18.

014 Problem Set 1 Solutions

Total: 24 points

Problem 1: If ab = 0, then a = 0 or b = 0.

Solution (4 points)
Suppose ab = 0 and b = � 0. By axiom 6, there exists a real number y such that
by = 1. Hence, we have

a = 1 · a = a · 1 = a(by) = (ab)y = 0 · y = 0

using axiom 4, axiom 1, axiom 2, and Thm. I.6. We conclude that a and b cannot
both be non-zero; thus, a = 0 or b = 0.

Problem 2: If a < c and b < d, then a + b < c + d.

Solution (4 points) By Theorem I.18, a + b < c + b and b + c < d + c. By the

commutative axiom for addition, we know that c+b = b+c, d+c = c+d. Therefore,
a + b < c + b, c + b < c + d. By Theorem I.17, a + b < c + d.

� �
Problem 3: For all real numbers x and y, �|x| − |y|� ≤ |x − y|.

Solution (4 points)
By part (i) of this exercise, |x|−|y| ≤ |x−y|. Now notice that −(|x|−|y|) = |y|−|x|.
By definition of the absolute value, either ||x|−|y|| = |x|−|y| or ||x|−|y|| = |y|−|x|.
In the first case, by part (i) of this problem, we see that ||x| − |y|| ≤ |x − y|. In
the second case, we can interchange the x and y from part (i) to get ||x| − |y|| =
|y| − |x| ≤ |y − x| = |x − y|, where the last equality comes from part (c) of this
problem. Thus, ||x| − |y|| ≤ |x − y|.

Problem 4: Let P be the set of positive integers. If n, m ∈ P , then nm ∈ P .

Solution (4 points)

1
Fix n ∈ P . We show by induction on m that nm ∈ P for all m ∈ P .
First, we check the base case. If m = 1, then

nm = n · 1 = 1 · n = n ∈ P

by axiom 4, axiom 1, and the hypothesis n ∈ P .

Next, we assume the statement for m = k and we prove it for m = k + 1. Assume
nk ∈ P . By theorem 5 of the course notes, nk+n ∈ P . By axiom 3, nk+n = n(k+1);
thus, n(k + 1) ∈ P and our induction is complete.

Problem 5: Let a, b ∈ R be real numbers and let n ∈ P be a positive integer. Then

an · bn = (a · b)n .

Solution (4 points)
Fix a, b ∈ R. We prove the statement by induction on n.
First, we must check the statement for n = 1. In that case, we must show a1 · b1 =
(a · b)1 . By the definition of exponents, we know a1 = a, b1 = b, and (a · b)1 = a · b
so our statement becomes the tautology a · b = a · b.
Next, we check the inductive step. Assume the statement is true for n = k; we must
prove it for n = k + 1.
Notice that (ab)k+1 = (ab)k · (ab)1 = ak · bk · a1 · b1 by Theorem 10 from the course
notes and the induction hypothesis. As ak · bk · a1 · b1 = ak · a1 · bk · b1 = ak+1 · bk+1
by commutativity and Theorem 10, we see that the statement holds for n = k + 1.

Problem 6: Let a and h be real numbers, and let m be a positive integer. Show
by induction that if a and a + h are positive, then (a + h)m ≥ am + mam−1 h.

Solution (4 points)
The first step is to prove the statement for m = 1. In this case (a + h)m = (a + h)1 =
a + h by the definition of exponents and

am + mam−1 h = a1 + 1 · a(1−1) h = a + a0 h = a + 1 · h = a + h

where the second to last inequality used the definition a0 = 1. Hence, for m = 1, we
have (a + h)m = am + mam−1 h, which in particular implies (a + h)m ≤ am + mam−1 h
by the definition of ≤.
Next, we assume the statement for m = k, and then we prove it for m = k + 1.
Thus, we assume (a + h)k ≥ ak + kak−1 h, which means (a + h)k = ak + kak−1 or

2
(a + h)k > ak + kak−1 h. In the first case, we can multiply both sides by (a + h) to
get (a + h)k · (a + h) = (ak + kak−1 )(a + h). In the second case, we can use Thm. I.19
and the fact that a + h > 0 to conclude (a + h)k · (a + h) > (ak + kak−1 h) · (a + h).
Thus, by the definition of exponents and the definition of ≥, we have
(a + h)k+1 = (a + h)k · (a + h) ≥ (ak + kak−1 h) · (a + h) = ak+1 + (k + 1)ak h + kak−1 h2 .
To finish the proof we need a lemma.

Lemma: If a is a positive real number, then al is positive for all positive inte­
gers l.

The proof is by induction on l. When l = 1, we know al = a1 = a using the

definition of exponents. However, a is positive by hypothesis so the base case is
true.
Now we assume the result for l and prove it for l + 1. Note al+1 = al · a by the def­
inition of exponents. Further, by the problem 5, al+1 is positive since al is positive
by the induction hypothesis and a is positive by the hypothesis of the lemma. The
lemma follows.

Now, back to our proof. By the lemma, we know ak−1 is positive since k − 1 is
a positive integer and a is positive. Moreover, all positive integers are positive, as is
remarked in the course notes; thus, kak−1 is positive by problem 5. Now, if h = 0,
then h2 = 0 by Thm. I.6; hence, (kak−1 )h2 = 0 again by Thm. I.6. Putting this
together with the expression (*) above yields
(a + h)k+1 ≥ ak+1 + (k + 1)ak h + kak−1 h2 = ak+1 + (k + 1)ak h.
On the other hand, if h �= 0, then h2 > 0 by Thm. I.20; hence, kak−1 h2 > 0
by Thm. I.19. Adding ak+1 + (k + 1)ak h to both sides (using Thm. I.18) yields
ak+1 + (k + 1)ak h < ak+1 + (k + 1)ak h + kak−1 h2 . Combining this with (*) and
applying the transitive property (Thm. I.1.7) implies (a + h)k+1 > ak+1 + (k + 1)ak h.
In particular,
(a + h)k+1 ≥ ak+1 + (k + 1)ak h.
Thus, regardless of whether h = 0 or h �= 0, we have proved the statement for
m = k + 1. The claim follows.

Bonus: If x1 , . . . , xn are positive real numbers, define

x1 + · · · + xn
An = , Gn = (x1 · · · xn )1/n .
n

3
(a) Prove that Gn ≤ An for n = 2.
(b) Use induction to show Gn ≤ An for any n = 2k where k is a positive integer.
(c) Now show Gn ≤ An for any positive integer n.

Solution (4 points)
Because this is a bonus problem, this solution is a bit less rigorous than the others.
However, you should be able to fill in all of the details on your own.
(a) Note (x1 − x2 )2 ≥ 0. Expanding, we get

(x1 − x2 )2 = x21 + x22 − 2x1 x2 = (x1 + x2 )2 − 4x1 x2

is also positive. Hence, (x1 + x2 )2 ≥ 4x1 x2 . Dividing by 4 and taking square roots,
we get
x1 + x2
≥ (x1 x2 )1/2 .
2
(b) We prove this part by induction on k. The base case k = 1 was done in part a.
Now, we assume Gn ≤ An for n = 2k , and we prove it for n = 2k+1 . The inductive
hypothesis tells us that
x1 + · · · + x2k k
Y1 = k
≥ (x1 · · · x2k )1/2
2
and
x2k +1 + · · · + x2k+1 1/2k
Y2 = ≥ (x 2k +1 · · · x2k+1 ) .
2k
Using part (a), we know
Y1 + Y2
≥ (Y1 Y2 )1/2 .
2
Writing this in terms of the xi , we have
x1 + · · · + x2k+1 k+1

k+1
≥ (x1 · · · x2k+1 )1/2 .
2
(c) Select a positive integer m such that 2m > n. Fix positive real numbers
x1 , . . . , xn , and let
x1 + · · · + xn
An = .
n
Now, put An = xn+1 = xn+1 = · · · = x2m . Applying part (b) for these real numbers
x1 , . . . , x2m yields

x1 + · · · + xn + (2m − n)An m m −n)/2m

m
≥ (x1 · · · xn )1/2 A(2
n .
2

4
 (2m −n)/2m n/2m
The left hand side is just An ; hence, dividing both sides by An yields An ≥
m
(x1 · · · xn )1/2 . Raising both sides to the power of 2m /n yields

An ≥ (x1 · · · xn )1/n .

This is what we wanted to show.

5
MIT OpenCourseWare
http://ocw.mit.edu

18.014 Calculus with Theory

Fall 2010

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.