Course 146 - Note 2
Basics about Complex Numbers -
Roots of Unity
J.Gilles
1 Basics about Complex Numbers
A complex number z is a number consisting of a real and imaginary part, respectively denoted a and b. Then we
write z = a + ıb where ı (sometimes denoted in engineering) is the imaginary unit with the property ı2 = −1.
The real and imaginary part of a complex number can be expressed by these notations:
ℜ(z) = a (1)
ℑ(z) = b (2)
The set of all complex number is denoted C. It’s easy to see that the set of real numbers R is the subset of C
of complex numbers with an imaginary part equal to zero.
We can represent complex numbers in the complex plane.
b
r
θ
a ℜ
From this representation, we can see that we can use a polar notation to denote complex numbers:
z = reıθ = r(cos θ + ı sin θ) (3)
where
arctan ab if a > 0
arctan ab + π if a < 0 and b > 0
p arctan b − π if a < 0 and b < 0
r = |r| = a2 + b 2 θ= π a (4)
2 if a = 0 and b > 0
− π2 if a = 0 and b < 0
undefined if a = 0 and b = 0
We define the complex conjugate of a complex number by
z̄ = a − ıb z̄ = re−ıθ (5)
1
It’s easy to check that z z̄ = |z|.
Now the addition, mutliplication and division of two complex numbers z1 = a1 + ıb1 , z2 = a2 + ıb2 are defined
by
• Addition:
z = z1 + z2 = a1 + ıb1 + a2 + ıb2 = (a1 + a2 ) + ı(b1 + b2 ) (6)
Equivalent to said that the real part of the addition is (a1 + a2 ) and its imaginary part b1 + b2 .
• Multiplication:
z = z1 z2 = (a1 + ıb1 )(a2 + ıb2 ) = a1 a2 + ıa1 b2 + ıa2 b1 + ı2 b1 b2 = (a1 a2 − b1 b2 ) + ı(a1 b2 + a2 b1 ) (7)
Multiplication is easier to apprehend by using the polar notation:
z = z1 z2 = r1 eıθ1 r2 eıθ2 = r1 r2 eı(θ1 +θ2 ) (8)
• Division: As we have z z̄ = |z|, the inverse of z is
1 z̄ z̄
= = ∀z 6= 0 (9)
z |z| r
Then the division
z1 z¯2
= z1 (10)
z2 |z2 |
Or with the polar notation
z1 r1 eıθ1 r1
= = eı(θ1 −θ2 ) (11)
z2 r2 eıθ2 r2
2 Roots of unity
Here we want to find the roots of the unity:
zn = 1 (12)
where z is assumed to be a complex number. The number 1 can also be viewed like a complex number (we
remind that R ⊂ C). Then this mean that unity can be written like
1 = eı2kπ k∈Z (13)
Now we have
2kπ
z n = eı2kπ ⇐⇒ z = eı n (14)
Concerning the choice of k it can be limited to the set of integers {0, 1, . . . , n − 1}. Indeed, for k ′ = k + mn, we
find the same roots. Then we have
2k′ π 2π(k+mn) 2mnπ 2kπ 2kπ 2kπ
z = eı n eı n = eı n eı n = eı2mπ eı n = eı n (15)
For example, we search the roots of z 3 = 1, we get z = {1, eı2π/3 , eı4π/3 }. This means that the roots of unity
are equi-distributed along the unit circle.
0
-1 0 1
-1
2
For z 3 = −1 = e−π+2kπ , by the same reasoning we get z = {eıπ/3 , −1, eı5π/3 } and on the unit circle:
0
-1 0 1
-1
Remark: When n is small, it could be more convenient tu use negative angle. For example 4π 3 = − 3 or
2π
5π π ı2π/3 −ı2π/3 ıπ/3 −ıπ/3
3 = − 3 and then the 3−roots of 1 are z = {1, e ,e } and the 3−roots of −1 are z = {e , −1, e }
