MHZ4553: Unit II Session 5: Implicit Functions

Session 5
Implicit Functions

Contents:
Introduction, p 64
5.1 Implicit functions, p 65
Solutions of Activities, p 69
Summary, p 69
Learning Outcomes, p 70

Introduction

Suppose 𝑧 = 𝑓(𝑥, 𝑦) is a function of 𝑥 and 𝑦 satisfying 𝑥 2 + 𝑦 2 + 𝑧 2 = 1

on some set 𝑆 ⊆ ℝ2 . It is evident that, the way how 𝑧 depends on 𝑥 and 𝑦 is
not readily seen from the equation. In a situation like this we say that 𝑧 is
defined implicitly. In this particular case, however, it is possible to solve the
equation 𝐹(𝑥, 𝑦, 𝑧) = 𝑥 2 + 𝑦 2 + 𝑧 2 − 1 = 0 for 𝑧 in terms of 𝑥 and 𝑦. This
gives 𝑧 = √1 − 𝑥 2 − 𝑦 2 and 𝑧 = −√1 − 𝑥 2 − 𝑦 2 .

But, there are situations in which obtaining an explicit formula for 𝑧 in

terms of 𝑥 and 𝑦 (such as 𝑧 = √1 − 𝑥 2 − 𝑦 2 and 𝑧 = −√1 − 𝑥 2 − 𝑦 2 in
the previous example) is not that easy. For example, there is no easy method
for solving for 𝑧 in the equation 𝑥 2 + 𝑥𝑦𝑧 + 𝑧 2 − 𝑒 𝑧 = 0.

In this session we are going to discuss about partial derivatives of functions

defined implicitly.

64
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

5.1 Implicit functions

5.1.1 Differentiation of Implicit functions

Theorem 5.1

Let 𝑇 ⊆ ℝ𝑛 , 𝑛 ≥ 2, be open and let 𝐹: 𝑇 → ℝ be such that 𝐹 is

differentiable on 𝑇. Assume that the equation 𝐹(𝑥1 , … , 𝑥𝑛 ) = 0
defines 𝑥𝑛 implicitly as a differentiable function of 𝑥1 , … , 𝑥𝑛−1, say
𝑥𝑛 = 𝑓(𝑥1 , … , 𝑥𝑛−1 ),
for all points (𝑥1 , … , 𝑥𝑛−1 ) in some open set 𝑆 ⊆ ℝ𝑛−1
(i.e. 𝑓 is differentiable on 𝑆 and 𝐹(𝑥1 , … , 𝑥𝑛−1 , 𝑓(𝑥1 , … , 𝑥𝑛−1 )) =
0 for each point (𝑥1 , … , 𝑥𝑛−1 ) ∈ 𝑆).
𝜕𝑓
Then for each 𝑘 = 1, 2, … , 𝑛 − 1, the partial derivative is given
𝜕𝑥𝑘
𝜕𝐹
𝜕𝑓 𝜕𝑥𝑘
by the formula =− 𝜕𝐹
𝜕𝑥𝑘
𝜕𝑥𝑛

𝜕𝐹
at those points where ≠ 0.
𝜕𝑥𝑛
𝜕𝐹 𝜕𝐹
The partial derivatives 𝜕𝑥 and 𝜕𝑥 which appear in the above
𝑘 𝑛

formula are to be evaluated at the point

(𝑥1 , 𝑥2 … , 𝑥𝑛−1 , 𝑓(𝑥1 , … , 𝑥𝑛−1 )).

Remark
Notice that this result allows us to find the partial derivative of 𝑥𝑛 =
𝜕𝑓
𝑓(𝑥1 , … , 𝑥𝑛−1 ) with respect to 𝑥𝑘 , i. e. 𝜕𝑥 , without expressing 𝑥𝑛 explicitly
𝑘

in terms of the remaining variables 𝑥1 , 𝑥2 … , 𝑥𝑛−1 .

Example 5.1
Let 𝑇 ⊆ ℝ3 and 𝑆 ⊆ ℝ2 be open sets and let 𝐹: 𝑇 → ℝ be a
differentiable function on 𝑇. Let 𝑧 = 𝑓(𝑥, 𝑦) be a differentiable function
which is defined implicitly by the equation 𝐹(𝑥, 𝑦, 𝑧) = 0 for each
(𝑥, 𝑦) ∈ 𝑆.

65
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

Prove that
𝜕𝐹 𝜕𝐹
𝜕𝑧 𝜕𝑥 𝜕𝑧 𝜕𝑦
=− 𝜕𝐹 and =− 𝜕𝐹
𝜕𝑥 𝜕𝑦
𝜕𝑧 𝜕𝑧
𝜕𝐹
at points where 𝜕𝑍 ≠ 0.

Solution
𝜕𝐹 𝜕𝐹 𝜕𝐹
Notice that since 𝐹 is differentiable on 𝑇, , and exist at each
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝜕𝑧 𝜕𝑧
point in 𝑇. Also, since 𝑧 is differentiable on 𝑆, and exist at each
𝜕𝑥 𝜕𝑦

point in 𝑆.
For each (𝑥, 𝑦) ∈ 𝑆, let 𝑔(𝑥, 𝑦) = 𝐹(𝑥, 𝑦, 𝑓(𝑥, 𝑦)). Then 𝑔(𝑥, 𝑦) = 0 on
𝑆.

𝜕𝑔 𝜕𝑔
Therefore = = 0 on 𝑆.
𝜕𝑥 𝜕𝑦

Now by chain rule,

𝜕𝑔 𝜕𝐹 𝜕𝑥 𝜕𝐹 𝜕𝑦 𝜕𝐹 𝜕𝑧
= ⋅ + ⋅ + ⋅ and
𝜕𝑥 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑥

𝜕𝑔 𝜕𝐹 𝜕𝑥 𝜕𝐹 𝜕𝑦 𝜕𝐹 𝜕𝑧
= ⋅ + ⋅ + ⋅ .
𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑦

𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥
However, = 𝜕𝑦 = 1 and 𝜕𝑥 = 𝜕𝑦 = 0.
𝜕𝑥

Hence,
𝜕𝑔 𝜕𝐹 𝜕𝐹 𝜕𝑧
0= = + ⋅ and
𝜕𝑥 𝜕𝑥 𝜕𝑧 𝜕𝑥

𝜕𝑔 𝜕𝐹 𝜕𝐹 𝜕𝑧
0= = + ⋅ .
𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑦

66
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

𝜕𝐹
Now, at those points where (𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ≠ 0 we get
𝜕𝑧
𝜕𝐹
𝜕𝑧
= − 𝜕𝑥 and
𝜕𝑥 𝜕𝐹
𝜕𝑧

𝜕𝐹
𝜕𝑧 𝜕𝑦
=− .
𝜕𝑦 𝜕𝐹
𝜕𝑧

Activity 5.1

Suppose that the equation 𝐹(𝑥, 𝑦) = 0 defines 𝑦 implicitly as a differentiable function of

𝑥, say 𝑦 = 𝑓(𝑥), for all 𝑥 in some open interval 𝐼. Suppose also that 𝐹 is differentiable
on an open set in ℝ2 containing the set {(𝑥, 𝑓(𝑥)): 𝑥 ∈ 𝐼}.

Show that
𝜕𝐹
𝑑𝑦
= − 𝜕𝑥
𝑑𝑥 𝜕𝐹
𝜕𝑦
𝜕𝐹
for each 𝑥 ∈ 𝐼, such that 𝜕𝑦 ≠ 0.

Example 5.2
Suppose 𝑥 2 𝑦 + 𝑦 2 𝑥 + 𝑧 2 − 𝑥𝑦𝑧 = 2 defines 𝑧 implicitly as a function
𝜕𝑧 𝜕𝑧 𝜕𝑧
of 𝑥 and 𝑦. Assuming that 𝜕𝑥 and 𝜕𝑦 exist, find expressions for 𝜕𝑥 and
𝜕𝑧
.
𝜕𝑦

Solution
Let 𝐹(𝑥, 𝑦, 𝑧) = 𝑥 2 𝑦 + 𝑦 2 𝑥 + 𝑧 2 − 𝑥𝑦𝑧 − 2. Clearly 𝐹 is defined for
each (𝑥, 𝑦, 𝑧) ∈ ℝ3 and 𝐹 is differentiable on ℝ3 . Since 𝑧 is a function
of 𝑥 and 𝑦, we write 𝑧 = 𝑓(𝑥, 𝑦).

67
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

For each (𝑥, 𝑦) ∈ Domn(𝑓), let 𝑔(𝑥, 𝑦) = 𝐹(𝑥, 𝑦, 𝑓(𝑥, 𝑦)).

Then 𝑔(𝑥, 𝑦) = 0 for each (𝑥, 𝑦) ∈ Domn(𝑓).

Notice that
𝜕𝑔
0=
𝜕𝑥
𝜕𝐹 𝜕𝑥 𝜕𝐹 𝜕𝑦 𝜕𝐹 𝜕𝑧
= ⋅ + ⋅ + ⋅
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑧 𝜕𝑥
𝜕𝐹 𝜕𝐹 𝜕𝑧
= + ⋅
𝜕𝑥 𝜕𝑧 𝜕𝑥
𝜕𝑧
= (2𝑥𝑦 + 𝑦 2 − 𝑦𝑧) + (2𝑧 − 𝑥𝑦) ⋅
𝜕𝑥
and
𝜕𝑔
0=
𝜕𝑦
𝜕𝐹 𝜕𝑥 𝜕𝐹 𝜕𝑦 𝜕𝐹 𝜕𝑧 𝜕𝐹 𝜕𝐹 𝜕𝑧
= ⋅ + ⋅ + ⋅ = + ⋅
𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑦
𝜕𝑧
= (𝑥 2 + 2𝑦𝑥 − 𝑥𝑧) + (2𝑧 − 𝑥𝑦) ⋅ .
𝜕𝑦

Therefore for each (𝑥, 𝑦) ∈ Domn(𝑓) such that 2𝑧 − 𝑥𝑦 ≠ 0, we get

𝜕𝑧 (2𝑥𝑦 + 𝑦 2 − 𝑦𝑧)
=− and
𝜕𝑥 (2𝑧 − 𝑥𝑦)

𝜕𝑧 (𝑥 2 + 2𝑦𝑥 − 𝑥𝑧)
=− .
𝜕𝑦 (2𝑧 − 𝑥𝑦)

68
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

Solutions of Activities

Activity 5.1

𝑑𝑦
Notice first that since 𝑦 = 𝑓(𝑥) is differentiable on 𝐼, exists on 𝐼. Also, since 𝐹 is
𝑑𝑥
𝜕𝐹 𝜕𝐹
differentiable on an open set in ℝ2 containing the set {(𝑥, 𝑓(𝑥)): 𝑥 ∈ 𝐼}, and exists
𝜕𝑥 𝜕𝑦

on that open set.

𝑑𝑔
For each 𝑥 ∈ 𝐼, define 𝑔: 𝐼 → ℝ by 𝑔(𝑥) = 𝐹(𝑥, 𝑓(𝑥)). Then 𝑔(𝑥) = 0 on 𝐼. Hence =0
𝑑𝑥

on 𝐼.

Now by chain rule,

𝑑𝑔 𝜕𝐹 𝑑𝑥 𝜕𝐹 𝑑𝑦 𝜕𝐹 𝜕𝐹 𝑑𝑦
0= = ⋅ + ⋅ = + ⋅ .
𝑑𝑥 𝜕𝑥 𝑑𝑥 𝜕𝑦 𝑑𝑥 𝜕𝑥 𝜕𝑦 𝑑𝑥

Therefore,
𝜕𝐹
𝑑𝑦
= − 𝜕𝑥
𝑑𝑥 𝜕𝐹
𝜕𝑦
𝜕𝐹
for each 𝑥 ∈ 𝐼, where ≠ 0.
𝜕𝑦

Summary

 If the function 𝑥𝑛 = 𝑓(𝑥1 , … , 𝑥𝑛−1 ) has been defined implicitly by the

equation 𝐹(𝑥1 , … , 𝑥𝑛 ) = 0, then it is possible to find partial derivatives of
𝑥𝑛 with respect to the remaining variables 𝑥1 , … , 𝑥𝑛−1, provided that both
𝐹 and 𝑓 are differentiable and satisfy certain additional conditions. The
process of finding partial derivatives, however, is based on chain rule.

69
Copyright © 2018, The Open University of Sri Lanka
MHZ4553: Unit II Session 5: Implicit Functions

Learning Outcomes

After studying session 5, you should be able to,

 Find partial derivatives of functions defined implicitly.

70
Copyright © 2018, The Open University of Sri Lanka