Lecture 4

• Roots of complex numbers

• Characterization of a polynomial by its roots

• Techniques for solving polynomial equations

 ROOTS OF COMPLEX NUMBERS
Def.:
• A number u is said to be an n-th root of complex number z
if un = z, and we write u = z 1/n .
Th.:
• Every complex number has exactly n distinct n-th roots.

Let z = r(cos θ + i sin θ); u = ρ(cos α + i sin α). Then

r(cos θ + i sin θ) = ρn (cos α + i sin α)n = ρn (cos nα + i sin nα)

⇒ ρn = r , nα = θ + 2πk (k integer)
Thus ρ = r 1/n , α = θ/n + 2πk/n .
n distinct values for k from 0 to n − 1. (z 6= 0)
    
θ 2πk θ 2πk
So u = z 1/n = r 1/n cos + + i sin + , k = 0, 1, . . . , n−1
n n n n
Note. f (z) = z 1/n is a “multi-valued” function.

!"#$%&'()(*(+,-(.//,0(/1(2+3,4(*(

x n = 1 (i.e. x n ! 1 = 0)
! x = 11/n

1 = e2m! ı " 11/ n = e2m! ı / n

2m! 2m!
=cos( )+ısin( )
n n

1/5 2m! 2m!

1 = cos( ) + ısin( ) (m = 0,1,2,3,4).
5 5
 Example 2. Find all cubic roots of z = −1 + i:
u = (−1 + i)1/3

√ 1/3
    
3π 1 2πk 3π 1 2πk
u = ( 2) cos + + i sin + , k = 0, 1, 2
4 3 3 4 3 3
that is,
1/6 π π

k=0
k=0: 2 cos 4 + i sin 4
1/6 11π 11π k=1


k=1: 2 cos 12 + i sin 12
1/6 19π 19π


k=2: 2 cos 12 + i sin 12

k=2
• Equivalently:
√
1/3 (1/3) ln(−1+i) (1/3)[ln 2+i(3π/4+2kπ)]
u = (−1 + i) =e =e
√ 1/3 i(π/4+2kπ/3)
= ( 2) e
 !""#$%"&%'"()*"+,-($%

P ( z ) " an z n + an !1 z n !1 + ! + a0 .

P( z = zi ) = 0 ! zi is a root

[Gauss, 1799]

• proof of fundamental theorem of algebra is given in the course

“Functions of a complex variable”, Short Option S1
3..&),.4,-./0*.1(#/),

P ( z ) " an z n + an !1 z n !1 + ! + a0 .

P( z = zi ) = 0 ! zi is a root

!"#$#%&'$()(*+,#,-./0*.1(#/,20,(&),$..&),

an z n + an !1 z n !1 + ! + a0 = an ( z ! z1 )( z ! z2 )! ( z ! zn )
n n
n
= an ( z ! z n !1
"z
j =1
j + ! + (!1) n
# z ).
j =1
j

!.1-#$(*+,%.'5%('*&),.4,6*78,#*9,6:,

an !1 a0
= !" z ; = (!1) n # z
an an
!

j j
 !"#"$%&'()'*+$!%&'*,-.$/$ a2 x 2 + a1 x + a0

=,,3./$ x1,2 =
(!a ±
1 a12 ! 4a2 a0 ) <$

2a2
8$

52$+,167!89$),,3.$+,1!$:-$+,167!8$+,-;&#'3!$6':).$ 8$

a1 a0
0&1$,2$),,3.$ = !( x1 + x2 ) 4),(&+3$,2$),,3.$ = x1.x2
a2 a2

2,)$),,3.$
• >!-!)'7$.,7&*,-.$-,3$'?':7'@7!$2,)$A:#A!)$,)(!)$6,7B-,1:'7.$C%&')*+.$'-($'@,?!D$
H$
E'-$F-($.,7&*,-.$:-$.6!+:'7$+'.!.G"$
 Example 1:
z 5 + 32 = 0

• The solutions of the given equation are the fifth roots of −32:
    
π 2πk π 2πk
(−32)1/5 = 321/5 cos + + i sin + , k = 0, 1, 2, 3, 4
5 5 5 5

that is,
Im z
k=1
k = 0 : 2 cos π5 π


+ i sin 5
k=0
k = 1 : 2 cos 3π 3π


5 + i sin 5
k = 2 : −2 k=2
7π 7π

Re
k = 3 : 2 cos 5 + i sin 5
9π 9π

k=4
k = 4 : 2 cos 5 + i sin 5
k=3
,-.!/01+&+2+34456+47+/40894!:.06+

7 7
(z + ı) + (z ! ı) = 0
z+ı 7
( ) = !1 = e(2m+1)" ı
z!ı
z+ı
! = e(2m+1)# ı /7
z"ı

! z(1 " e(2m+1)# ı /7 ) = "ı(1 + e(2m+1)# ı /7 )

e(2m+1)" ı /7 + 1
! z = ı (2m+1)" ı /7
e #1

e(2m+1)! ı /14 + e "(2m+1)! ı /14 2cos( 2m+1

14
!) 2m+1
= ı (2m+1)! ı /14 "(2m+1) ! ı /14
= ı 2m+1
= cot( 14
!)
e "e 2ısin( 14 ! )
!"#$%$&$'$($)$*+
=>27?&"#@#;#2)#2&+"3)2A4"#/'37#
7 7
(z + ı) + (z ! ı) = 0
!"#$%&&#'(")#)""*#+,"#-'".-%")+#'/## x r y n!r %)# ( x + y)n

0,"1"#23"#-')4")%")+&5#'6+2%)"*#/3'7#821-2&91#+3%2):&"#;#

( x + y )0 1
( x+ y )1 1 1 Solution:
( x + y )2 1 2 1 z = cot( 2m+1
14
!)
( x+ y )3 1 3 3 1
( x + y )4 1 4 6 4 1
( x+ y )5 1 5 10 10 5 1

<+,#3'$#'/#821-2&91#+3%2):&"#%1# 1 7 21 35 35 21 7 1 1'#

(z + ı)7 + (z ! ı)7 = 0 " z 7 ! 21z 5 + 35z 3 ! 7z = 0

:;*5<+#$=$6$>#3$*)%3"#&$4%&5$

!"#$%&'(')*+$#,-*.%)$ (z + ı)7 + (z ! ı)7 = 0 " z 7 ! 21z 5 + 35z 3 ! 7z = 0

/*)$0#$1&'2#)$')$*)%3"#&$4%&5$6$

z 7 ! 21z 5 + 35 z 3 ! 7 z = 0 zm , m = 3
" z 6 ! 21z 4 + 35 z 2 ! 7 = 0 or z=0
" w3 ! 21w2 + 35w ! 7 = 0 ( w # z 2 )

7#)/#$3"#$&%%38$%4$ w3 ! 21w2 + 35w ! 7 = 0 *&#$

w = cot 2 ( 2 m14+1 ! ) (m = 0,1, 2)

2
9-5$%4$&%%38$ " # m =0 cot 2 ( 2 m14+1 ! ) = 21
  
n n!
Pascal’s triangle = table of binomial coefficients r
= r!(n−r)!
i.e., coefficients of xr y n−r in (x + y)n
[Pascal, 1654]

1
1 1
1 2 1
1 3 3 1
1 4 6 4 1
1 5 10 10 5 1
1 6 15 20 15 6 1
1 7 21 35 35 21 7 1

• r-th element of n-th row given by sum of two elements above it in (n − 1)-th row:
! ! !
n n−1 n−1
= +
r r−1 r

[use (x + y)n = (x + y)n−1 x + (x + y)n−1 y ]

 Historical note

♦ binomial coefficients already known in the middle ages:

• “Pascal’s triangle” first discovered by Chinese mathematicians of 13th

century to find coefficients of (x + y)n
 
n n!
• = r!(n−r)! in Hebrew writings of 14th century is
r
number of combinations of n objects taken r at a time

♦ Pascal (1654) rediscovers triangle and most importantly unites

algebraic and combinatorial viewpoints
−→ theory of probability; proof by induction
 !"#$# %&'()*+#*",-./*#0)*+*#()*#1&2*+/34&5#*61,7'&#48#&'(#'9:4'18#;#
( + y )0 1
( x+ y )1 1 1
3 2
z + 7 z + 7 z + 1 = 0. ( x + y )2
( x+ y )3 1
1
3
2
3
1
1
( x + y )4 1 4 6 4 1
( x+ y )5 1 5 10 10 5 1

<()#+'0#'=#>,8?,/@8#(+4,&5/*#48# 1 8 28 56 70 56 28 8 1 8'#

1
2 [( z + 1)8 ! ( z ! 1)8 ] = 8 z 7 + 56 z 5 + 56 z 3 + 8 z
= 8 z[ w3 + 7 w2 + 7 w + 1] ( w " z 2 ).

A'0## ( z + 1)8 ! ( zC##! 1)8 = 0 #0)*&##

z +1
z "1 = e 2 m! i/ 8
e m! ıB#/4# + 1
#4B*B#0)*&## z =
m! ı /4
= "ıcot(m! / 8) (m = 1,2,…,7),
e "1

8'#()*#+''(8#'=#()*#54:*&#*61,7'&#,+*##

z = " cot 2 (m!/8) m = 1, 2, 3

 Example
Question from 2008 Paper

Find all the solutions of the equation

 n
z+i
= −1 ,
z−i
and solve
z 4 − 10z 2 + 5 = 0 .

 n
z+i z+i
= −1 = ei(π+2N π) , N integer ⇒ = ei(π/n+2N π/n) , N = 0, 1, . . . , n−1
z−i z−i
ei(π/n+2N π/n) + 1 cos[π(1 + 2N )/(2n)] π(1 + 2N )
Then : z = i i(π/n+2N π/n) =i = cotg .
e −1 i sin[π(1 + 2N )/(2n)] 2n
For n = 5 : (z+i)5 = −(z−i)5 ⇒ z(z 4 −10z 2 +5) = 0 . Then the 4 roots of z 4 −10z 2 +5 = 0 are
π 3π 7π 9π
cotg , cotg , cotg , cotg .
10 10 10 10
 !"#$####%&'(#)&*)##
z 2m " a 2m ! 2! (m " 1)!
2 2
= ( z 2 " 2az cos + a 2 )( z 2 " 2az cos + a 2 ) ! ( z 2 " 2az cos + a 2 ).
z "a m m m

0=.=#%&'(#)&*)## P(z) = Q(z) where

P(z) ! z 2m " a 2m 6+''),#<# zr = ae r! i / m )

! 2! (m " 1)!
Q( z ) # ( z 2 " a 2 )( z 2 " 2az cos + a 2 )( z 2 " 2az cos + a 2 ) ! ( z 2 " 2az cos + a 2 ).
m m m

r! r!
+''),# zr = a cos ± a 2 cos 2 " a2
m m
r! r!
=a(cos ±ı 1 " cos 2 )=ae±ır! / m (r=0,1,…,m).
m m

-.*/012#3'.430.1)# a2 m = 1 5678#*1/#9678#0/.1:3*;#
• This concludes part A of the course.

A. Complex numbers

1 Introduction to complex numbers

2 Fundamental operations with complex numbers

3 Elementary functions of complex variable

4 De Moivre’s theorem and applications

5 Curves in the complex plane

6 Roots of complex numbers and polynomials