Math 140a - HW 1 Solutions

Problem 1 (WR Ch 1 #1). If r is rational (r 6= 0) and x is irrational, prove that r + x and

rx are irrational.
a
Solution. Given that r is rational, we can write r = b for some integers a and b. We are
also given that x is irrational. From here, we proceed with a proof by contradiction. We
first assume that r + x is rational, and then we use this fact in some way to show that x is
rational, contradicting one of the facts we were given. This will prove that r + x is instead
irrational.
c
So if r + x is rational, we can write r + x = d for some relatively prime integers c and d.
But then
c c a bc − ad
x= −r = − = ,
d d b bd
and thus x is rational, which is a contradiction. Therefore, r + x is irrational.
Next, we prove that rx is irrational using a similar contradiction proof. Assume that rx
c
is rational. Then we can write rx = d for some integers c and d. But then
c c bc
x= = a = ,
rd b d ad
and thus x is rational, which is a contradiction. Therefore, rx is irrational.

Problem 2 (WR Ch 1 #2). Prove that there is no rational number whose square is 12.
a
Solution. Let x be a rational number such that x2 = 12. Then we can write x = b, and
furthermore, we can choose a and b to be relatively prime (which means there is no prime
a
number dividing both a and b), so that the fraction b is written in lowest terms. With a
little algebraic manipulation,
a2
12 = x2 = =⇒ 12b2 = a2 .
b2
Now, the prime factorization of 12 is 22 · 31 , so since there is an odd number of factors of
3 (just 1), we’ll concentrate on 3 and how it divides both sides of the equation 12b2 = a2 .
Notice that 3 divides the left side since it has a factor of 12. Therefore, 3 must divide the
right side of the equation, a2 . From here, the crucial step is realizing that if 3 divides a2 ,
then it must also divide a. This is because if we factor a2 into its prime factors, saying that
3 divides a2 is equivalent to saying that 3 is one of those prime factors, but a square of an
integer must have an even number of each factor (it can’t have just 1), so that means 32 must
divide a2 , and 3 must divide a.
Since we have shown that 32 divides the right side, 32 must divide the left side, but there
is only one factor of 3 in 12, so that means 3 divides b2 . Using the same logic as before, this
means that 3 must divide b.

1
 Therefore, we have shown that 3 divides both a and b, but this contradicts the fact that
a
we already chose a and b to be relatively prime (so that b would be expressed in lowest
terms). Since our initial assumption leads to a contradiction, we have instead that there is
no rational number whose square is 12.

Problem 3 (WR Ch 1 #7). Fix b > 1, y > 0, and prove that there is a unique real x such
that bx = y, by completing the following outline. (This is called the logarithm of y to the
base b.)

(a) For any positive integer n, bn − 1 ≥ n(b − 1).

Solution. First, we factorize the left hand side:

bn − 1 = (b − 1)(bn−1 + bn−2 + · · · + b2 + b + 1).

Then, since b > 1, we know that bn−1 + bn−2 + · · · + b2 + b + 1 ≥ n. So

bn − 1 = (b − 1)(bn−1 + bn−2 + · · · + b2 + b + 1) ≥ (b − 1)n.

(b) Hence b − 1 ≥ n(b1/n − 1).

Solution. If b > 1 then b1/n > 1, so since we proved that bn − 1 ≥ n(b − 1) for any
b > 1, we can substitute b1/n for b in that equation to get that b − 1 ≥ n(b1/n − 1).

(c) If t > 1 and n > (b − 1)/(t − 1), then b1/n < t.

Solution. n > (b − 1)/(t − 1) implies that n(t − 1) > (b − 1). Using the previous result
for the second inequality,

n(t − 1) > (b − 1) ≥ n(b1/n − 1).

Therefore, n(t − 1) > n(b1/n − 1). Dividing by n 6= 0 we have that

t − 1 > b1/n − 1.

Then we add 1 to both sides to get the result.

(d) If w is such that bw < y, then bw+(1/n) < y for sufficiently large n; to see this, apply
part (c) with t = y · b−w .
Solution. First of all, bw < y implies that yb−w > 1. Therefore, for t = yb−w , we have
b−1
t > 1, so if we also choose some n > yb−w −1 we can then use part (c) to get that

b1/n < t =⇒ b1/n < yb−w =⇒ bw+(1/n) < y.

2
 (e) If bw > y, then bw−(1/n) > y for sufficiently large n.
Solution. First of all, bw > y implies that y −1 bw > 1. Therefore, for t = y −1 bw , we
b−1
have t > 1, so if we also choose some n > y −1 bw −1 we can then use part (c) to get that

b1/n < t =⇒ b1/n < y −1 bw =⇒ y < bw−(1/n) .

(f ) Let A be the set of all w such that bw < y, and show that x = sup A satisfies bx = y.
Solution. To show that bx = y, we will first show that bx is not greater than y and then
show that it is not less than y.
Assume (by way of contradiction) that bx > y. Then by part (e), there is some integer
n such that bx−(1/n) > y. However, this means that x − (1/n) is an upper bound for A,
but since x − (1/n) < x, this means that x is not the least upper bound, a contradiction
of the definition of x as the supremum of A. Therefore, instead we have that bx is not
greater than y, or equivalently, bx ≤ y.
Next assume (by way of contradiction) that bx < y. Then by part (d), there is some
integer n such that bx+(1/n) < y. However, this means that x + (1/n) ∈ A, but
x < x + (1/n), which means that x is not an upper bound, a contradiction of the
definition of x as the supremum of A. Therefore, instead we have that bx is not less
than y, or equivalently, bx ≥ y.

(g) Prove that this x is unique.

Solution. Assume there is some other z ∈ R such that y = bz . Then

bx = y = bz .

If we assume that x 6= z, then either x > z or z > x. In the first case, we divide both
sides of the above equation by bz to get bx−z = 1 (and note that x − z is a positive real
number). However, b > 1, so that bw > 1 for any positive real number w, contradicting
the fact that bx−z = 1. In the second case, we divide both sides of the above equation
by bx to get bz−x = 1 (and note that z − x is a positive real number). However, b > 1,
so that bw > 1 for any positive real number w, contradicting the fact that bx−z = 1.

Problem 4 (WR Ch 1 #9). Suppose z = a + bi, w = c + di. Define z < w if a < c, and also
if a = c but b < d. Prove that this turns the set of all complex numbers into an ordered set
(This type of order relation is called a dictionary order, or lexicographic order, for obvious
reasons.) Does this ordered set have the least-upper-bound property?

Solution. To prove that “<” is an order on the set of all complex numbers, we need to check
the two axioms of an ordered set.

3
 The first is that for any z, w ∈ C, one and only one of the statements z < w, z = w, or
z > w is true. To show that this is true, we need to break the proof up into cases using the
first axiom of the usual ordering for real numbers:

Case 1 : a < c. This results in z < w.

Case 2 : a = c. Here we are again essentially using the first axiom of the usual ordering for
real numbers. We must have exactly one of the following cases: b < d, b = d, or b > d.
From here, b < d if and only if z < w; b = d if and only if z = w; and b > d if and only
if z > w.

Case 3 : a > c. This results in z > w.

Next we prove the second axiom, which is that for any z, y, w ∈ C,

z < y and y < w =⇒ z < w.

For notation, let y = e + f i. Then if z < y we know that either a < e or a = e and b < f .
Likewise, if y < w we know that either e < c or e = c and f < d. This give us four possible
cases to check:

Case 1 : a < e and e < c. By the transitive property of the usual ordering for the real numbers,
we then know that a < c, which implies that z < w.

Case 2 : a < e and e = c. Then a < c, which implies that z < w.

Case 3 : a = e and e < c. Then a < c, which implies that z < w.

Case 4 : a = e and e = c, with b < f and f < d. Then by the transitive property of the usual
ordering for the real numbers, we then know that b < d, which means that a = c and
b < d, so that z < w.

Lastly, in order to show that C doesn’t have the least upper bound property, we need to
find a subset of C that has no least upper bound. There are plenty, but let’s consider the set

A = {z ∈ C|z = a + bi for some a, b ∈ R, and a < 0}.

Assume there is a least upper bound w = c + di. Then if c > 0, we know that 0 + di < w,
and 0 + di is an upper bound for A, contradicting the fact that w is a least upper bound. If
c < 0, then c/2 + di is a complex number with a negative real part, so c/2 + di is in A and
c + di < c/2 + di, contradicting the fact that w is an upper bound. Therefore c = 0, and thus
w is of the form di for some d ∈ R.
However, for any choice of d ∈ R, the complex number (d − 1)i is an upper bound for A
which is less than di under the order, contradicting the fact that w is a a least upper bound
once again. Hence, A has no least upper bound, and more generally C does not have the
least upper bound property with the order described above.

4
Problem 5 (WR Ch 1 #17). Prove that

|x + y|2 + |x − y|2 = 2|x|2 + 2|y|2

if x ∈ Rk and y ∈ Rk . Interpret this geometrically, as a statement about parallelograms.

Solution. Let x = (x1 , . . . , xn ) and y = (y1 , . . . , yn ). Then

p 2 p 2
|x + y|2 + |x − y|2 = (x1 + y1 )2 + · · · + (xn + yn )2 + (x1 − y1 )2 + · · · + (xn − yn )2

= (x1 + y1 )2 + · · · + (xn + yn )2 + (x1 − y1 )2 + · · · + (xn − yn )2

= (x21 + 2x1 y1 + y12 ) + · · · + (x2n + 2xn yn + yn2 ) + (x21 − 2x1 y1 + y12 ) + · · · + (x2n − 2xn yn + yn2 )
= (x21 + 
2x  2 2
1 y1 + y1 ) + · · · + (xn + 
2x  2 2
n yn + yn ) + (x1 − 
2 2
1 y1 + y1 ) + · · · + (xn − 
2x  2x 2
n yn + yn )


= 2(x21 + · · · + x2n ) + 2(y12 + · · · + yn2 )

q 2 q 2
=2 x21 + · · · + x2n + 2 y12 + · · · + yn2

= 2|x|2 + 2|y|2 .

Interpreted geometrically, the statement simply says that:

The sum of the squares of the diagonals of a parallelogram is equal to twice the
sum of the squares of the sides.