Continuous functions

A function of the complex variable 𝑧, written as 𝑤 = 𝑓(𝑧), is a rule that assigns a complex
number w to each complex number 𝑧 in a given subset 𝐷 of the complex plane. The set 𝐷 is
called the domain of definition of the function. The collection of all possible values 𝑤 of the
function is called the range of the function.
A function 𝑤 = 𝑓(𝑧) is called one-to-one on a set 𝐷 if from 𝑓(𝑧1 ) = 𝑓(𝑧2 ) with 𝑧1 , 𝑧2 ∈ 𝐷 it
must follow 𝑧1 = 𝑧2 . A function 𝑤 = 𝑓(𝑧) is called onto a set ℝ if is a subset of the range of this
function.
Let 𝑓 be a function defined on a set Ω of complex numbers. We say that 𝑓 is continuous at
the point 𝑧0 ∈ Ω if for every 𝜖 > 0 there exists δ > 0 such that whenever 𝑧 ∈ Ω and |𝑧 − 𝑧0 | < 𝛿
then |𝑓(𝑧) − 𝑓(𝑧0 )| < 𝜖. An equivalent definition is that for every sequence {𝑧1 , 𝑧2 , … } ⊂ Ω
such that lim 𝑧𝑛 = 𝑧0, then lim 𝑓(𝑧𝑛 ) = 𝑓(𝑧0 ).
The function 𝑓 is said to be continuous on Ω if it is continuous at every point of Ω. Sums and
products of continuous functions are also continuous.
Since the notions of convergence for complex numbers and points in ℝ2 are the same, the
function 𝑓 of the complex argument 𝑧 = 𝑥 + 𝑖𝑦 is continuous if and only if it is continuous
viewed as a function of the two real variables 𝑥 and 𝑦.
By the triangle inequality, it is immediate that if 𝑓 is continuous, then the real-valued function
defined by 𝑧 → |𝑓(𝑧)| is continuous. We say that 𝑓 attains a maximum at the point 𝑧0 ∈ Ω if
|𝑓(𝑧)| ≤ |𝑓(𝑧0 )| for all 𝑧 ∈ Ω,
with the inequality reversed for the definition of a minimum.
Theorem 2.1 A continuous function on a compact set Ω is bounded and attains a maximum and
minimum on Ω. This is of course analogous to the situation of functions of a real variable, and
we shall not repeat the simple proof here.

Holomorphic functions
We now present a notion that is central to complex analysis, and in distinction to our previous
discussion we introduce a definition that is genuinely complex in nature.
Let Ω be an open set in ℂ and 𝑓 a complex-valued function on Ω. The function 𝑓 is
holomorphic at the point 𝑧0 ∈ Ω if the quotient
𝑓(𝑧0 + ℎ) − 𝑓(𝑧0 )
(1)
ℎ
converges to a limit when ℎ → 0. Here ℎ ∈ ℂ and ℎ ≠ 0 with 𝑧0 + ℎ ∈ Ω, so that the quotient
is well defined. The limit of the quotient, when it exists, is denoted by 𝑓 ′ (𝑧0 ), and is called the
derivative of 𝒇 at 𝑧0 :
𝑓(𝑧0 + ℎ) − 𝑓(𝑧0 )
𝑓 ′ (𝑧0 ) = lim .
ℎ→0 ℎ

It should be emphasized that in the above limit, ℎ is a complex

number that may approach 0 from any direction. The function 𝑓 is
said to be holomorphic on Ω if 𝑓 is holomorphic at every point of Ω.
If ℂ is a closed subset of ℂ, we say that f is holomorphic on ℂ if f is
holomorphic in some open set containing ℂ. Finally, if f is
holomorphic in all of ℂ we say that 𝑓 is entire. Sometimes the terms
regular or complex differentiable are used instead of holomorphic.
The latter is natural in view of (1) which mimics the usual definition
of the derivative of a function of one real variable. But despite this
resemblance, a holomorphic function of one complex variable will
satisfy much stronger properties than a differentiable function of one
real variable. For example, a holomorphic function will actually be
infinitely many times complex differentiable, that is, the existence of
the first derivative will guarantee the existence of derivatives of any
order. This is in contrast with functions of one real variable, since
there are differentiable functions that do not have two derivatives.
In fact more is true: every holomorphic function is analytic, in the
sense that it has a power series expansion near every point (power
series will be discussed in the next section), and for this reason we
also use the term analytic as a synonym for holomorphic. Again, this is in contrast with the fact
that there are indefinitely differentiable functions of one real variable that cannot be expanded
in a power series.

Example 1. The function 𝑓(𝑧) = 𝑧 is holomorphic on any open set in ℂ, and 𝑓 ′ (𝑧) = 1. In fact,
any polynomial
𝑝(𝑧) = 𝑎0 + 𝑎1 𝑧 + 𝑎2 𝑧 2 + ⋯ + 𝑎𝑛 𝑧 𝑛
is holomorphic in the entire complex plane and
𝑝′ (𝑧) = 𝑎1 + ⋯ + 𝑛𝑎𝑛 𝑧 𝑛−1 .
Example 2. The function 1/𝑧 is holomorphic on any open set in ℂ that does not contain the
origin, and 𝑓 ′ (𝑧) = −1/𝑧 2. Example 3. The function 𝑓(𝑧) = 𝑧 ) is not holomorphic. Indeed, we
have
𝑓(𝑧𝑛 + ℎ) − 𝑓(𝑧0 ) ℎ̅
=
ℎ ℎ
which has no limit as ℎ → ∞, as one can see by first taking ℎ real and then ℎ purely imaginary.
An important family of examples of holomorphic functions, which we discuss later, are the
power series. They contain functions such as 𝑒 𝑧 , sin 𝑧, or cos 𝑧, and in fact power series play a
crucial role in the theory of holomorphic functions, as we already mentioned in the last
paragraph. Some other examples of holomorphic functions that will make their appearance in
later chapters were given in the introduction to this book.
It is clear that a function 𝑓 is holomorphic at 𝑧0 ∈ Ω if and only if there exists a complex
number 𝑎 such that
𝑓(𝑧0 + ℎ) − 𝑓(𝑧0 ) − 𝑎ℎ = ℎ𝜓(ℎ), (2)
where 𝜓 is a function defined for all small ℎ and lim 𝜓(ℎ) = 0. Of course, we have 𝑎 = 𝑓 ′ (𝑧0 ).
ℎ→0
From this formulation, it is clear that 𝑓 is continuous wherever it is holomorphic. Arguing as in
the case of one real variable, using formulation (2) in the case of the chain rule (for example),
one proves easily the following desirable properties of holomorphic functions.

Proposition 2.2 If 𝑓 and 𝑔 are holomorphic in Ω, then:

I. 𝑓 + 𝑔 is holomorphic in Ω and (𝑓 + 𝑔)′ = 𝑓 ′ + 𝑔′.
II. 𝑓𝑔 is holomorphic in Ω and (𝑓𝑔)′ = 𝑓 ′ 𝑔 + 𝑓 ′ 𝑔.
𝑓 ′𝑔−𝑓𝑔′
III. If 𝑔(𝑧0 ) = 0, then 𝑓/𝑔 is holomorphic at 𝑧0 and (𝑓 ⁄𝑔)′ = .
𝑔2

Moreover, if 𝑓 ∶ Ω → 𝑈 and 𝑔 ∶ 𝑈 → ℂ are holomorphic, the chain rule holds

(𝑓 ∘ 𝑔)′ (𝑧) = 𝑔 ′ (𝑓(𝑧))𝑓 ′ (𝑧) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑧 ∈ Ω
Complex-valued functions as mappings
We now clarify the relationship between the complex and real derivatives. In fact, the third
example above should convince the reader that the notion of complex differentiability differs
significantly from the usual notion of real differentiability of a function of two real variables.
Indeed, in terms of real variables, the function 𝑓(𝑧) = 𝑧 corresponds to the map 𝐹 ∶ (𝑥, 𝑦) →
(𝑥, −𝑦), which is differentiable in the real sense. Its derivative at a point is the linear map given
by its Jacobian, the 2 × 2 matrix of partial derivatives of the coordinate functions. In fact, 𝐹 is
linear and is therefore equal to its derivative at every point. This implies that 𝐹 is actually
indefinitely differentiable. In particular the existence of the real derivative need not guarantee
that 𝑓 is holomorphic.
This example leads us to associate more generally to each complexvalued function 𝑓 = 𝑢 +
𝑖𝑣 the mapping 𝐹(𝑥, 𝑦) = (𝑢(𝑥, 𝑦), 𝑣(𝑥, 𝑦)) from ℝ2 to ℝ2 . Recall that a function 𝐹(𝑥, 𝑦) =
(𝑢(𝑥, 𝑦), 𝑣(𝑥, 𝑦)) is said to be differentiable at a point 𝑃0 = (𝑥0 , 𝑦0 ) if there exists a linear
transformation 𝐽 ∶ ℝ2 → ℝ2 such that
|𝐹(𝑃0 + 𝐻) − 𝐹(𝑃0 ) − 𝐽(𝐻)|
→ 0 𝑎𝑠 |𝐻| → 0, 𝐻 ∈ ℝ2 . (3)
|𝐻|
Equivalently, we can write
𝐹(𝑃0 + 𝐻) − 𝐹(𝑃0 ) = 𝐽(𝐻) + |𝐻|Ψ(𝐻).
With |Ψ(𝐻)| → 0 as |𝐻| → 0. The linear transformation 𝐽 is unique and is called the derivative
of 𝐹 at 𝑃0 . If 𝐹 is differentiable, the partial derivatives of 𝑢 and 𝑣 exist, and the linear
transformation 𝐽 is described in the standard basis of ℝ2 by the Jacobian matrix of 𝐹

𝜕𝑢 𝜕𝑢
𝜕𝑥 𝜕𝑦
𝐽 = 𝐽𝐹 (𝑥, 𝑦) =
𝜕𝑣 𝜕𝑣
(𝜕𝑥 𝜕𝑦)

In the case of complex differentiation the derivative is a complex number 𝑓 ′ (𝑧0 ), while in the
case of real derivatives, it is a matrix. There is, however, a connection between these two
notions, which is given in terms of special relations that are satisfied by the entries of the
Jacobian matrix, that is, the partials of 𝑢 and 𝑣. To find these relations, consider the limit in (1)
when ℎ is first real, say ℎ = ℎ1 + 𝑖ℎ2 with ℎ2 = 0. Then, if we write 𝑧 = 𝑥 + 𝑖𝑦, 𝑧0 = 𝑥0 + 𝑖𝑦0,
and 𝑓(𝑧) = 𝑓(𝑥, 𝑦), we find that
𝑓(𝑥0 + ℎ1 , 𝑦0 ) − 𝑓(𝑥0 , 𝑦0 ) 𝜕𝑓
𝑓 ′ (𝑧0 ) = lim = (𝑧 ),
ℎ1 →0 ℎ1 𝜕𝑥 0
𝜕
where denotes the usual partial derivative in the 𝑥 variable. (We fix 𝑦0 and think of 𝑓 as a
𝜕𝑥
complex-valued function of the single real variable 𝑥.) Now taking h purely imaginary, say ℎ =
𝑖ℎ2, a similar argument yield
𝑓(𝑥0 , 𝑦0 + ℎ2 ) − 𝑓(𝑥0 , 𝑦0 ) 1 𝜕𝑓
𝑓 ′ (𝑧0 ) = lim = (𝑧 ),
ℎ2 →0 𝑖ℎ2 𝑖 𝜕𝑦 0
where 𝜕/𝜕𝑦 is partial differentiation in the 𝑦 variable. Therefore, if 𝑓 is holomorphic we have
shown that
 𝜕𝑓 1 𝜕𝑓
= .
𝜕𝑥 𝑖 𝜕𝑥
1
Writing 𝑓 = 𝑢 + 𝑖𝑣, we find after separating real and imaginary parts and using = −𝑖, that
𝑖
the partials of 𝑢 and 𝑣 exist, and they satisfy the following non-trivial relations
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= 𝑎𝑛𝑑 =−
𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥
These are the Cauchy-Riemann equations, which link real and complex analysis.
We can clarify the situation further by defining two differential operators
𝜕 1 𝜕 𝜕 𝜕 1 𝜕 1 𝜕
= ( +𝑖 ) 𝑎𝑛𝑑 = ( − ).
𝜕𝑧 2 𝜕𝑥 𝜕𝑦 𝜕𝑧̅ 2 𝜕𝑥 𝑖 𝜕𝑦
Proposition 2.3 If 𝑓 is holomorphic at 𝑧0 , then
𝜕𝑓 𝜕𝑓 𝜕𝑢
(𝑧 ) = 0 𝑎𝑛𝑑 𝑓 ′ (𝑧0) = (𝑧0 ) = 2 (𝑧0 ).
𝜕𝑧 0 𝜕𝑧 𝜕𝑧
Also, if we write 𝐹(𝑥, 𝑦) = 𝑓(𝑧), then 𝐹 is differentiable in the sense of real variables, and
det 𝐽𝐹 (𝑥0 , 𝑦0 ) = |𝑓 ′ (𝑧0 )|2
Proof. Taking real and imaginary parts, it is easy to see that the Cauchy-Riemann equations are
𝜕𝑓
equivalent to = 0. Moreover, by our earlier observation
𝜕𝑧

1 𝜕𝑓 1 𝜕𝑓 𝜕𝑓
𝑓 ′ (𝑧0 ) = ( (𝑧0 ) + (𝑧0 )) = (𝑧 ),
2 𝜕𝑥 𝑖 𝜕𝑦 𝜕𝑧 0
𝜕𝑓 𝜕𝑢
and the Cauchy-Riemann equations give =2 . To prove that 𝐹 is differentiable it suffices
𝜕𝑧 𝜕𝑧
to observe that if 𝐻 = (ℎ1 , ℎ2 ) and ℎ = ℎ1 + 𝑖ℎ2 , then the Cauchy-Riemann equations imply
𝜕𝑢 𝜕𝑢
𝐽𝐹 (𝑥0, 𝑦0 )(𝐻) = ( − 𝑖 ) (ℎ1 + 𝑖ℎ2 ) = 𝑓 ′ (𝑧0 )ℎ,
𝜕𝑥 𝜕𝑦
where we have identified a complex number with the pair of real and imaginary parts. After a
final application of the Cauchy-Riemann equa tions, the above results imply that
𝜕𝑢 𝜕𝑣 𝜕𝑣 𝜕𝑢 𝜕𝑢 2 𝜕𝑢 2 𝜕𝑢 2
det 𝐽𝐹 (𝑥0 , 𝑦0 ) = − = ( ) + ( ) = |2 | = |𝑓 ′ (𝑧0 )|2 . (4)
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑧
So far, we have assumed that 𝑓 is holomorphic and deduced relations satisfied by its real and
imaginary parts. The next theorem contains an important converse, which completes the circle
of ideas presented here.
Theorem 2.4 Suppose 𝑓 = 𝑢 + 𝑖𝑣 is a complex-valued function defined on an open set Ω. If 𝑢
and 𝑣 are continuously differentiable and satisfy the Cauchy-Riemann equations on Ω, then 𝑓 is
𝜕𝑓
holomorphic on Ω and 𝑓 ′ (𝑧) = .
𝜕𝑧

Proof. Write
𝜕𝑢 𝜕𝑢
𝑢(𝑥 + ℎ1 , 𝑦 + ℎ2 ) − 𝑢(𝑥, 𝑦) = ℎ1 + ℎ + |ℎ|𝜓1 (ℎ)
𝜕𝑥 𝜕𝑦 2
And
𝜕𝑣 𝜕𝑣
𝑣(𝑥 + ℎ1 , 𝑦 + ℎ2 ) − 𝑣(𝑥, 𝑦) = ℎ1 + ℎ + |ℎ|𝜓2 (ℎ),
𝜕𝑥 𝜕𝑦 2
Where 𝜓𝑗 (ℎ) → 0 (for 𝑗 = 1, 2, … ) as |ℎ| tends to 0, and ℎ = ℎ1 + 𝑖ℎ2 . Using the Cauchy-
Riemann equations we find that
𝜕𝑢 𝜕𝑢
𝑓(𝑧 + ℎ) − 𝑓(𝑧) = ( − 𝑖 ) (ℎ1 + 𝑖ℎ2 ) + |ℎ|𝜓(ℎ),
𝜕𝑥 𝜕𝑦
Where 𝜓(ℎ) = 𝜓1 (ℎ) + 𝜓2 (ℎ) → 0 , as |ℎ| → 0. Therefore 𝑓 is holomorphic and
𝜕𝑢 𝜕𝑓
𝑓 ′ (𝑧) = 2 = .
𝜕𝑧 𝜕𝑧