Complex analysis and power series ∗

Jean-Marie Dufour †
McGill University

First version: March 1992

Revised: January 2002, October 2016
This version: October 2016
Compiled: January 10, 2017, 15:38

∗ This work was supported by the William Dow Chair in Political Economy (McGill University), the Bank of Canada
(Research Fellowship), the Toulouse School of Economics (Pierre-de-Fermat Chair of excellence), the Universitad Car-
los III de Madrid (Banco Santander de Madrid Chair of excellence), a Guggenheim Fellowship, a Konrad-Adenauer
Fellowship (Alexander-von-Humboldt Foundation, Germany), the Canadian Network of Centres of Excellence [program
on Mathematics of Information Technology and Complex Systems (MITACS)], the Natural Sciences and Engineering Re-
search Council of Canada, the Social Sciences and Humanities Research Council of Canada, and the Fonds de recherche
sur la société et la culture (Québec).
† William Dow Professor of Economics, McGill University, Centre interuniversitaire de recherche en analyse des

organisations (CIRANO), and Centre interuniversitaire de recherche en économie quantitative (CIREQ). Mailing address:
Department of Economics, McGill University, Leacock Building, Room 414, 855 Sherbrooke Street West, Montréal,
Québec H3A 2T7, Canada. TEL: (1) 514 398 4400 ext. 09156; FAX: (1) 514 398 4800; e-mail: jean-marie.dufour@
mcgill.ca . Web page: http://www.jeanmariedufour.com
Contents

List of Definitions, Assumptions, Propositions and Theorems iii

1. Analytic functions 1

2. Power series 3

3. Taylor and Laurent series 6

4. Sums, products and ratios of power series 9

5. Singularities 11

6. Partial fractions 13

7. Proofs and references 14

i
List of Definitions, Assumptions, Propositions and Theorems
1.2 Definition : Limit of a complex function . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Definition : Right and left limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.4 Definition : Continuous function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.5 Definition : Derivative of a complex function . . . . . . . . . . . . . . . . . . . . . 1
1.7 Proposition : Continuity of differentiable functions . . . . . . . . . . . . . . . . . . 2
1.8 Theorem : Properties of differentiation . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.9 Theorem : Chain rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.10 Theorem : Derivatives of important functions . . . . . . . . . . . . . . . . . . . . . 2
1.11 Theorem : Derivative of a real function of a complex variable . . . . . . . . . . . . . 2
1.12 Definition : Analytic function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.14 Definition : Singular point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.15 Theorem : Operations on analytic functions . . . . . . . . . . . . . . . . . . . . . . 3
1.16 Theorem : Composition of analytic functions . . . . . . . . . . . . . . . . . . . . . 3
1.17 Theorem : Infinite differentiability of analytic functions . . . . . . . . . . . . . . . . 3
1.18 Theorem : Important analytic functions . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 Definition : Power series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.3 Theorem : Convergence radius of a power series (Abel-Hadamard) . . . . . . . . . . 4
2.5 Corollary : Absolute convergence of power series . . . . . . . . . . . . . . . . . . . 4
2.6 Corollary : Bounds on coefficients of power series . . . . . . . . . . . . . . . . . . . 4
2.7 Theorem : Uniform absolute convergence of power series . . . . . . . . . . . . . . . 4
2.8 Proposition : Convergence radius and ratio criterion . . . . . . . . . . . . . . . . . . 4
2.9 Theorem : Convergence condition on the unit circle . . . . . . . . . . . . . . . . . . 4
2.10 Theorem : Continuity of power series on the unit circle (Abel) . . . . . . . . . . . . 5
2.11 Corollary : Continuity of real power series on the unit circle . . . . . . . . . . . . . . 5
2.13 Theorem : Criterion for convergence and continuity of real power series on the unit
cirle (Tauber) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.14 Theorem : Unicity of power series coefficients . . . . . . . . . . . . . . . . . . . . . 5
2.15 Corollary : Unicity of power series coefficients in a circle . . . . . . . . . . . . . . . 5
2.16 Theorem : Differentiability of power series . . . . . . . . . . . . . . . . . . . . . . 5
2.18 Corollary : Differentiability at all orders of power series . . . . . . . . . . . . . . . . 6
2.19 Theorem : Integrability of power series . . . . . . . . . . . . . . . . . . . . . . . . 6
2.20 Definition : Two-sided power series . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.21 Proposition : Convergence annulus of two-sided power series . . . . . . . . . . . . . 6
3.1 Theorem : Taylor series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.3 Corollary : Cauchy inequalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.5 Corollary : Equivalence between analyticity and the existence of a Taylor series . . . 7
3.6 Theorem : Zeros of analytic functions . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.8 Theorem : Factorization of an analytic function . . . . . . . . . . . . . . . . . . . . 8
3.10 Theorem : Simplification rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.12 Theorem : Local separability of analytic functions . . . . . . . . . . . . . . . . . . . 8

ii
3.14 Theorem : Laurent series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.16 Corollary : Laurent series near an isolated singularity . . . . . . . . . . . . . . . . . 9
3.18 Corollary : Generalized Cauchy inequalities . . . . . . . . . . . . . . . . . . . . . . 9
3.19 Definition : Principal and regular parts of a Laurent series . . . . . . . . . . . . . . . 9
4.1 Theorem : Pointwise convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 Theorem : Convergence in a circle . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.3 Theorem : MacLaurin series for a rational function . . . . . . . . . . . . . . . . . . 11
5.1 Definition : Pole and essential singularity . . . . . . . . . . . . . . . . . . . . . . . 11
5.2 Theorem : Characterization of isolated singularities . . . . . . . . . . . . . . . . . . 11
5.3 Definition : Order of a pole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.4 Theorem : Singularities and Laurent series . . . . . . . . . . . . . . . . . . . . . . . 12
5.5 Theorem : Behavior of an analytic function near an essential singularity (Picard) . . . 12
6.1 Theorem : Partial fraction expansion of rational functions . . . . . . . . . . . . . . . 13
6.2 Theorem : Factorization of an analytic function with finite number of poles . . . . . . 13
6.3 Theorem : Partial fraction expansion of an analytic function with finite number of poles 13

iii
1. ANALYTIC FUNCTIONS 1

1. Analytic functions
1.1 Notation In this text, z refers to a complex number
(z ∈ C), while f and

g represent functions f :
E → C and g : F → C, where B(a; δ̄ ) ⊆ E ⊆ C, B a; δ̄ ⊆ F ⊆ C, B a; δ̄ = z ∈ C : |z − a| < δ̄ ,


0 < δ̄ ≤ ∞ and a ∈ C. In other words, f and g are functions with complex values whose domains
are subsets E and F of the complex numbers containing an open ball centered at the point a.

1.2 Definition L IMIT OF A COMPLEX FUNCTION . Let b ∈ C. We say that f (z) converges to b
when z tends to a, denoted
lim f (z) = b ,
z→a

iff the following property holds: for any ε > 0, there exists δ > 0 such that

|z − a| < δ and z 6= a ⇒ | f (z) − b| < ε .

1.3 Definition R IGHT AND LEFT LIMITS. Let b ∈ C, x ∈ R and f : E → C, where B a; δ̄ ⊆ E ⊆ R

and a ∈ R. We say that f (x) converges to b when x tends to a from the left, denoted

lim f (x) = b or f (a−) = b ,

x→a−

iff the following property holds: for any ε > 0, there exists δ > 0 such that

|x − a| < δ and x < a ⇒ | f (x) − b| < ε .

Similarly, we say that f (x) converges to b when x tends to a from the right, denoted

lim f (x) = b or f (a+) = b ,

x→a+

iff the following property holds: for any ε > 0, there exists δ > 0 such that

|x − a| < δ and x > a ⇒ | f (x) − b| < ε .

1.4 Definition C ONTINUOUS FUNCTION . We say that the function f is continuous at point a iff

lim f (z) = f (a) .

z→a

1.5 Definition D ERIVATIVE OF A COMPLEX FUNCTION. We say that the function f is differen-
tiable at point a iff there exists a number f ′ (a) ∈ C such that

f (z) − f (a)
lim = f ′ (a) .
z→a z−a
We call f ′ (a) the derivative of f (z) at a.

1.6 Remark We also denote f ′ (z) by d

dz f (z) .
1. ANALYTIC FUNCTIONS 2

1.7 Proposition C ONTINUITY OF DIFFERENTIABLE FUNCTIONS . If the function f is differen-

tiable at point a, then it is continuous at point a.

1.8 Theorem P ROPERTIES OF DIFFERENTIATION. Let z ∈ B a; δ̄ ⊆ E ∩ F . If the functions f

and g are differentiable at point z, then

(1) d
dz [c f (z)] = c f ′ (z) ,

(2) d
dz [ f (z) + g (z)] = f ′ (z) + g′ (z) ,

(3) d
dz[ f (z) g (z)] = f ′ (z) g (z) + f (z) g′ (z) ,
f ′ (z)g(z)− f (z)g′ (z)
h i
d f (z)
(4) dz g(z) = 2 , provided g (z) 6= 0 .
g(z)

1.9 Theorem C HAIN RULE. Let h : G → C where B ( f (a) ; δ 0 ) ⊆ f (E) ⊆ G ⊆ C, B ( f (a) ; δ 0 ) =

{z ∈ C : |z − f (a)| < δ 0 } and 0 < δ 0 ≤ ∞. If the function f is differentiable at point a and if h is
differentiable at point f (a), then the composite function H (z) = h [ f (z)] is differentiable at point a
and
H ′ (a) = h′ [ f (a)] f ′ (a) .

1.10 Theorem D ERIVATIVES OF IMPORTANT FUNCTIONS .

(1) If c is a complex constant, then

d
(c) = 0 .
dz
(2) If n is a real constant,

d n
(z ) = n zn−1 , provided z 6= 0 when n < 1 .
dz

(3) d
dz (ez ) = ez .

1.11 Theorem D ERIVATIVE OF A REAL FUNCTION OF A COMPLEX
VARIABLE. Suppose the

function f only takes real values at all points of the open ball B a; δ̄ , i.e. f (z) ∈ R for any z ∈
B a; δ̄ . If f is differentiable at point a, then f ′ (a) = 0.

1.12 Definition A NALYTIC FUNCTION. If there exists a positive constant ε > 0 such that the
function f is differentiable at all points z such that |z − a| < ε, we say that the function is analytic
at point a. If the function f is analytic at all points of a domain D ⊆ C, we say that f is analytic on
the domain D.

1.13 Remark An analytic function is also called a holomorphic function.

2. POWER SERIES 3

1.14 Definition S INGULAR POINT. If a function f is not analytic at point z0 , but for any ε > 0
there exists a point z1 such that |z1 − z0 | < ε and f is analytic at z1 , we say that z0 is a singular
point (or a singularity) of the function f . If, furthermore, there exists a radius R > 0 such that f is
analytic on the disk 0 < |z − z0 | < R, we say that z0 is an isolated singular point of the function f .

1.15 Theorem O PERATIONS ON ANALYTIC FUNCTIONS. If the functions f and g are analytic at
point a, then

(1) the functions f (z) + g (z) and f (z) g (z) are analytic at point a ;

(2) the function f (z) /g (z) is analytic at point a provided g (a) 6= 0.

1.16 Theorem C OMPOSITION OF ANALYTIC FUNCTIONS. Let h : G → C where B ( f (a) ; δ 0 ) ⊆

f (E) ⊆ G ⊆ C, B ( f (a) ; δ 0 ) = {z ∈ C : |z − f (a)| < δ 0 } and 0 < δ 0 ≤ ∞. If the function f is
analytic at point a and if h is analytic at point f (a), then the composed function (h ◦ f ) (z) = h [ f (z)]
is analytic at point a.

1.17 Theorem I NFINITE DIFFERENTIABILITY OF ANALYTIC FUNCTIONS. If the function f is

analytic at point a ∈ C, then f has derivatives of all orders at a, and the derivative functions are also
analytic at point a.

1.18 Theorem I MPORTANT ANALYTIC FUNCTIONS.

(1) Any polynomial of degree n,

f (z) = a0 + a1 z + · · · + an zn (1.1)

where a0 , a1 , . . . , an ∈ C, is analytic at all points z ∈ C.

(2) A rational function

f (z) = P (z) /Q (z) (1.2)
where P (z) and Q (z) are polynomials of degrees p and q, is analytic everywhere, except
when Q (z) = 0.

(3) The functions ez , cos(z) and sin(z) are analytic everywhere.

(4) The function log(z) is analytic everywhere except at z = 0.

2. Power series
2.1 Definition P OWER SERIES. Let {an }∞ n=0 ⊆ C, z0 ∈ C and z ∈ C. We call the series
∑∞
n
n=0 an (z − z 0 ) a power series centered at z 0 . The numbers an are the coefficients of the series.

2.2 Remark In this definition and the sequel, we will use the convention 00 = 1.
2. POWER SERIES 4

2.3 Theorem C ONVERGENCE RADIUS OF A POWER SERIES (A BEL -H ADAMARD ). Let

∑∞n=0 an (z − z0 ) a power series and
n

α = lim sup |an |1/n , R = 1/α ,

n→∞

where R = ∞ when α = 0, and R = 0 when α = ∞. Then the series ∑∞ n=0 an (z − z0 ) converges

n

absolutely if |z − z0 | < R and diverges if |z − z0 | > R. Further, if 0 ≤ ρ < R, the convergence is

uniform for |z − z0 | ≤ ρ .

2.4 Remark We call R the convergence radius of the series ∑∞ n

n=0 an (z − z0 ) . The expression 1/R =
lim sup |an |1/n is the Hadamard formula for the convergence radius.
n→∞

2.5 Corollary A BSOLUTE CONVERGENCE OF POWER SERIES. If the power series

∞ n
∑n=0 na (z − z 0 ) converges for z = z 1 , where z 1 =
6 z 0 , then it converges absolutely for any z such
that |z − z0 | < |z1 − z0 | .

2.6 Corollary B OUNDS ON COEFFICIENTS OF POWER SERIES . Let ∑∞ n

n=0 an (z − z0 ) a power
series whose convergence radius is R, and let ε > 0.
1

n
(1) If 0 < R ≤ ∞, there exists an integer N , such that |an | < R +ε
for n > N.

n
(2) If 0 < R < ∞, there is an infinity of values of n for which |an | > R1 − ε .

(3) If R = 0, there is an infinity of values of n for which |an | > ε n .

2.7 Theorem U NIFORM ABSOLUTE CONVERGENCE OF POWER SERIES. If the power series
∑∞n=0 an (z − z0 ) converges absolutely for z = z1 , where z1 6= z0 , then it converges absolutely and
n

uniformly on the closed disk D = {z ∈ C : |z − z0 | ≤ |z1 − z0 |} .

2.8 Proposition C ONVERGENCE RADIUS AND RATIO CRITERION . Let ∑∞ n

n=0 an (z − z0 ) be a
power series whose convergence radius is R. Then

an+1 an+1
lim inf ≤ R ≤ lim sup .
n→∞ an n→∞ an

Further, if lim |an+1 /an | exists or lim |an+1 /an | = ∞, then R = lim |an+1 /an | .
n→∞ n→∞ n→∞

2.9 Theorem C ONVERGENCE CONDITION ON THE UNIT CIRCLE. Let ∑∞ n=0 an z be a power
n
∞
series whose convergence radius is 1. If {an }n=0 is a sequence of real numbers such that

(a) an+1 ≤ an , ∀n , and

(b) lim an = 0,
n→∞

then the series ∑∞

n=0 an z converges at any point of the circle |z| = 1, except possibly at z = 1.
n
2. POWER SERIES 5

2.10 Theorem C ONTINUITY OF POWER SERIES ON THE UNIT CIRCLE (A BEL ). If the series
∑∞ ∞ ∞
n=0 an converges, then the function ∑n=0 an z , where |z| < 1, tends to ∑n=0 an when z → 1 so that
n

|1 − z| / (1 − |z|) remains bounded.

2.11 Corollary C ONTINUITY OF REAL POWER SERIES ON THE UNIT CIRCLE. If {an }∞ n=0 is a
∞ ∞
sequence of real numbers such that ∑n=0 an converges, and if the power series f (x) = ∑n=0 an xn
converges for |x| < 1, where x ∈ R, then lim− f (x) exists and
x→1

∞
lim− f (x) =
x→1
∑ an .
n=0

2.12 Remark If the series ∑∞

n=0 an does not converge, the limit lim− f (x) may or may not exist.
x→1
In general, the existence of the limit lim− f (x) does not guarantee the convergence of the series
x→1
∑∞
n=0 an . There are however cases where the existence of the limit lim− f (x) implies the convergence
x→1
of ∑∞
n=0 an (Tauberian theorems). The following theorem provides an example.

2.13 Theorem C RITERION FOR CONVERGENCE AND CONTINUITY OF REAL POWER SERIES ON
∞
THE UNIT CIRLE (TAUBER ) . If {an }n=0 is a sequence of real numbers such that f (x) = ∑∞ n=0 an x
n

converges for |x| < 1, where x ∈ R, if lim (n an ) = 0 and if lim− f (x) exists, then the series ∑∞
n=0 an
n→∞ x→1
converges and lim− f (x) = ∑∞
n=0 an .
x→1

2.14 Theorem U NICITY OF POWER SERIES COEFFICIENTS. If ∑∞ n

n=0 an (z − z0 ) and
∑∞
n
n=0 bn (z − z0 ) are two power series which converge for |z − z0 | < R, where R > 0, and if the
limits of these series coincide on a sequence of points {zk }∞
k=1 such that 0 < |zk | < R, ∀k, and
lim zk = z0 , then
k→∞
an = bn , ∀n .

2.15 Corollary U NICITY OF POWER SERIES COEFFICIENTS IN A CIRCLE. If ∑∞ n=0 an (z − z0 )

n
∞ n
and ∑n=0 bn (z − z0 ) are two power series which converge for |z − z0 | < R, where R > 0, and if the
limits of these series coincide for any z in the circle |z − z0 | < R, then

an = bn , ∀n .

2.16 Theorem D IFFERENTIABILITY OF POWER SERIES. Let f (z) = ∑∞ n

n=0 an (z − z0 ) for
∞ n
|z − z0 | < R, where R > 0 and ∑n=0 an (z − z0 ) is a power series whose convergence radius is R.
Then the function f (z) is analytic (and thus differentiable) on the disk |z − z0 | < R, and
∞
f ′ (z) = ∑ n an (z − z0 )n−1
n=1

where the power series ∑∞ n=1 n an (z − z0 )

n−1
has convergence radius R. If, furthermore, 0 <
R < ∞ and f (z) is a function such that f (z) = ∑∞ n
n=0 an (z − z0 ) at every point where the series
3. TAYLOR AND LAURENT SERIES 6

∑∞
n
n=0 an (z − z0 ) converges, then there is at least one point on the circle |z − z0 | = R where the
function f (z) is not analytic.

2.17 Remark In other words, we can obtain the derivative of the function f (z) = ∑∞ n=1 an (z − z0 )
n

by differentiating the series term by term, and the derivative series has the same convergence radius
as the original series.

2.18 Corollary D IFFERENTIABILITY AT ALL ORDERS OF POWER SERIES. Let f (z) =

∑∞ ∞
n n
a
n=0 n (z − z 0 ) for |z − z 0 | < R , where ∑ a
n=0 n (z − z 0 ) is a power series whose convergence ra-
dius is R. Then the function f (z) has derivatives of all orders, and these derivatives can be obtained
by differentiating the series term by term. The derivative series all have the same convergence radius
R, and
f (n) (z0 )
an = , n = 0, 1, 2, ...
n!
where f (n) (z) is the derivative of order n of f (z) .

2.19 Theorem I NTEGRABILITY OF POWER SERIES. Let ∑∞ n

n=0 an (z − z0 ) be a power series
∞ n
whose convergence radius is R, let f (z) = ∑n=0 an (z − z0 ) for |z − z0 | < R, C a contour (continuous
curve) in the interior of the convergence circle |z − z0 | < R, and g (z) a continuous function on C.
Then Z ∞ Z
f (z) g (z) dz = ∑ an g (z) (z − z0 )n dz .
n=0
C C

2.20 Definition T WO - SIDED POWER SERIES. Let {an }∞ n=−∞ , z0 ∈ C and z ∈ C. We call two-sided
power series a series of the form ∑∞ n=−∞ an (z − z 0 )n
. This series converges when the two series
∞ n and −1 n
∑n=0 n
a (z − z 0 ) ∑n=−∞ n
a (z − z 0 ) converge. Otherwise, we say it diverges.

2.21 Proposition C ONVERGENCE ANNULUS OF TWO - SIDED POWER SERIES. Let

∑∞ ∞
n n
a
n=0 n (z − z 0 ) and ∑ a
n=1 −n (z − z 0 ) be power series whose convergence radii are R 1 and R2
respectively, where R1 > 0 and R2 > 0.
(1) If 1/R2 < R1 , the series ∑∞ n
n=−∞ an (z − z0 ) converges for 1/R2 < |z − z0 | < R1 and diverges
when |z − z0 | > R1 or |z − z0 | < 1/R2 .

(2) If R1 < 1/R2 , the series ∑∞ n

n=−∞ an (z − z0 ) diverges everywhere.

(3) If R1 = 1/R2 , the series ∑∞

n=−∞ an (z − z0 ) diverges everywhere except possibly on the circle
n

|z − z0 | = R1 .

3. Taylor and Laurent series

3.1 Theorem TAYLOR SERIES . Let f be an analytic function at any point of the open disk

D = {y ∈ C : |z − z0 | < R} , where z0 ∈ C and 0 < R ≤ ∞ .

3. TAYLOR AND LAURENT SERIES 7

Then there exists a unique sequence {an }∞

n=0 in C such that

∞
f (z) = ∑ an (z − z0 )n , ∀z ∈ D .
n=0

Further,
1 f (z) dz
Z
an = f (n) (z0 ) /n! = , n = 0, 1, 2, . . .
2π i (z − z0 )n+1
C

where C = {z ∈ C : |z − z0 | = ρ } and ρ is any radius such that 0 < ρ < R.

3.2 Remark In other words, an analytic function on the interior of a circle centered at z0 can be
written in the interior of this circle as a power series of (z − z0 ). Further, this series is unique. The
integral is evaluated counterclockwise.
R
C

3.3 Corollary C AUCHY INEQUALITIES. Under the conditions of Theorem 3.1, suppose that
| f (z)| ≤ M for z ∈ C (ρ ), where C (ρ ) = {z ∈ C : |z − z0 | = ρ } and 0 < ρ < R. Then

|an | = | f n (z0 )| /n! ≤ M/ρ n , n = 0, 1, 2, . . . .

3.4 Remark The Cauchy inequalities entail: for ρ < 1, the coefficients of the Taylor series must
decline at an exponential rate which depends on the convergence radius.

3.5 Corollary E QUIVALENCE BETWEEN ANALYTICITY AND THE EXISTENCE OF A TAYLOR SE -

RIES . Let D = {z ∈ C : |z − z0 | < R} where z0 ∈ C and 0 < R ≤ ∞. Then a function f is analytic on
the domain D iff there exists a unique sequence {an }∞
n=0 in C such that

∞
f (z) = ∑ an (z − z0 )n , ∀z ∈ D .
n=0

3.6 Theorem Z EROS OF ANALYTIC FUNCTIONS. Let f be an analytic function at point z0 , such
that f (z0 ) = 0. If f (n) (z0 ) = 0, n = 1, 2, . . . , m − 1, but f (m) (z0 ) 6= 0, where m ≥ 1, then there
exists a radius R > 0 such that the function f can be written

f (z) = (z − z0 )m g (z)

for |z − z0 | < R, where the function g (z) is analytic at z0 , and g (z) 6= 0 for |z − z0 | < R. If f (n) (z0 ) =
0, n = 1, 2, . . . , then there exists a radius R > 0 such that f (z) = 0 for |z − z0 | < R.

3.7 Remark The latter theorem implies that the zeros of a non-zero analytic function are isolated:
unless all the derivatives of f are zero, we can find a radius R > 0 such that z0 is the only point
where the function cancels in the disk |z − z0 | < R. We call z0 a root of the function f , and m its
multiplicity.
3. TAYLOR AND LAURENT SERIES 8

3.8 Theorem FACTORIZATION OF AN ANALYTIC FUNCTION. Let f be an analytic function on an

open convex domain U ⊆ C. If the function f has only a finite number p of distinct roots z1 , . . . , z p ,
then the function f can be written

f (z) = (z − z1 )m1 ... (z − z p )m p g (z) , z ∈ U

where m1 , . . . , m p are the multiplicities of the roots z1 , . . . , z p and g (z) is an analytic function on U
such that g (z) 6= 0 for any z ∈ U.

3.9 Remark In other words, an analytic function with a finite number of roots is finite can be
written as the product of a polynomial with the same roots and an analytic function which is different
from zero everywhere. An open disk C = {z ∈ C : 0 ≤ (z − z0 ) < R} where R > 0 is a convex set.
The latter theorem remains valid when U is a convex and connected set.

3.10 Theorem S IMPLIFICATION RULE. Let U ⊆ C an open and connected set. If f and g are two
analytic functions on U such that

f (z) g (z) = 0 , ∀z ∈ U ,

then f (z) = 0 , ∀z ∈ U , or g (z) = 0 , ∀z ∈ U.

3.11 Remark If f , g and h are three analytic functions on U such that f (z) h (z) = g (z) h (z), ∀z ∈ U,
and if h (z) 6= 0 for at least one value of z ∈ U, then

[ f (z) − g (z)] h (z) = 0

and we can conclude that f (z) = g (z) , ∀z ∈ U.

3.12 Theorem L OCAL SEPARABILITY OF ANALYTIC FUNCTIONS. Let f be an analytic function

which is not constant on an open connected set U . Then, for w ∈ C and z0 ∈ U , there exists a radius
R > 0 such that f (z) 6= w for 0 < (z − z0 ) < R.

3.13 Remark In other words, if the function is not constant, we can find a radius R > 0 such that
f (z) takes the value w at least one time in the disk 0 ≤ |z − z0 | < R.

3.14 Theorem L AURENT SERIES . Let C0 and C1 be two circles centered at z0 such that C0 is
contained in C1 , i.e.

C0 = {z ∈ C : |z − z0 | = R0 } ,C1 = {z ∈ C : |z − z0 | = R1 } where 0 ≤ R0 < R1 ≤ ∞ .

Let f be an analytic function on C0 and C1 as well as on the domain between these two circles. Then
there exists a unique two-sided sequence {an }∞n=−∞ in C such that

∞
f (z) = ∑ an (z − z0 )n
n=−∞
4. SUMS, PRODUCTS AND RATIOS OF POWER SERIES 9

for any z such that R0 < |z − z0 | < R1 , where

1 f (z) dz
Z
an = , for n = 0, 1, 2, . . .
2π i (z − z0 )n+1
C1
1 f (z) dz
Z
= , for n = −1, −2, . . . .
2π i (z − z0 )n+1
C0

Further, for any circle C = {z ∈ C : |z − z0 | = R} where R0 < R < R1 ,

1 f (z) dz
Z
an = , n = 0, ±1, ±2, . . . .
2π i (z − z0 )n+1
C

3.15 Remark The line integrals and are evaluated counterclockwise.

R R R
,
C0 C1 C

3.16 Corollary L AURENT SERIES NEAR AN ISOLATED SINGULARITY. If f is an analytic func-

tion at any point of the disk |z − z0 | < R, where R > 0, except possibly at z0 , then there exists a
unique two-sided sequence {an }∞ n=−∞ in C such that

∞
f (z) = ∑ an (z − z0 )n
n=−∞

for any z such that 0 < |z − z0 | < R.

3.17 Remark In other words, if z0 is a singular point of the function f , the function f can be
represented by a Laurent series on the disk 0 < |z − z0 | < R. If, furthermore, an = 0 for n < 0, the
Laurent series reduces to a Taylor series, and we can redefine the function f at z0 so that the latter
is analytic at z0 and thus everywhere on the disk 0 ≤ |z − z0 | < R. In such a case, we say that the
singular point z0 is removable. When a function f is analytic at any point of the disk |z − z0 | < R, it
is clear we must have an = 0 for n < 0.

3.18 Corollary G ENERALIZED C AUCHY INEQUALITIES. Under the conditions of Theorem 3.14,
suppose that | f (z)| ≤ M for z ∈ C(R), where C(R) = {z ∈ C : |z − z0 | = R} and R0 < R < R1 . Then

|an | ≤ M/Rn , n = 0, ±1, ±2, . . . .

3.19 Definition P RINCIPAL AND REGULAR PARTS OF A L AURENT SERIES. In a Laurent series
∑∞
n −1 n
n=−∞ an (z − z0 ) , we call the series ∑n=−∞ an (z − z0 ) the principal part of the series, while the
series ∑∞ n
n=0 an (z − z0 ) is called the regular part of the series.

4. Sums, products and ratios of power series

4.1 Theorem P OINTWISE CONVERGENCE. Let ∑∞ n ∞ n
n=0 an (z − z0 ) and ∑n=0 bn (z − z0 ) be two
convergent power series whose limits are f (z) and g (z) respectively at a given point z. Then the
4. SUMS, PRODUCTS AND RATIOS OF POWER SERIES 10

following properties hold:

(1) c f (z) = ∑∞ n
n=0 c an (z − z0 ) , ∀c ∈ C ;

(2) f (z) + g (z) = ∑∞ n

n=0 (an + bn ) (z − z0 ) ;

(3) if f (z) or g (z) converges absolutely, then

∞
f (z) g (z) = ∑ cn (z − z0 )n (4.1)
n=0

where cn = ∑nk=0 ak bn−k ; furthermore, if the two series f (z) and g (z) converge absolutely, the
series ∑∞ n
n=0 cn (z − z0 ) converges absolutely;

(4) if

(a) b0 6= 0 ,
(b) the series h (z) = ∑∞ n
n=0 dn (z − z0 ) where the coefficients dn are obtained by solving the
equations ∑k=0 ak bn−k = an , n = 0, 1, . . . , converges,
n

(c) g (z) or h (z) converges absolutely,

(d) g (z) 6= 0 ,

then ∞
f (z)
= ∑ dn (z − z0 )n . (4.2)
g (z) n=0

4.2 Theorem C ONVERGENCE IN A CIRCLE. Let f (z) = ∑∞ n

n=0 an (z − z0 ) and g (z) =
∑∞
n
n=0 bn (z − z0 ) be two power series whose convergence radii are R1 and R2 respectively, where
R1 > 0 and R2 > 0. Then

(1) for any c ∈ C, the series ∑∞ n

n=0 can (z − z0 ) converges absolutely for |z − z0 | < R1 and

∞
∑ c an (z − z0 )n = c f (z) for |z − z0 | < R1 ; (4.3)
n=0

(2) the series ∑∞ n

n=0 (an + bn ) (z − z0 ) converges absolutely for |z − z0 | < min {R1 , R2 } and

∞
∑ (an + bn ) (z − z0 )n = f (z) + g (z) for |z − z0 | < min {R1 , R2 } ; (4.4)
n=0

(3) the series ∑∞ n

n=0 cn (z − z0 ) , where cn = ∑k=0 ak bn−k , converges absolutely for |z − z0 | <
n

min {R1 , R2 } , and

∞
∑ cn (z − z0 )n = f (z) g (z) , for |z − z0 | < min {R1 , R2 } ; (4.5)
n=0
5. SINGULARITIES 11

(4) if g (z) 6= 0 for |z − z0 | < R, where 0 < R ≤ min {R1 , R2 } , and {dn }∞
n=0 is the sequence of coef-
ficients obtained by solving the equations
n
∑ dk bn−k = an , n = 0, 1, . . . , (4.6)
k=0

then the series ∑∞ n

n=0 dn (z − z0 ) converges absolutely for |z − z0 | < R, and

∞
∑ dn (z − z0 )n = f (z) /g (z) , for |z − z0 | < R; (4.7)
n=0

when g (z0 ) = b0 6= 0, the coefficients dn are unique and there exists a radius R > 0 such that
g (z) 6= 0 for |z − z0 | < R.

4.3 Theorem M AC L AURIN SERIES FOR A RATIONAL FUNCTION. If

f (z) = P (z) /Q (z) (4.8)

p
where P (z) = ∑n=0 an zn and Q (z) = ∑qn=0 an zn are polynomials of degree p and q respectively, and
Q (0) 6= 0, then
∞
f (z) = ∑ dn zn , for |z| < R ,
n=0

where R = min |z∗1 | , . . . , z∗q > 0, z∗1 , . . . , z∗q are the roots (possibly non distinct) of polynomial


Q (z) and the coefficients dn are obtained by solving the equations

n
∑ dk bn−k = an , n = 0, 1, . . . , (4.9)
k=0

with an ≡ 0 for n > p and bn ≡ 0 for n > q. Further, the series ∑∞

n=0 dn z converges absolutely for
n

|z| < R.

5. Singularities
5.1 Definition P OLE AND ESSENTIAL SINGULARITY. Let f be an analytic function on the disk
0 < |z − z0 | < R. We say that f has a pole at point z0 if lim | f (z)| = ∞. If the point z0 is a singular
z→z0
point which is neither removable nor a pole, we say that it is an essential singular point.

5.2 Theorem C HARACTERIZATION OF ISOLATED SINGULARITIES. Let f be an analytic function

with an isolated singular point at z0 . Then

(1) z0 is a removable singular point

⇔ lim (z − z0 ) f (z) = 0
z→z0
5. SINGULARITIES 12

⇔ lim f (z) = c , for some c ∈ C .

z→z0

(2) z0 is a pole

⇔ the function g (z) = 1/ f (z) has a removable singular point at z0

⇔ there is a positive integer m (m > 0) and an analytic function h (z) on a disk |z − z0 | < R,
where R > 0, such that h (z0 ) 6= 0 and f (z) = h (z) / (z − z0 )m
⇔ there is a positive integer m such that lim (z − z0 )m f (z) = c, where c ∈ C
z→z0

⇔ there is a positive integer m such that the function g (z) = (z − z0 )m f (z) has a removable
singular point at z0 .

5.3 Definition O RDER OF A POLE. If z0 is a pole of the function f such that

lim (z − z0 )m f (z) = c 6= 0, for some c ∈ C ,

z→z0

we say that z0 is a pole of order m.

5.4 Theorem S INGULARITIES AND L AURENT SERIES. Let f be an analytic function with an
isolated singular point at z0 with Laurent series is
∞
f (z) = ∑ an (z − z0 )n for 0 < |z − z0 | < R . (5.1)
n=−∞

Then

(1) z0 is a removable singular point ⇔ an = 0, ∀n < 0 ;

(2) z0 is a pole of order m ⇔ a−m 6= 0 and an = 0 for n < −m ;

(3) z0 is an essential singular point

⇔ an 6= 0 for an infinite number of negative values of n.

5.5 Theorem B EHAVIOR OF AN ANALYTIC FUNCTION NEAR AN ESSENTIAL SINGULARITY (P I -

CARD ) . Let f be an analytic function on the disk 0 < |z − z0 | < R. If z0 is an essential singular
point, then for any complex number c ∈ C, except possibly one, there exists a sequence {zn }∞ n=1
converging to z0 such that f (zn ) = c, ∀n.

5.6 Remark Picard’s theorem means that in any neighborhood of z0 and for any complex number
c (except possibly one), the function f takes the value c an infinite number of times.
6. PARTIAL FRACTIONS 13

6. Partial fractions
6.1 Theorem PARTIAL FRACTION EXPANSION OF RATIONAL FUNCTIONS. Consider the rational
p
function f (z) = P (z) /Q (z) where P (z) = ∑n=0 an zn is a polynomial of degree p (a p 6= c) and
Q (z) = (z − z1 ) (z − z2 ) · · · (z − zq ) is a polynomial of degree q∗ = ∑qj=1 m j with q distinct
m1 m2 mq

roots z1 , . . . , zq of multiplicities m1 , . . . , mq respectively (q ≥ 1, m j ≥ 1 for j = 1, . . . , q). Then the

function f (z) can be uniquely written in the form
q
f (z) = G (z) + ∑ G j [1/ (z − z j )]
j=1

for any z ∈ C such that z 6= z j , j = 1, . . . , q, where

mj
G j [1/ (z − z j )] = ∑ A jk / (z − z j )k ,
k=1

A jk ∈ C, and G (z) is a polynomial. Further,

(1) if p < q∗ , G (z) ≡ 0,

(2) if p ≥ q∗ and the polynomials P (z) and Q (z) have no common root, the degree of G (z) is
p − q∗ .

6.2 Theorem FACTORIZATION OF AN ANALYTIC FUNCTION WITH FINITE NUMBER OF POLES.

Let f be an analytic function everywhere on an open domain U ⊆ C except at a finite number
of singular points z1 , . . . , zq which are poles of orders m1 , . . . , mq respectively (q ≥ 1, m j ≥ 1 for
j = 1, . . . , q). Then there exists a function g (z) analytic everywhere on U such that g (z j ) 6= 0, j =
1, . . . , p, and
f (z) = g (z) / [(z − z1 )m1 (z − z2 )m2 · · · (z − zq )mq ]
for z ∈ U and z 6= z j , j = 1, . . . , q. If, furthermore, the function f has a finite number of zeros, the
function f can be written
P (z)
f (z) = h (z)
Q (z)
for z ∈ U and z 6= z j , j = 1, . . . , q, where P (z) and Q (z) are polynomials with no common root,
Q (z) = (z − z1 )m1 (z − z2 )m2 · · · (z − zq )mq and h (z) 6= 0 for z ∈ U.

6.3 Theorem PARTIAL FRACTION EXPANSION OF AN ANALYTIC FUNCTION WITH FINITE NUM -
BER OF POLES . Let f be an analytic function everywhere on an open domain U ⊆ C except at a
finite finite number of singular points z1 , . . . , zq which are poles of orders m1 , . . . , mq (q ≥ 1, m j ≥ 1
for j = 1, . . . , q). Then the function f can be written in a unique way in the form
q
f (z) = g (z) + ∑ G j [1/ (z − z j )]
j=1
7. PROOFS AND REFERENCES 14

for any z ∈ U such that z 6= z j , j = 1, . . . , q, where

mj
G j [1/ (z − z j )] = ∑ A jk / (z − z j )k ,
k=1

A jk ∈ C, and G (z) is analytic everywhere on U.

7. Proofs and references

1. Churchill and Brown (1984, chapters 2 and 3) and Ahlfors (1979, chapter 2).
1.2. Ahlfors (1979, section 1.1, p. 22).
1.4. Ahlfors (1979, section 1.1, p. 23).
1.5. Ahlfors (1979, section 1.1, p. 23).
1.7. Ahlfors (1979, section 1.2, p. 24).
1.8 - 1.9. Churchill and Brown (1984, section 15, p. 41, and section 21, pp. 57-58).
1.10. Churchill and Brown (1984, section 15, p. 41 et section 21, pp. 57-58).
1.11. Ahlfors (1979, section 1.1, p. 23).
1.12. Churchill and Brown (1984, section 19, p. 50).
1.14. Churchill and Brown (1984, section 54, p. 156).
1.15 - 1.16. Churchill and Brown (1984, section 19, p. 51).
1.17. Churchill and Brown (1984, section 39, pp. 111-114).
1.18. Churchill and Brown (1984, section 9, p. 27, and Chapter 3).
2. Ahlfors (1979, Chapter 2), Churchill and Brown (1984, Chapter 5), and Rudin (1976, Chap-
ter 3).
2.3. Ahlfors (1979, section 2.4, p. 38) and Rudin (1976, section 3.39, p. 69).
2.6. Wilf (1990, section 2.4, Theorem 2.4.2, p. 44).
2.7 - 2.8. Deshpande (1986, section 6.1, pp. 62-64).
2.9. Rudin (1976, section 3.44, p. 71).
2.10. Ahlfors (1979, section 2.5, pp. 41-42).
2.11 - 2.13. Devinatz (1968, section 4.5, pp. 170-171).
2.14. Gillert, Küstner, Kellwich and Kästner (1986, section 21.2, p. 527).
2.16. Ahlfors (1979, section 2.4, p. 38), Churchill and Brown (1984, section 49, p. 144) and
Wilf (1990, section 2.4, Theorem 2.4.2, pp. 44-45).
2.18. Churchill and Brown (1984, section 50, pp. 146-147).
2.19. Churchill and Brown (1984, section 44, pp. 126-128, and section 39,).
3.1. Churchill and Brown (1984, section 44, pp. 126-128, and section 39, pp. 111-114.).
3.3. Silverman (1974, section 10.1, p. 139).
3.5. This is a direct consequence of Theorem 3.1.
3.6. Churchill and Brown (1984, section 53, pp. 152-153).
3.8. Deshpande (1986, section 10.2, p. 139).
3.10 - 3.12. Deshpande (1986, section 10.1, Propositions 10.3 and 10.4, p. 135).
3.14. Churchill and Brown (1984, sections 46 and 50, pp. 132-136 and 146-148).
7. PROOFS AND REFERENCES 15

3.16. Silverman (1974, section 11.1, p. 158).

4. Churchill and Brown (1984, Section 51, pp. 148-153) and Deshpande (1986, Section 6.1).
4.1. Statement (3) is a consequence of the Cauchy-Mertens theorem; see Devinatz (1968,
Section 4.5, pp. 168-169). Statement (4) is a consequence of (3).
5. Deshpande (1986, Chapter 12).
5.1. Deshpande (1986, Section 12.2, Propositions 12.2 - 12.3, pp. 154-155).
5.2 - 5.3. Deshpande (1986, section 10.2, p. 139).
5.4. Deshpande (1986, Section 12.3, Proposition 12.9, p. 163).
5.5. Silverman (1974, Section 57, p. 241) and Churchill and Brown (1984, Section 56, p.
161).
6.1. Ahlfors (1979, Section 1.4, pp. 31-32) and Lentin and Rivaud (1964, chapitre II, section
19, pp. 234-238).
6.2. Deshpande (1986, Section 13.1, Proposition 13.1, pp. 169-170).
6.3. Deshpande (1986, Section 13.1, p. 171).

Other useful references include: Cartan (1961), Gillert et al. (1986), Gradshteyn and Ryzhik
(1980), Knopp (1956), Knopp (1990), Rudin (1987), Silverman (1972), Spiegel (1964).
REFERENCES 16

References
Ahlfors, L. V. (1979), Complex Analysis: An Introduction to the Theory of Analytic Functions
of One Complex Variable, International Series in Pure and Applied Mathematics, third edn,
McGraw-Hill, New York.

Cartan, H. (1961), Théorie élémentaire des fonctions analytiques d’une ou plusieurs variables com-
plexes, Hermann, Paris.

Churchill, R. V. and Brown, J. W. (1984), Complex Variables and Applications, fourth edn,
McGraw-Hill, New York.

Deshpande, J. V. (1986), Complex Analysis, Tata McGraw-Hill Publishing Company, New Delhi,
India.

Devinatz, A. (1968), Advanced Calculus, Holt, Rinehart and Winston, New York.

Gillert, W., Küstner, Kellwich, H. and Kästner, H. (1986), Petite encyclopédie des mathématiques,
Éditions K. Pagoulatos, Paris and Athens.

Gradshteyn, I. S. and Ryzhik, I. M. (1980), Table of Integrals, Series, and Products. Corrected and
Enlarged Edition, Academic Press, New York.

Knopp, K. (1956), Infinite Sequences and Series, Dover Publications, New York.

Knopp, K. (1990), Theory and Application of Infinite Series, Dover Publications, New York.

Lentin, A. and Rivaud, J. (1964), Leçons d’algèbre moderne, second edn, Vuibert, Paris.

Rudin, W. (1976), Principles of Mathematical Analysis, Third Edition, McGraw-Hill, New York.

Rudin, W. (1987), Real and Complex Analysis, third edn, McGraw-Hill, New York.

Silverman, R. A. (1972), Introductory Complex Analysis, Dover Publications, New York.

Silverman, R. A. (1974), Complex Analysis with Applications, Dover Publications, New York.

Spiegel, M. R. (1964), Theory and Problems of Complex Variables with an Introduction to Confor-
mal Mapping and its Applications, Schaum Publishing Company, New York.

Wilf, H. S. (1990), Generatingfunctionology, Academic Press, New York.