James Stankowicz - 132A - Due: Jan.

16, 2013

Homework #1 Solutions
Problem 1
Write each of the following complex numbers in the form x + iy .

Solution 1
Part a.
 √ 3
1 + 2i =
√ √ 2 √ 3
= 1 + 3 2i + 3 2i + 2i =
1 3
= 1 + 3 · 2 2 i − 6 − i2 2 =
1
= −5 + 2 2 (3i − i2) =
√
= −5 + i 2 .
Note: the rst line used the binomial expansion:
n  
n
X n
(a + b) = xn−k y k ,
k
k=0
√
3
with n = 3. 1 + 2i by hand:
It's also possible (although longer) to expand
 √ 3  √  √ 2
1 + 2i = 1 + 2i 1 + 2i + · · · .

Part b.
2
(1 − i)
=
(1 + i)
3
(1 − i) 1 3
= = (1 − i) =
(1 + i) (1 − i) 2
1 2 3

= 1 + 3 (−i) + 3 (−i) + (−i) =
2
1
1 − 3i − 3 − i3 =


=
2
1
= (−2 − 2i) = −1 − i .
2
Part c.
√
3 √
First, recognize the result from part a where 1 + 2i = −5 + i 2. Then:
√ √
√ 3 √ 



 2−i 2−i
1 + 2i 2 = −5 + i 2
2 =
(1 + i) (1 + i)
√
√
√ √
−5 + i 2 2−i −5 2 + 5i + i2 + 2
= = =
1 + 2i + i2 2i
1  √ 
= −4 2 + 7i =
2i
√
= 72 + i2 2 .

Problem 2 [BC 5#11]

Solve the equation z2 + z + 1 = 0 for z = (x, y) by writing

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

and then solving a pair of simultaneous equations in x and y. Suggestion: Use the fact that no real number x satises
the given equation to show that y 6= 0.
1
James Stankowicz 132 Due: Jan. 16, 2014

Solution 2
By expanding.
(0, 0) = (x, y) (x, y) + (x, y) + (1, 0) =
= x2 − y 2 , 2xy + (x + 1, y) =

= x2 − y 2 + x + 1, 2xy + y ,

arrive at the two equations

0 = 2xy + y,
0 = x2 − y 2 + x + 1.
Solve the rst for x:
2xy = −y ⇒ x = − 21 .
The second is then straightforward using the solution for x:
 2  
1 1
0= − − y2 + − +1=
2 2
1 1
= − y2 − + 1 ⇒
4 2
3 √
y 2 = ⇒ y = ± 23 .
4
Compiling the answer:
 √
3
 √

z = (x, y) = − 12 , ± 2 = 1
2 −1 + 3i .

Problem 3 [BC 13#4]

√
Verify 2z ≥ |Re {z}| + |Im {z}|.
2
Suggestion: Reduce this inequality to (|x| − |y|) ≥ 0.

Solution 3
√
The goal is to verify 2 |z| ≥ |Re {z}| + |Im {z}|. Let
q z = x + iy, for x, y ∈ R. This means |Re {z}| = |x| and
2 2
p
|Im {z}| = |y|. The modulus of z is |z| = 2 2
x + y = |x| + |y| . There is nothing profound about the last equality;
it only says that some number squared is the same as squaring the absolute value of the number. Make sure this makes
sense to you!The goal is now to show
√
2 |z| ≥ |Re {z}| + |Im {z}| ⇒
√ q 2 2
2 |x| + |y| ≥ |x| + |y| .
Clean this up by squaring both sides, then see if the inequality can be manipulated into an obviously true inequality:
 
2 2 2
2 |x| + |y| ≥ (|x| + |y|) .
Rewrite the right hand side:
2 2 2
(|x| + |y|) = |x| + 2 |x| |y| + |y| .
Then the inequality is
2 2 2 2
2 |x| + 2 |y| ≥ |x| + 2 |x| |y| + |y| .
Move everything to the left side of the inequality:
2 2
|x| − 2 |x| |y| + |y| ≥ 0 ⇒
2
(|x| − |y|) ≥ 0.
This statement is true because the quantity in parentheses is always some real number, so that its square is always
non-negative.

As a rigorous mathematical treatment, this is somewhat backwards.

A formal proof would start from the statement

2
(|x| − |y|) ≥ 0,
then perform all the operations in reverse to arrive at the desired result
√
2 |z| ≥ |Re {z}| + |Im {z}| .
2
James Stankowicz 132 Due: Jan. 16, 2014

Doing so is a bit tricky, particularly when trying to go from

2 2
|x| − 2 |x| |y| + |y| ≥ 0
to
2 2 2 2
2 |x| + 2 |y| ≥ |x| + 2 |x| |y| + |y| ,
which is why the reverse engineering in this solution is easier.

Problem 4 [BC 13#5]

In each case, sketch the set of points determined by the given condition.

Solution 4

(a) |z − 1 + i| = 1 (b) |z + i| ≤ 3 (c) |z − 4 i| ≥ 4

y y y

x
r=1
(1, − i)
(0, 4 i)
r=4
(0, − i) r=3

Problem 5
Write each of the following complex numbers in the form reiθ .

Solution 5
Part a. Again utilize the result of problem 1a:

 √ 3
1+ 2i =
r
√ √ 2 √
2 ei arctan(− 2/5) =
2
= −5 + i 2 = (−5) +

√ √ √ √
= 27ei(π−arctan( 2/5))
= 3 3ei(π−arctan( 2/5)) .
√ √ 
2
The number −5 + i 2 is in quadrant II (negative x, positive y ), and forms an angle of arctan 5 with respect to
√ 
2
the negative x-axis; this gives a corresponding angle of Arg (z) = π − arctan 5 .

It is also possible to try writing

√ √ √
1+ 2i = 3ei arctan( 2)
;
this result, however, is somewhat muddled when trying to simplify, since it is not clear how to treat the
√
φ =
3 arctan 2 that appears as the argument of the exponent.
3
James Stankowicz 132 Due: Jan. 16, 2014

Part b.
  
−i i 1 1−i i (1 − i) 1
=− =− = − (1 + i) =
2 + 2i 2 1+i 1−i 2 2 4
1 1 √
(−1 − i) = · 2e−i 4 = 2− 2 e−i 4 .
3π 3 3π
=
4 4
π
The number −1 − i is in quadrant III (negative x, negative y ), and forms an angle of
4 with respect to the negative
3π
x-axis; this gives a corresponding angle of Arg (z) = − 4 . Another way to approach this problem is to convert both i
and −2 − 2i to polar form, then manipulate the numbers that way (as in problem 6 of this assignment).

Part c.
1
7 =
(1 − i)
√ −7
2e−iπ/4 = 2− 2 ei7π/4 =
−7 7
= (1 − i) =

= 2− 2 e−iπ/4 .
7

π
The number 1−i is in quadrant IV (positive x, negative y ), and forms an angle of
4 with respect to the x-axis. After
exponentiating, the 7π/4 that comes out is equivalent, modulo 2π , to −iπ/4.

Problem 6 [BC 24#5b,5d]

By writing the individual factors on the left in exponential form, performing the needed operations, and nally changing
back to rectangular coordinates, show that:

b: 5i/ (2 + i) = 1 + 2i
√
−10 √

d: 1 + 3i = 2−11 −1 + 3i .

Solution 6
Part b.
5i
=
(2 + i)

5eiπ/2 √
5ei( 2 −arctan( 2 )) =
π 1
=√ =
5ei arctan(1/2)
√ q
5ei arctan(2) = (1) + (2) ei arctan(2/1) =
2 2
=

= 1 + 2i.
π 1


Converting
2 − arctan 2 to arctan
(2) may be understood geometrically. The angle π/2 is straight up along the
y -axis. Geometrically,
 π2 − arctan 12  then says to go straight along the y -axis, and then subtract from that angle
1
the angle of arctan
2 (that's the triangle with opposite length 1 and adjacent length 2). The picture:

Want this angle!

π
θ= 2
1

x
2

arctan 1/2

4
James Stankowicz 132 Due: Jan. 16, 2014

Part d.
√ −10 π
−10
= 2ei 3 = 2−10 e−i 3 = 2−10 e−i 3 =
 10π 4π
1+i 3
√ 
= 2−11 · 2ei 3 = 2−11 −1 + i 3
2π


Note that for angles − 10π 4π 2π

3 = − 3 = 3 . These angles are equivalent mod (2π) for those who speak math. Of these
three equivalent mod (2π) angles, 3 is the only one that meets the criteria to be in Arg (z). Further, the angle 2π
2π
3
π
is in quadrant II, with angle
3 with respect to the negative x-axis. Drawing the corresponding right triangle gives
the x and y components.

Problem 7 [BC 24#9]

Establish the identity
1 − z n+1
1 + z + z2 + · · · + zn = , (z 6= 1) ,
1−z
and then use it to derive Lagrange's trigonometric identity:

1 sin ((2n + 1) θ/2)

1 + cos (θ) + cos (2θ) + · · · + cos (nθ) = + , (0 < θ < 2π) .
2 2 sin (θ/2)
. Suggestion: As for the rst identity, write S = 1 + z + z2 + · · · + zn, and consider the dierence S − zS . To derive
the second identity, write z= eiθ in the rst one.

Solution 7
Starting as hinted:
S − zS = (1 + z + · · · + z n ) − z + z 2 + · · · + z n + z n+1 =

= 1 − z n+1 ,
so that
(1 − z) S = 1 − z n+1 ⇒
1 − z n+1
S= .
1−z
By plugging in z = eiθ , the left hand side is

S = 1 + eiθ + ei(2θ) + · · · + ei(nθ) =

= 1 + (cos (θ) + i sin (θ)) + · · · + (cos (nθ) + i sin (nθ)) =
which has real part
Re {S} = 1 + cos (θ) + · · · + cos (nθ) .
This means Re {S} will be equal to the real part of the other side of the expression:

1 − z n+1 1 − ei(n+1)θ
   
Re = Re .
1−z 1 − eiθ
θ 1
eiθ/2 − e−iθ/2



To brute force the issue, attempt to write the denominator in the form sin 2 = 2i by multiplying

through in both numerator and denominator by e−iθ/2 :

1 − ei(n+1)θ
   −iθ/2 
e
Re =
1 − eiθ e−iθ/2
 −iθ/2
− ei(n+1/2)θ

e
= Re =
e−iθ/2 − eiθ/2
 −iθ/2
− ei(n+1/2)θ

e
= Re =
−2i sin (θ/2)
1 −iθ/2
− ei(n+1/2)θ
n  o
= Re i e =
2 sin (θ/2)
     
1 θ θ
= Re i cos − i sin −
2 sin (θ/2) 2 2
      
1 1
− cos n+ θ + i sin n+ θ =
2 2
   
1 1
= sin (θ/2) + sin n+ θ =
2 sin (θ/2) 2
5
James Stankowicz 132 Due: Jan. 16, 2014

1 sin((2n+1)θ/2)
= 2 + 2 sin(θ/2) .

Problem 8
Find all values of the following roots and write them in the form x + iy :
1
Part a: (32i)5
1
Part b: 83 1
Part c: (3 + 4i) 2

Solution 8
Part a.
1
(32i) 5 =
 15
= 2 ei(π/2+2nπ)

=

= 2ei( 10 + 5 πn) = 2ei(π(1+4n)/10)

π 2

Consider dierent possible n values.

2eiπ/10 = 2 cos π π




n=0 : 10 + i sin 10 ,

n=1 : 2eiπ/2 = 0 + 2i ,

2ei9π/10 = 2 cos 9π 9π




n=2 : 10 + i sin 10 ,

n=3 : 2ei13π/10 = 2ei−7π/10 =

= −2 cos 7π 7π




10 + i sin 10 ,

n=4 : 2eiπ17/10 = 2e−iπ3/10 =

= 2 cos 3π 7π




10 − i sin 10 ,

n=5 : 2eiπ21/10 = 2eiπ/10 =

π π




= 2 cos 10 + i sin 10 ,

at which point the solutions repeat cyclically.

Part b.

31
8 3 = 8ei2πn = 2ei2πn/3
1

Consider dierent possible values of n:

n=0 : 2ei2π0/3 = 2 ,
    
2π 2π
n=1 : 2ei2π/3 = 2 cos + i sin =
3 3
√
= −1 + i 3 ,
√
n=2 : 2ei4π/3 = 2e−i2π/3 = −1 − i 3 ,
at which point the solutions repeat cyclically.

Part c.
1
(3 + 4i) 2 =
 21
= 5ei(arctan(4/3)+2πn)

=
√
5ei( 2 arctan(4/3)+nπ)
1
=
6
James Stankowicz 132 Due: Jan. 16, 2014

Consider dierent possible values for n:

√
5ei( 2 arctan ( 3 )) =
1 4
n=0 :
√
      
1 4 1 4
= 5 cos arctan + i sin arctan ,
2 3 2 3
√ i 1 arctan 4 +π √
n=1 : 5e ( 2 ( 3 ) ) = 5eiπ e 2i arctan ( 34 ) =
√
      
1 4 1 4
= 5 − sin arctan + i cos arctan ,
2 3 2 3
at which point the solutions repeat cyclically.

Problem 9 [BC 31#8a,8b]

Part a: Prove that the usual formula solves the quadratic equation

az 2 + bz + c = 0, (a 6= 0)
when the coecients a, b, and c are complex numbers. Specically, by completing the square on the left-hand
side, derive the quadratic formula

1
−b + b2 − 4ac 2
z= ,
2a
2
where both square roots are to be considered when b − 4ac 6= 0.
Part b: Use the result in part (a) to nd the roots of the equation z 2 + 2z + (1 − i) = 0.

Solution 9
Part a.

The rst step in completing the square is factoring out the overall factor on the quadratic term:
 
b c b c
0 = a z2 + z + ⇒ 0 = z2 + z + ,
a a a a
since a 6= 0. Then add and subtract a term so that a perfect square appears:
 2 !  2
2 b 1 b 1 b c
0= z + z+ − + =
a 4 a 4 a a
 2  2
b 1 b c
= z+ − + ,
2a 4 a a
where the parentheses in the rst line contain the same terms as the parentheses in the second line. Then solve for z:
 2  2
b 1 b c
z+ = − ⇒
2a 4 a a
s  
2
r
b 1 b c b2 − 4ac 1p 2
z+ =± − =± 2
=± b − 4ac ⇒
2a 4 a a 4a 2a
√
−b ± b2 − 4ac
z= .
2a

Part b.

Comparing to z 2 + 2z + (1 − i) = 0, obtain:

a = 1, b = 2, c = (1 − i) .
Then
1 p 
z=− 2 ± 4 − 4 (1 − i) =
2
1 p 
=− 2 ± 2 1 − (1 − i) =
2
 √
=− 1± i .

This can be cleaned up:

  21 
z = − 1 ± eiπ/2

.
7
James Stankowicz 132 Due: Jan. 16, 2014

The last term is a bit hairy to deal with:

 21  12
eiπ/2 = ei( 2 +2πn)
  π
=

= eiπ(n+ 4 ) ,
1

where n = 0, ±1, ±2, . . .. For dierent n's:

1
: ei 4 = √ (1 + i) ,
π
n=0
2
i 54 π 1
n=1 : e = − √ (1 + i) = −eiπ/4 ,
2
with repeats after this. Using this result in the expression for z :
 
1
z = − 1 ± √ (1 + i) .
2

Problem 10 [BC p.35 Problem 5]

Let S be the open set consisting of all points z such that |z| < 1 or |z =2| < 1. State why S is not connected.

Solution 10
The region S is not connected because it is impossible to connect the two points z± = 1 ±  ∈ S , for  ∈ R and
0 <   1 with any polygonal line that remains entirely in S . In English, the point z− is just to the left of z = 1 on the
x-axis, and z+ is just to the right of z0 = 1. The only point where the two circles dening S are almost connected is
0
at z = 1, however neither of the two circles cover the point z = 1. The region S = S ∪ {z = 1} would be (one possible)
connected region.