = an−1 + 2an−2 . Show that for all n ⩾ 1, an = 3 · 2n−1 + 2(−1)n .

Solution of Problem 6.3. Note that 3 · 21−1 + 2(−1)1 = 1 = a1 , 3 · 22−1 + 2(−1)2 = 8 = a2 . Now assume
that for some integer m ⩾ 2, we have ak = 3 · 2k−1 + 2(−1)k whenever 1 ⩽ k ⩽ m. Then we have

am+1 = am + 2am−1

= (3 · 2m−1 + 2(−1)m ) + 2(3 · 2m−2 + 2(−1)m−1 )

= 3 · 2 · 2m−1 + 2((−1)m + (−1)m−1 ) + 2(−1)m−1

= 3 · 2m + 2(−1)m+1 .

Therefore, by the principle of strong mathematical induction it follows that an = 3 · 2n−1 + 2(−1)n for all
integers n ⩾ 1.

Problem 6.4. Consider the sequence (an )∞

n=1 recursively defined as a1 = 2, a2 = 4 and for all n ⩾ 3,

an = 3an−1 − 2an−2 . For all n ⩾ 1, find a closed formula for an .

Solution of Problem 6.4. After performing a few calculations, it should be clear that an = 2n . We prove by
induction that this is the correct formula.
Clearly 21 = 2 = a1 and 22 = 4 = a2 . Now suppose that for some m ∈ N with m ⩾ 2, we have ak = 2k
whenever 1 ⩽ k ⩽ m. Then we have

am+1 = 3am − 2am−2 = 3 · 2m − 2 · 2m−1 = 3 · 2m − 2m = 2m+1 .

Thus, by the principle of strong mathematical induction it follows that an = 2n for all n ∈ N.

7 Set Theory

7.1 Subsets

Problem 7.1. Prove that X ⊆ Y where X = {n ∈ Z | n is a multiple of 6} and Y = {n ∈ Z | n is even}.

Solution of Problem 7.1. Let n ∈ X be given. Then n is a multiple of 6, so there is some k ∈ Z such that
n = 6k. Thus n = 2(3k), so n is even and therefore x ∈ Y .

Problem 7.2. Consider the sets

A = {n ∈ Z | (∃k ∈ Z)(n = 12k + 11)},

B = {n ∈ Z | (∃ j ∈ Z)(n = 4 j + 3)}.

27
(a) Is A ⊆ B? Prove or disprove.

(b) Is B ⊆ A? Prove or disprove.

Solution of Problem 7.2. (a) We will show that A ⊆ B. Suppose n ∈ A. Then there is some k ∈ Z such that
n = 12k + 11. Put j = 3k + 2. Then j ∈ Z and

n = 12k + 11 = 4(3k) + 4(2) + 3 = 4 j + 3,

and therefore n ∈ B.

(b) We will show that B ̸⊆ A, that is, there is some n ∈ B such that n ∈
/ A. Indeed, put n = 7. Then
n = 4(1) + 3 so n ∈ B. Now assume (towards a contradiction) that n ∈ A. Then there is some k ∈ Z such
that 7 = 12k + 11. But then −4 = 12k, which is impossible. Thus n ∈
/ A and therefore B ̸⊆ A.

Problem 7.3. Consider the sets

A = {n ∈ Z | (∃k ∈ Z)(n = 4k + 1)},

B = {n ∈ Z | (∃ j ∈ Z)(n = 4 j − 7}.

Prove that A = B.

Solution of Problem 7.3. Suppose n ∈ A. Then there is some k ∈ Z such that n = 4k + 1. Hence we have

n = 4k + 1 = 4(k + 2) − 7 ∈ B,

and therefore A ⊆ B.
Conversely, assume n ∈ B. Then there is some j ∈ Z such that n = 4 j − 7. Hence we have

n = 4 j − 7 = 4( j − 2) + 1 ∈ A

and therefore B ⊆ A.

Problem 7.4. Consider the sets

A = {n ∈ Z | (∃k ∈ Z)(n = 3k)},

B = {n ∈ Z | (∃i, j ∈ Z)(n = 15i + 12 j)}.

Prove that A = B.

28
Solution of Problem 7.4. Suppose n ∈ A. Then there is some k ∈ Z such that n = 3k. Note that k = 5k − 4k,
so we have
n = 3k = 3(5k − 4k) = 15(k) + 12(−k) ∈ B

and therefore A ⊆ B.
Conversely, suppose n ∈ B. Then there exist i, j ∈ Z such that n = 15i + 12 j. Then n = 3(5i + 4 j) ∈ A,
and therefore B ⊆ A.

Problem 7.5. Prove that X = {n ∈ Z | n + 5 is odd} is the set of all even integers.

Solution of Problem 7.5. Let n ∈ X be given. Then n + 5 is odd, so there is some k ∈ Z such that n + 5 =
2k + 1. Thus n = 2k − 4 = 2(k − 2), so n is even.
Now suppose n ∈ Z is even. Then there is some k ∈ Z such that n = 2k. Thus n + 5 = 2k + 5 =
2(k + 2) + 1, so n + 5 is odd and therefore n ∈ X.

7.2 Complements

Problem 7.6. Let A and B be subsets of an ambient set U. Prove that (A − B) ∪ (B − A) = (A ∪ B) − (A ∩ B).

Solution of Problem 7.6. Using the identities found in section Section 4.2, this can be done with a few
manipulations:

(A − B) ∪ (B − A) = (A ∩ B) ∪ (B ∩ A)

= (A ∪ (B ∩ A)) ∩ (B ∪ (B ∩ A))

= [(A ∪ B) ∩ (A ∪ A)] ∩ [(B ∪ B) ∩ (B ∪ A)]

= (A ∪ B) ∩ (B ∪ A)

= (A ∪ B) ∩ (A ∩ B)

= (A ∪ B) − (A ∩ B).

You could also give a a double-inclusion proof.

First, we show that (A − B) ∪ (B − A) ⊆ (A ∪ B) − (A ∩ B). Note that A ⊆ A ∪ B and A ∩ B ⊆ B, so
(A − B) ⊆ (A ∪ B) − (A ∩ B). Similarly, since B ⊆ A ∪ B and A ∩ B ⊆ A, we have (B − A) ⊆ (A ∪ B) − (A ∩ B),
and therefore (A − B) ∪ (B − A) ⊆ (A ∪ B) − (A ∩ B).
Conversely, assume that x ∈ (A ∪ B) − (A ∩ B). Then in particular, x ∈
/ A ∩ B and x ∈ (A ∩ B). If x ∈ A,
then x ∈
/ B, for otherwise x ∈ A∩B and we obtain a contradiction. Thus x ∈ A−B ⊆ (A−B)∪(B−A). If now
x ∈ B, then x ∈
/ A, for otherwise x ∈ A ∩ B and we obtain a contradiction. Thus x ∈ B − A ⊆ (A − B) ∪ (B − A).
Therefore, we have (A ∪ B) − (A ∩ B) ⊆ (A ∪ B) − (A ∩ B) and equality holds.

29
7.3 Arbitrary unions and intersections
S∞
Problem 7.7. For i ∈ N, let Ai = (−i, i). Compute i=1 Ai .

S∞
Solution of Problem 7.7. We will show that i=1 Ai =R
S∞
• i=1 Ai ⊆R
S∞ S∞
If x ∈ i=1 Ai , then x ∈ Ai0 for some i0 . Since Ai0 ⊆ R, then x ∈ R. Since x ∈ i=1 Ai was arbitrary, it
S∞
follows that i=1 Ai ⊆ R.
S∞
• R⊆ i=1 Ai

If x ∈ R, then there is some i ∈ N such that −i < x < i. Indeed if x = 0 any i ⩾ 1 would work, otherwise
if x ̸= 0 then |x| > 0 and by the Archimedean principle there is an integer i ⩾ 1 such that |x| < i and
S∞ S∞
then −i < x < i. Thus x ∈ (−i, i) = Ai and x ∈ i=1 Ai . Therefore R ⊆ i=1 Ai

S∞
By combining the two inclusions one has that R = i=1 Ai .

T∞
Problem 7.8. For i ∈ N, let Ai = (−i, i). Compute i=1 Ai .

T∞
Solution of Problem 7.8. We will show that i=1 Ai = (−1, 1).

T∞
• i=1 Ai ⊆ (−1, 1).
T∞ T∞
If x ∈
/ (−1, 1), then x ∈
/ A1 , and thus x ∈
/ i=1 Ai . Therefore i=1 Ai ⊆ (−1, 1).
T∞
• (−1, 1) ⊆ i=1 Ai .
T∞
If x ∈ (−1, 1), then x ∈ (−i, i) for all i ∈ N and thus (−1, 1) ⊆ i=1 Ai .

Problem 7.9. For i ∈ N, let Ai = 0, 1 − 1i . Compute i∈N Ai .

  S

S
Solution of Problem 7.9. We will show that i∈N Ai = [0, 1).

⊆ [0, 1)
S
• i∈N Ai

If x ∈ then x ∈ A j for some j ∈ N. Thus 0 ⩽ x ⩽ 1 − 1j < 1, and hence x ∈ [0, 1). Therefore,
S
i∈N Ai ,

⊆ [0, 1).
S
i∈N Ai

30
 • [0, 1) ⊆
S
i∈N Ai .

1
If now x ∈ [0, 1), then 0 ⩽ x < 1. Thus 0 < 1 − x and 1−x > 0, and by the Archimedean principle there
1 1
is some j ∈ N such that 1−x ⩽ j, and hence j ⩽ 1 − x. Thus 0 ⩽ x ⩽ 1 − 1j , and x ∈ [0, 1 − 1j ] = A j .
Therefore x ∈ and [0, 1) ⊆
S S
i∈N Ai , i∈N Ai .

S
By definition of equality between sets we have proven that i∈N Ai = [0, 1).

Problem 7.10. For i ∈ N, let Ai = 0, 1 − 1i . Compute i∈N Ai .

  T

= {0}.
T
Solution of Problem 7.10. We will show that i∈N Ai

⊆ {0}.
T
• i∈N Ai

Suppose x ∈ Then in particular x ∈ A1 = [0, 1 − 1] = {0}, and therefore ⊆ {0}.

T T
i∈N Ai . i∈N Ai

• {0} ⊆
T
i∈N Ai
h 
Conversely, 0 ∈ 0, 1 − 1j = A j for all j ∈ N, and therefore {0} ⊆ i∈N Ai .
T

Problem 7.11. Let Xn = ( 2n , 2n] for every integer n ⩾ 2.

S∞
1. Compute n=2 Xn .

T∞
2. Compute n=2 Xn .

S∞
Solution of Problem 7.11. 1. We will show that n=2 Xn = (0, ∞).

= ( 2k , 2k]
S∞ S∞
First, we show that n=1 Xn ⊆ (0, ∞). Let x ∈ n=2 Xn , then there exists k ⩾ 2 such that x ∈ Xk
2 2
and hence k < x ⩽ 2k. Since it follows from k ⩾ 2 that k ⩾ 1 > 0 and 2k < ∞ one has 0 < x < ∞ and
S∞
thus x ∈ (0, ∞). Therefore n=2 Xn ⊆ (0, ∞)

Assume now that x ∈ (0, ∞), then x > 0 and also 2x > 0. On one hand, if follows from the Archimedean
2
principle that there is some n1 ⩾ 2 such that n1 > 2x , so 2n1 ⩾ x. On the other hand, x > 0 and it
2 2
follows from the Archimedean principle that there exists n2 ⩾ 2 such that x < n2 and hence x > n2 .
2 2 S∞
Let k = max{n1 , n2 } ⩾ 2 then k ⩽ n2 < x ⩽ 2n1 ⩽ k and hence x ∈ Xk . Therefore, (0, ∞) ⊆ n=2 Xn .

T∞
2. We will show that n=2 Xn = (1, 4].

then x ∈ Xn = ( 2n , 2n] for all integers n ⩾ 2. In particular, x ∈ X2 = ( 22 , 2 · 2] = (1, 4].

T∞
Let x ∈ n=2 Xn
T∞
Therefore, n=2 Xn ⊆ (1, 4].

31
 2
Now, let x ∈ (1, 4] then 1 < x ⩽ 4 and for all n ⩾ 2, it follows that n ⩽ 1 < x ⩽ 4 ⩽ 2n. Therefore
x ∈ ( n2 , 2n] = Xn for all n ⩾ 2, and (1, 4] ⊆
T∞
n=2 Xn .

Problem 7.12. Let I be a nonempty set and let {Ai : i ∈ I} be an indexed family of sets. Let X be a non-empty
set. Suppose that for all i ∈ I, X ⊆ Ai . Prove that X ⊆
T
i∈I Ai .

Solution of Problem 7.12. Suppose x ∈ X, and let i ∈ I be given. Since X ⊆ Ai , x ∈ Ai . Since i ∈ I was
arbitrary, x ∈ Ai for all i, thus x ∈ and therefore X ⊆
T T
i∈I Ai , i∈I Ai .

Problem 7.13. Let {Ai : i ∈ N} be an indexed family of sets. Assume that for all i ∈ N, Ai+1 ⊆ Ai . Prove
S
that i∈N Ai = A1 .

Solution of Problem 7.13. If x ∈ A1 , then there is some i ∈ N such that x ∈ Ai (namely, i = 1), and thus
x∈
S
i∈N Ai .

Next, we show that Ai ⊆ A1 for all i ∈ N. Clearly A1 ⊆ A1 . If now Am ⊆ A1 , then Am+1 ⊆ Am ⊆ A1 , so

Am+1 ⊆ A1 . Thus it follows by induction that An ⊆ A1 for all n ∈ N.
Now assume x ∈ Then there is some n ∈ N such that x ∈ An . Since An ⊆ A1 , it follows that
S
i∈N Ai .

x ∈ A1 .

Problem 7.14. Let (Xi )i∈I be a collection of subsets of an ambient set U. Show that
\ [
Xi = X i.
i∈I i∈I

⊆
T S
Solution of Problem 7.14. We first prove the inclusion i∈I Xi i∈I X i .

= 0/ then the inclusion holds, otherwise let z ∈ Then z ∈

T T T
If i∈I Xi i∈I Xi . / i∈I Xi (by definition of the
complement), and it follows that z ∈
/ Xi0 for some i0 ∈ I (by definition of the intersection). Thus, z ∈ Xi0 (by
definition of the complement), which means that z ∈
S
i∈I X i (by definition of the union).

= 0/ then the inclusion holds, otherwise let z ∈ Then z ∈ Xi0

S S
For the reverse inclusion, if i∈I X i i∈I X i .

for some i0 ∈ I (by definition of the union), and thus z ∈

/ Xi0 (by definition of the complement). It follows
that z ∈ (by definition of the intersection), and hence z ∈
T T
/ i∈I Xi i∈I Xi (by definition of the complement).
S T
Therefore, it follows from the definition of equality between sets that i∈I Xi = i∈I X i .

Problem 7.15. Let (Xi )i∈I be a collection of subsets of an ambient set U. Show that
[ \
Xi = X i.
i∈I i∈I

32
 ⊆
S T
Solution of Problem 7.15. We first prove the inclusion i∈I Xi i∈I X i .

= 0/ then the inclusion holds, otherwise let z ∈ Then z ∈

S S S
If i∈I Xi i∈I Xi . / i∈I Xi (by definition of the
complement), and it follows that z ∈
/ Xi for all i ∈ I (by definition of the union). Thus, z ∈ Xi for all i ∈ I (by
definition of the complement), which means that z ∈
T
i∈I X i (by definition of the intersection).

= 0/ then the inclusion holds, otherwise let z ∈ Then z ∈ Xi

T T
For the reverse inclusion, if i∈I X i i∈I X i .

for all i ∈ I (by definition of the intersection), and thus z ∈

/ Xi for all i ∈ I (by definition of the complement). It
follows that z ∈ (by definition of the union), and hence z ∈
S S
/ i∈I Xi i∈I Xi (by definition of the complement).
S T
Therefore, it follows from the definition of equality between sets that i∈I Xi = i∈I X i .

7.4 More problems

√
Problem 7.16. Let A = {x + y 2 | x, y ∈ Q} ⊆ R.
√
(a) Prove that for all x, y ∈ Q, x + y 2 = 0 if and only if x = y = 0.

z1
(b) Prove that for all z1 , z2 ∈ A, z1 + z2 , z1 z2 ∈ A and , for z2 ̸= 0, z2 ∈ A.
√ √
Solution of Problem 7.16. (a) Clearly if x = y = 0 then x+y 2 = 0. Conversely, assume x+y 2 = 0. Then
√ √ √
x = −y 2. If y ̸= 0 then 2 = − xy ∈ Q, a contradiction. Thus y = 0, and therefore 0 = x + y 2 = x.
√
(b) Suppose z1 , z2 ∈ A. Then there exist x1 , x2 , y1 , y2 ∈ Q such that zi = xi + yi 2 for i = 1, 2. Then we have
√
z1 + z2 = (x1 + x2 ) + (y1 + y2 ) 2 ∈ A,

and
√
z1 z2 = (x1 x2 + 2y1 y2 ) + (x1 y2 + x2 y1 ) 2 ∈ A.

Now assume z2 ̸= 0. Then we have

√
z1 x1 + y1 2
= √
z2 x2 + y2 2
√ √
x1 + y1 2 x2 − y2 2
= √ · √
x2 + y2 2 x2 − y2 2
x1 x2 − 4y1 y2 x2 y1 − x1 y2 √
= + 2 2.
x22 + 4y22 x2 + 4y22
x1 x2 −4y1 y2 x2 y1 −x1 y2 z1
Since x22 +4y22
, x2 +4y2 ∈ Q, it follows that z2 ∈ A.
2 2

33
 n=1 is increasing, or an ascending chain, if X1 ⊆ X2 ⊆
Problem 7.17. We say that the sequence of sets (Xn )∞
X3 ⊆ · · · ⊆ Xn ⊆ Xn+1 ⊆ . . . . Formally, (Xn )∞
n=1 is increasing if

(∀n ∈ N)[Xn ⊆ Xn+1 ].

Show that the sequence of sets (Xn )∞

n=1 is increasing if and only if

(∀n ∈ N)(∀k ∈ N)[(n ⩽ k) =⇒ (Xn ⊆ Xk )].

Hint. Your goal is to show that

(∀r ∈ N)[Xr ⊆ Xr+1 ] ⇐⇒ (∀k ∈ N)(∀n ∈ N)[(1 ⩽ n ⩽ k) =⇒ (Xn ⊆ Xk )]

For the implication ⇐= simply put n = r and k = r + 1. For the implication =⇒ you need to assume
that (∀r ∈ N)[Xr ⊆ Xr+1 ] and show by induction on k that ∀k ∈ N P(k) is true, where P(k) is the predicate
(∀n ∈ N)[(1 ⩽ n ⩽ k) =⇒ (Xn ⊆ Xk )]. For the base case observe that P(1) is simply X1 ⊆ X1 . For the
inductive step P(k + 1) is the predicate (∀n ∈ N)[(1 ⩽ n ⩽ k + 1) =⇒ (Xn ⊆ Xk+1 )] and you need to
distinguish two cases: either 1 ⩽ n ⩽ k and you can use the induction hypothesis together with the other
assumption, or n = k + 1

n=1 is decreasing, or a descending chain, if X1 ⊇ X2 ⊇

Problem 7.18. We say that the sequence of sets (Xn )∞
X3 ⊇ · · · ⊇ Xn ⊇ Xn+1 ⊇ . . . . Formally, (Xn )∞
n=1 is increasing if

(∀n ∈ N)[Xn ⊆ Xn+1 ].

n=1 is decreasing if and only if for all n, k ∈ N if n ⩽ k then Xn ⊇ Xk .

Show that the sequence of sets (Xn )∞

Hint. Your goal is to show that

(∀r ∈ N)[Xr ⊇ Xr+1 ] ⇐⇒ (∀k ∈ N)(∀n ∈ N)[(1 ⩽ n ⩽ k) =⇒ (Xn ⊇ Xk )]

For the implication ⇐= simply put n = r and k = r + 1. For the implication =⇒ you need to assume
that (∀r ∈ N)[Xr ⊇ Xr+1 ] and show by induction on k that ∀k ∈ N P(k) is true, where P(k) is the predicate
(∀n ∈ N)[(1 ⩽ n ⩽ k) =⇒ (Xn ⊇ Xk )]. For the base case observe that P(1) is simply X1 ⊇ X1 . For the
inductive step P(k + 1) is the predicate (∀n ∈ N)[(1 ⩽ n ⩽ k + 1) =⇒ (Xn ⊇ Xk+1 )] and you need to
distinguish two cases: either 1 ⩽ n ⩽ k and you can use the induction hypothesis together with the other
assumption, or n = k + 1

34
Problem 7.19. Let X and Y be subsets of a universal set U. Show that X ∩Y = X ∪Y .

Solution of Problem 7.19. We first prove the inclusion X ∩Y ⊆ X ∪Y . If X ∩Y = 0/ then the inclusion holds,
otherwise let z ∈ X ∩Y . Then z ∈
/ X ∩Y (by definition of the complement), and it follows that z ∈
/ X or z ∈
/Y
(by definition of the intersection). Thus, z ∈ X or z ∈ Y (by definition of the complement), which means that
z ∈ X ∪Y (by definition of the union). We just proved that X ∩Y ⊆ X ∪Y .

For the reverse inclusion, if X ∪Y = 0/ then the inclusion holds, otherwise let z ∈ X ∪Y . Then z ∈ X or
z ∈ Y (by definition of the union), and thus z ∈
/ X or z ∈
/ Y (by definition of the complement). It follows that
z∈
/ X ∩ Y (by definition of the intersection), and hence z ∈ X ∩Y (by definition of the complement). This
shows the reverse inclusion.

8 Functions

8.1 Composition

Problem 8.1. Let f , g : R → R be defined for all x ∈ R as f (x) = x2 − 3x and g(x) = 5x − 2.

1. Is it possible to define f ◦ g? If it is, what is f ◦ g.

2. Is it possible to define g ◦ f ? If it is, what is g ◦ f .

3. Are f ◦ g and g ◦ f equal? (Justify your answer)

Solution of Problem 8.1. 1. It is possible to define f ◦ g : R → R and for all x ∈ R

f ◦ g(x) = f (g(x)) = (g(x))2 − 3(g(x)) = (5x − 2)2 − 3(5x − 2) = 25x2 − 35x + 10.

2. It is possible to define g ◦ f : R → R and for all x ∈ R

(g ◦ f )(x) = g( f (x)) = 5( f (x)) − 2 = 5(x2 − 3x) − 2 = 5x2 − 15x − 2.

3. Let x = 0 then (g ◦ f )(0) = −2 ̸= 10 = ( f ◦ g)(0) and thus f ◦ g ̸= g ◦ f .

Problem 8.2. Let f , g : Z → Z be defined for all n ∈ Z as f (n) = 2n + 3 and


 2n − 1 if n is even,
g(n) =
 n + 1 if n is odd.

35
 1. Is it possible to define f ◦ g? If it is, what is f ◦ g.

2. Is it possible to define g ◦ f ? If it is, what is g ◦ f .

3. Are f ◦ g and g ◦ f equal? (Justify your answer)

Solution of Problem 8.2. 1. It is possible to define f ◦ g : Z → Z and for all n ∈ Z, since we have f (2n −
1) = 4n + 1 and f (n + 1) = 2n + 5, it follows that

 4n + 1 if n is even,
( f ◦ g)(n) =
 2n + 5 if n is odd.

2. Since f (n) is odd for any n ∈ Z, we have

(g ◦ f )(n) = f (n) + 1 = (2n + 3) + 1 = 2n + 4.

3. Let n = 0 then n is even and (g ◦ f )(0) = 4 ̸= 1 = ( f ◦ g)(0). Therefore, g ◦ f ̸= f ◦ g.

8.2 Injectivity, surjectivity, bijectivity

Problem 8.3. For f : R → R defined by f (x) = x + |x|, determine if:

1. f is injective,

2. f is surjective,

3. f is bijective.

Solution of Problem 8.3. First, note that


 2x : x ⩾ 0,
f (x) =
 0 : x < 0.

1. f is not injective, since f (−1) = 0 = f (0).

2. f is not surjective, since y ∈ R is not in the range of f whenever y < 0. Indeed, for any x ∈ R we have
x + |x| ⩾ x + (−x) = 0, so f (x) ⩾ 0.

3. f is not bijective because it is not injective (nor surjective).

36
8.3 Composition and injectivity/surjectivity

Problem 8.4. Let W, X,Y be nonempty sets. Let f : W → X, g : X → Y be functions. Show that if g ◦ f is
surjective, then g is surjective.

Solution of Problem 8.4. Fix y ∈ Y . Since g ◦ f is surjective, there is some w ∈ W such that (g ◦ f )(w) = y.
Put x = f (w). Then x ∈ X and g(x) = g( f (w)) = (g ◦ f )(w) = y. Therefore, g is surjective.

Problem 8.5. Let W, X,Y be nonempty sets. Let f : W → X, g : X → Y be functions. Show that if g ◦ f is
injective, then f is injective.

Solution of Problem 8.5. Assume that g ◦ f is injective. Let w1 , w2 ∈ W such that f (w1 ) = f (w2 ). Since g
is a function one has g( f (w1 )) = g( f (w2 )) and (g ◦ f )(w1 ) = (g ◦ f )(w2 ) (by definition of the composition).
Since g ◦ f is injective it implies that w1 = w2 , and f is injective.

Problem 8.6. Let X and Y be nonempty sets and let f : X → Y be a function. Prove that f is injective if and
only if for all sets Z, for all functions h : Z → X and k : Z → X, if f ◦ h = f ◦ k, then h = k.

Solution of Problem 8.6. First suppose f is injective. Let Z be a set, and let h, k : Z → X be functions such
that f ◦ h = f ◦ k. Given z ∈ Z, since f (h(z)) = f (k(z)), it follows that h(z) = k(z). Thus h = k.
Conversely assume f is not injective. Then there exist x1 , x2 ∈ X such that x1 ̸= x2 while f (x1 ) = f (x2 ).
Define h, k : X → X by h(x) = x1 and k(x) = x2 for all x ∈ X. Then for all x ∈ X, f (h(x)) = f (x1 ) = f (x2 ) =
f (k(x)), so f ◦ h = f ◦ k, but h ̸= k.

Problem 8.7. Let X and Y be nonempty sets and let f : X → Y be a function. Prove that f is surjective if
and only if for all sets Z, for all functions h : Y → Z and k : Y → Z, if h ◦ f = k ◦ f , then h = k.

Solution of Problem 8.7. First assume f is surjective. Let Z be a set, and let h, k : Y → Z be functions such
that h ◦ f = k ◦ f . Given y ∈ Y , there is some x ∈ X such that y = f (x), and thus h(y) = h( f (x)) = k( f (x)) =
k(y). Therefore, h = k.
Conversely, assume f is not surjective. Then there is some y0 ∈ Y such that the set {x ∈ X : f (x) = y0 }
is empty. Put h = iY , and let k : Y → Y be defined by k(y) = y if y ̸= y0 and k(y0 ) = y1 for some y1 ∈ Y with
y0 ̸= y1 . Then h ̸= k since h(y0 ) = y0 ̸= y1 = k(y0 ), but h ◦ f = k ◦ f .

9 Injectivity, surjectivity, and one-sided invertibility

Problem 9.1. Let X and Y be nonempty sets and f : X → Y be a function. We say that f is left-invertible
(or admits a left-inverse) if there exists a function g : Y → X such that g ◦ f = iX . Prove that f is injective if

37
and only if f is left-invertible.

Solution of Problem 9.1. Suppose first that f is left-invertible. Then there is some function g : Y → X such
that g ◦ f = iX . Let x1 , x2 ∈ X such that f (x1 ) = f (x2 ), then g( f (x1 )) = g( f (x2 )) (since g is a function),
and thus (g ◦ f )(x1 ) = (g ◦ f )(x2 ) (by definition of the composition). It follows from the assumption that
iX (x1 ) = iX (x2 ) and hence x1 = x2 (by definition of the identity function on X). Therefore f is injective.
Conversely, assume f is injective. Define a function g : Y → X as follows: if y = f (x) for some (and
hence only one by injectivity) x ∈ X, put g(y) = x, and otherwise define g(y) arbitrarily. Then for each x ∈ X,
let y = f (x). Then g(y) = x, that is, g( f (x)) = x and hence (g ◦ f )(x) = x (by definition of the composition).
By definition of the identity function (g◦ f )(x) = iX (x) and thus g◦ f = iX . Therefore f is left-invertible.

Problem 9.2. Let X and Y be nonempty sets, and f : X → Y be a function. We say that f is right-invertible
(or admits a right-inverse) if there exists a function g : Y → X such that f ◦ g = iY . Prove that if f has a
right-inverse then f is surjective.

Solution of Problem 9.2. Suppose first that f is right-invertible. Then there is some function h : Y → X such
that f ◦ h = iY . Let y ∈ Y , then

y = iY (y) (by definition of the identity function on Y )

= ( f ◦ h)(y) (since f ◦ h(y) = iY (y) by definition of h being a right-inverse of f )

= f (h(y)) (by definition of the composition).

If we let x = h(y) then x ∈ X (since the codomain of h is X) and y = f (x). We just proved that for all y ∈ Y ,
there is x ∈ X such that y = f (x), which means that f is surjective.

10 Functions and sets

Problem 10.1. Let X and Y be nonempty sets, and f : X → Y be an injective function. Let A be a subset of
X. Prove that f −1 ( f (A)) = A.

Solution of Problem 10.1. The result is proved by a double inclusion argument. We first prove that f −1 ( f (A)) ⊆
A. If f −1 ( f (A)) = 0/ then the inclusion holds. Otherwise let x ∈ f −1 ( f (A)), then f (x) ∈ f (A) (by definition
of the inverse image of a subset), and there exists a ∈ A such that f (x) = f (a) (by definition of the image of
a subset). Since f is injective it follows that x = a, and hence x ∈ A (because a ∈ A).

We now prove that A ⊆ f −1 ( f (A)). If A = 0/ the inclusion holds. Otherwise, let x ∈ A, then f (x) ∈ f (A)
(by definition of the image of a subset) and x ∈ f −1 ( f (A)) (by definition of the inverse image of a subset).

38
Problem 10.2. Let X and Y be nonempty sets, and f : X → Y be an surjective function. Let A be a subset of
Y . Prove that f ( f −1 (A)) = A.

Solution of Problem 10.2. The result is proved by a double inclusion argument. We first prove that f ( f −1 (A)) ⊆
A. If f ( f −1 (A)) = 0/ the inclusion holds. Otherwise let y ∈ f ( f −1 (A)), then y = f (x) for some x ∈ f −1 (A)
(by definition of the image), and f (x) ∈ A (by definition of the inverse image). But y = f (x) belongs to A
since f (x) does. Therefore f −1 ( f (A)) ⊆ A.

We now prove that A ⊆ f ( f −1 (A)). If A = 0/ the inclusion holds. Otherwise let a ∈ A, then a ∈ Y since
A is a subset of Y . By surjectivity of f , there exists x ∈ X such that a = f (x), and f (x) ∈ A (since a is in A).
It follows that x ∈ f −1 (A) (by definition of the inverse image) and f (x) ∈ f ( f −1 (A)) (by definition of the
image). Therefore a ∈ f ( f −1 (A)).

11 Supplementary problems

Problem 11.1. Let f1 : X1 → X2 , f2 : X2 → X3 , f3 : X3 → X4 and f4 : X4 → X5 . Show that (( f4 ◦ f3 )◦ f2 )◦ f1 =

f4 ◦ ( f3 ◦ ( f2 ◦ f1 )).

Problem 11.2. Let X and Y be nonempty sets, and f : X → Y be a function. Prove that f is surjective then
f is right-invertible.

Solution of Problem 11.3. Assume f is surjective. For each y ∈ Y , the set {x ∈ X : f (x) = y} is non-empty.
Note that if y1 ̸= y2 , then {x ∈ X : f (x) = y1 } ∩ {x ∈ X : f (x) = y2 } = ∅. By the axiom of choice, there is a
function h : Y → X such that for each y ∈ Y , h(y) ∈ {x ∈ X : f (x) = y}. Hence f (h(y)) = y for each y ∈ Y ,
so f ◦ h = iY and therefore f is right-invertible.

Problem 11.3. Let f1 : X1 → X2 , f2 : X2 → X3 , f3 : X3 → X4 be three injective functions. Show that f3 ◦ f2 ◦ f1

is injective.

39