Mathematical Methods

(MTH 201)
Dr. Julius Ehigie
 Partial Derivatives
Table: Heat Index as a function of temperature 𝑇°𝐶 and humidity 𝐻%

(𝑇, 𝐻) 50 55 60 65 70 75 80 85 90

90 96 98 100 103 106 109 112 115 119

92 100 103 105 108 112 115 119 123 128

94 104 107 111 114 118 122 127 132 137

96 109 113 116 121 125 130 135 141 146

98 114 118 123 127 133 138 144 150 157

100 119 124 129 135 141 147 154 161 168

• At what rate is temperature changing at 𝑇 = 92 when 𝐻 = 60?

𝑓 92 + ∆𝑇, 60 − 𝑓 92,60
𝑓𝑇 92 = lim
∆𝑇→0 ∆𝑇
 Partial Differentiation
Let 𝑍 = 𝑓 𝑥, 𝑦 be function of two independent variables 𝑥 and 𝑦.
Differentiation can be done with respect to an independent variable
at a time.

𝜕𝑍 𝑓 𝑥 + ℎ, 𝑦 − 𝑓(𝑥, 𝑦) 𝜕𝑍 𝑓 𝑥, 𝑦 + ℎ − 𝑓(𝑥, 𝑦)
= lim = lim
𝜕𝑥 ℎ→0 ℎ 𝜕𝑦 ℎ→0 ℎ

• To find 𝑓𝑥 , regard 𝑦 as a constant and differentiate 𝑓(𝑥, 𝑦) w.r.t. 𝑥

• To find 𝑓𝑦 , regard 𝑥 as a constant and differentiate 𝑓(𝑥, 𝑦) w.r.t. 𝑦
𝜕𝑍
= 𝑍𝑥 Differentiation of 𝑍 partially w.r.t 𝑥
𝜕𝑥
𝜕𝑍
= 𝑍𝑦 Differentiation of 𝑍 partially w.r.t 𝑦
𝜕𝑦
 Second Derivatives

𝜕2𝑍 𝜕 𝜕𝑍 Differentiation of 𝑍 partially

= = 𝑍𝑥𝑥
𝜕𝑥 2 𝜕𝑥 𝜕𝑥 w.r.t 𝑥 twice
𝜕2𝑍 𝜕 𝜕𝑍 Differentiation of 𝑍 partially
2 = = 𝑍𝑦𝑦 w.r.t 𝑦 twice
𝜕𝑦 𝜕𝑦 𝜕𝑦
𝜕2𝑍 𝜕 𝜕𝑍 Differentiation of 𝑍 partially
= = 𝑍𝑦𝑥 w.r.t 𝑦 first, then w.r.t 𝑥
𝜕𝑥𝜕𝑦 𝜕𝑥 𝜕𝑦
𝜕2𝑍 𝜕 𝜕𝑍 Differentiation of 𝑍 partially
= = 𝑍𝑥𝑦 w.r.t 𝑥 first, then w.r.t 𝑦
𝜕𝑦𝜕𝑥 𝜕𝑦 𝜕𝑥

• Higher-order derivatives can be evaluated

 𝜕𝑢 𝜕𝑢 𝜕2 𝑢 𝜕2 𝑢
Ex: Find , , and , if 𝑢 = 𝑥 3 𝑦 − sin 𝑦
𝜕𝑥 𝜕𝑦 𝜕𝑥𝜕𝑦 𝜕𝑦 2

Solution:
𝜕𝑢 𝜕 𝜕
- =𝑦∙ 𝑥3 − sin 𝑦 = 𝑦 ∙ 3𝑥 2 − 0 = 3𝑥 2 𝑦
𝜕𝑥 𝜕𝑥 𝜕𝑥

𝜕𝑢 𝜕 𝜕
- = 𝑥3 ∙ 𝑦 − sin 𝑦 = 𝑥 3 ∙ 1 − cos 𝑦 = 𝑥 3 − cos 𝑦
𝜕𝑦 𝜕𝑦 𝜕𝑦

𝜕2 𝑢 𝜕 𝜕𝑢 𝜕 𝜕 𝜕
- = = 𝑥 3 − cos 𝑦 = 𝑥3 − cos 𝑦
𝜕𝑥𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑥

= 3𝑥 2 − 0 = 3𝑥 2
𝜕2 𝑢 𝜕 𝜕𝑢 𝜕 𝜕 𝜕
- = = 𝑥3 − cos 𝑦 = 𝑥3 − cos 𝑦
𝜕𝑦 2 𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑦 𝜕𝑦

= 0 − (−sin 𝑦) = sin 𝑦
 𝜕𝑢 𝜕𝑢
Ex: Find and if 𝑢 = 3𝑥 2 sin 𝑦
𝜕𝑥 𝜕𝑦

Solution:
- If 𝑢 = 3𝑥 2 sin 𝑦
𝜕𝑢 𝜕
Then, = [sin 𝑦] ∙ 3𝑥 2
𝜕𝑥 𝜕𝑥

=sin 𝑦 ∙ 6𝑥 = 6𝑥 sin 𝑦
- If 𝑢 = 3𝑥 2 sin 𝑦
𝜕𝑢 𝜕
Then, = [3𝑥 2 ] ∙ sin 𝑦
𝜕𝑦 𝜕𝑥

=3𝑥 2 ∙ cos 𝑦 = 3𝑥 2 cos 𝑦

 Application: Cobb-Douglas Production Function

The Cobb-Douglas production function is denoted by 𝑃 = 𝑃(𝐿, 𝐾)

given by
𝑃 𝐿, 𝐾 = 𝑏𝐿𝛼 𝐾 1−𝛼

where 𝐿 is amount of Labour and 𝐾 is capital investment.

𝜕𝑃
• We can obtain ,the rate at which production changes
𝜕𝐿
w.r.t. amount of labour, called marginal productivity of labor
𝜕𝑃
• Similarly, we can obtain , the rate of change of production
𝜕𝐾
w.r.t. capital, called marginal productivity of capital
Ex: Show that the Cobb-Douglas production function 𝑃 = 𝑏𝐿𝛼 𝐾𝛽
𝜕𝑃 𝜕𝑃
satisfies the equation 𝐿 + 𝐾 = 𝛼 + 𝛽 𝑃
𝜕𝐿 𝜕𝐾

Solution:
𝜕𝑃 𝜕𝑃
𝛼−1 𝛽
= 𝛼𝑏𝐿 𝐾 = 𝛽𝑏𝐿𝛼 𝐾𝛽−1
𝜕𝐿 𝜕𝐾
𝜕𝑃 𝜕𝑃
𝐿 +𝐾 = 𝐿 ∙ 𝛼𝑏𝐿𝛼−1 𝐾𝛽 + 𝐾 ∙ 𝛽𝑏𝐿𝛼 𝐾𝛽−1
𝜕𝐿 𝜕𝐾
= 𝛼𝑏𝐿𝛼 𝐾𝛽 + 𝛽𝑏𝐿𝛼 𝐾𝛽
= (𝛼 + 𝛽)𝑏𝐿𝛼 𝐾𝛽
𝜕𝑃 𝜕𝑃
֜𝐿 +𝐾 = 𝛼+𝛽 𝑃
𝜕𝐿 𝜕𝐾
 Total Differentiation

• In Partial differentiation only one of the independent variable varies

• In Total differentiation, all the independent variables varies

• Increments are given in all the variables

𝜕𝑍 𝑓 𝑥+ℎ,𝑦 −𝑓(𝑥,𝑦)
• Partial derivative formula from first principle = lim
𝜕𝑥 ℎ→0 ℎ

• Given a function 𝑍 = 𝑓(𝑥, 𝑦), let 𝛿𝑥 and 𝛿𝑦 be increments such that

𝑍 + 𝛿𝑧 = 𝑓 𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦

Then 𝑍 + 𝛿𝑧 − 𝑍 = 𝛿𝑧 = 𝑓 𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦 − 𝑓(𝑥, 𝑦)

 Derivation

𝛿𝑧 = 𝑓 𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦 − 𝑓(𝑥, 𝑦)
• Adding and subtracting 𝑓(𝑥, 𝑦 + 𝛿𝑦) to RHS
𝛿𝑧 = 𝑓 𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦 − 𝑓 𝑥, 𝑦 + 𝛿𝑦 + 𝑓(𝑥, 𝑦 + 𝛿𝑦) − 𝑓(𝑥, 𝑦)

𝑓 𝑥 + 𝛿𝑥, 𝑦 + 𝛿𝑦 − 𝑓 𝑥, 𝑦 + 𝛿𝑦 𝑓(𝑥, 𝑦 + 𝛿𝑦) − 𝑓(𝑥, 𝑦)

𝛿𝑧 = 𝛿𝑥 + 𝛿𝑦
𝛿𝑥 𝛿𝑦
• Taking limits of both sides when 𝛿𝑥 → 0, 𝛿𝑦 → 0 and 𝛿𝑧 → 0
𝜕𝑓 𝜕𝑓
𝑑𝑧 = 𝑑𝑥 + 𝑑𝑦
𝜕𝑥 𝜕𝑦
• 𝑑𝑧 is called the total differential of 𝑧
 Theorems and applicable formulas

Chain Rule Case 1.

If 𝑧 = 𝑓 𝑥, 𝑦 , 𝑥 = 𝜑 𝑡 , 𝑦 = 𝜓(𝑡), Here 𝑧 is a composition function, then
𝑑𝒛 𝜕𝒛 𝑑𝑥 𝜕𝒛 𝑑𝑦
= +
𝑑𝒕 𝜕𝑥 𝑑𝒕 𝜕𝑦 𝑑𝒕
𝑧
𝜕𝑧 𝜕𝑧

Chain Rule Case 2 𝜕𝑥 𝜕𝑦

If 𝑧 = 𝑓 𝑥, 𝑦 , 𝑥 = 𝜑 𝑢, 𝑣 , 𝑦 = 𝜓(𝑢, 𝑣), then 𝑥 𝑦

𝜕𝒛 𝜕𝒛 𝜕𝑥 𝜕𝒛 𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
= + 𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
𝜕𝒖 𝜕𝑥 𝜕𝒖 𝜕𝑦 𝜕𝒖 𝑣
𝑢 𝑣 𝑢

𝜕𝒛 𝜕𝒛 𝜕𝑥 𝜕𝒛 𝜕𝑦
= +
𝜕𝒗 𝜕𝑥 𝜕𝒗 𝜕𝑦 𝜕𝒗
 3 3 𝑑𝑢
Ex: If 𝑢 = 𝑥 + 𝑦 where 𝑥 = 𝑎 cos 𝑡, 𝑦 = 𝑏 sin 𝑡. Find .
𝑑𝑡

Solution:
𝑢 = 𝑥3 + 𝑦3 𝑥 = 𝑎 cos 𝑡 𝑦 = 𝑏 sin 𝑡

𝜕𝑢 𝑑𝑥 𝑑𝑦
= 3𝑥 2 = −𝑎 sin 𝑡 = 𝑏 cos 𝑡
𝜕𝑥 𝑑𝑡 𝑑𝑡
𝜕𝑢
= 3𝑦 2
𝜕𝑦
𝑑𝑢 𝜕𝑢 𝑑𝑥 𝜕𝑢 𝑑𝑦
= + = 3𝑥 2 ∙ (−𝑎 sin 𝑡) + 3𝑦 2 ∙ (𝑏 cos 𝑡)
𝑑𝑡 𝜕𝑥 𝑑𝑡 𝜕𝑦 𝑑𝑡
𝑑𝑢
֜ = 3(𝑎 cos 𝑡)2 ∙ (−𝑎 sin 𝑡) + 3(𝑏 sin 𝑡)2 ∙ (𝑏 cos 𝑡)
𝑑𝑡
𝑑𝑢
֜ = 3(𝑎 cos 𝑡)2 ∙ (−𝑎 sin 𝑡) + 3(𝑏 sin 𝑡)2 ∙ (𝑏 cos 𝑡)
𝑑𝑡
 Ex: Compute the error in evaluation of z = 𝑓 𝑥, 𝑦 = 𝑥 2 + 3𝑥𝑦 − 𝑦 2 .
If 𝑥 changes from 2 to 2.05 and 𝑦 changes from 3 to 2.96, compare
∆𝑧 and 𝑑𝑧.

Solution:
𝜕𝑧 𝜕𝑧
𝑑𝑧 = 𝑑𝑥 + 𝑑𝑦 𝑑𝑥 = 2.05 − 2 = 0.05 𝑑𝑦 = 2.96 − 3 = −0.04
𝜕𝑥 𝜕𝑦
𝜕𝑧 𝜕𝑧
= 2𝑥 + 3𝑦 = 3𝑥 − 2𝑦
𝜕𝑥 𝜕𝑦
𝑑𝑧 = 2𝑥 + 3𝑦 𝑑𝑥 + 3𝑥 − 2𝑦 𝑑𝑦
֜ 𝑑𝑧 = 2 2 + 3 3 0.05 + 3 2 − 2 3 −0.04 = 0.65
• This can be verified by using the correct formula
𝑓 2.05,2.96 − 𝑓 2,3 = 0.6449
 2 4 𝑑𝑧
If 𝑧 = 𝑥 𝑦 + 3𝑥𝑦 , where 𝑥 = sin 2𝑡 and 𝑦 = cos 𝑡. Find when 𝑡 = 0.
𝑑𝑡

Solution:
• Apply the Chain Rule 1
𝑑𝑧 𝜕𝑧 𝑑𝑥 𝜕𝑧 𝑑𝑦
= +
𝑑𝑡 𝜕𝑥 𝑑𝑡 𝜕𝑦 𝑑𝑡
𝑑𝑧
= (2𝑥𝑦 + 3𝑦 4 )(2 cos 2𝑡) + (𝑥 2 + 12𝑥𝑦 3 )(− sin 𝑡)
𝑑𝑡
𝑥(0) = sin 2 0 = 0 and 𝑦 0 = cos 0 = 1
𝑑𝑧
|𝑡=0 = (2(0)(0) + 3 1 4 )(2 cos 2(0)) + ( 0 2 + 12(0)(0)3 )(− sin(0))
𝑑𝑡
𝑑𝑧
֜ |𝑡=0 = 0 + 3 2 + 0 + 0 0 = 6
𝑑𝑡
 Tangents

• Suppose 𝑓 has continuous partial derivatives. An equation of the

tangent plane to the surface 𝑧 = 𝑓 𝑥, 𝑦 at the point 𝑃 𝑥0 , 𝑦0 , 𝑧0 is
given by
𝑧 − 𝑧0 = 𝑓𝑥 𝑥0 , 𝑦0 𝑥 − 𝑥0 + 𝑓𝑦 𝑥0 , 𝑦0 𝑦 − 𝑦0

• Suppose 𝑓 has continuous partial derivatives. An equation of the

tangent plane to the surface 𝐹 𝑥, 𝑦, 𝑧 at the point 𝑃 𝑥0 , 𝑦0 , 𝑧0 is
𝐹𝑥 𝑥0 , 𝑦0 , 𝑧0 𝑥 − 𝑥0 + 𝐹𝑦 𝑥0 , 𝑦0 , 𝑧0 𝑦 − 𝑦0 + 𝐹𝑧 𝑥0 , 𝑦0 , 𝑧0 𝑧 − 𝑧0 = 0
 Linear Approximation

Given the function 𝑧 = 𝑓(𝑥, 𝑦), an equation of the tangent plane to the
graph of function 𝑓 of two variables at the point 𝑎, 𝑏, 𝑓 𝑎, 𝑏 is given by

𝑧 − 𝑧0 = 𝑓𝑥 𝑥0 , 𝑦0 𝑥 − 𝑥0 + 𝑓𝑦 𝑥0 , 𝑦0 𝑦 − 𝑦0

Therefore, the linear function whose graph is the tangent plane, namely

𝐿 𝑥, 𝑦 = 𝑓 𝑎, 𝑏 + 𝑓𝑥 𝑎, 𝑏 𝑥 − 𝑎 + 𝑓𝑦 𝑎, 𝑏 𝑦 − 𝑏

is called the linearization of 𝑓 at (𝑎, 𝑏) and the approximation

𝑓(𝑥, 𝑦) ≈ 𝑓 𝑎, 𝑏 + 𝑓𝑥 𝑥0 , 𝑦0 𝑥 − 𝑥0 + 𝑓𝑦 𝑥0 , 𝑦0 𝑦 − 𝑦0

is called the linear approximation of 𝑓 at (𝑎, 𝑏)

Ex: Find the tangent plane to the elliptic paraboloid 𝑧 = 2𝑥 2 + 𝑦 2
at the point (1,1,3)

Equation of the tangent passing through a point is

𝑧 − 𝑧0 = 𝑓𝑥 𝑥0 , 𝑦0 𝑥 − 𝑥0 + 𝑓𝑦 𝑥0 , 𝑦0 𝑦 − 𝑦0
𝑓𝑥 𝑥, 𝑦 = 4𝑥 ֜ 𝑓𝑥 𝑥0 , 𝑦0 = 4 1 = 4
𝑓𝑦 𝑥, 𝑦 = 2𝑦 ֜ 𝑓𝑦 𝑥0 , 𝑦0 = 2 1 = 2

Therefore, the equation of the tangent at (1,1,3) is

plot3d({4*x+2*y-3, 2*x^2+y^2}, x = -4 .. 4, y = -4 .. 4)
𝑧−3=4 𝑥−1 +2 𝑦−1
𝑧 − 3 = 4𝑥 − 4 + 2𝑦 − 2
֜ 𝑧 = 4𝑥 + 2𝑦 − 3
is the equation of the tangent at (1,1,3)
Ex: Linearize the function 𝑓(𝑥, 𝑦) = 𝑥𝑒 𝑥𝑦 at the point 1,0 .
Estimate 𝑓(1.1, −0.1).

The linearization of 𝑓 at (1,0)

𝐿 𝑥, 𝑦 = 𝑓 𝑎, 𝑏 + 𝑓𝑥 𝑥0 , 𝑦0 𝑥 − 𝑥0 + 𝑓𝑦 𝑥0 , 𝑦0 𝑦 − 𝑦0
𝑓𝑥 𝑥, 𝑦 = 𝑥𝑦𝑒 𝑥𝑦 + 𝑒 𝑥𝑦 ֜ 𝑓𝑥 1,0 = (1)(0)𝑒 (1)(0) + 𝑒 (1)(0) = 1
𝑓𝑦 𝑥, 𝑦 = 𝑥 2 𝑒 𝑥𝑦 ֜ 𝑓𝑦 1,0 = (12 )𝑒 (1)(0) = 1
𝑓 1,0 = 1 𝑒 1 0 =1
𝐿 𝑥, 𝑦 = 1 + 1 𝑥 − 1 + 1 𝑦 − 0
֜ 𝐿 𝑥, 𝑦 = 1 + 𝑥 − 1 + 𝑦 = 𝑥 + 𝑦
Approximation at 𝑓 1.1, −0.1 ֜ 𝑓 𝑥, 𝑦 ≈
1.1 + (−0.1) ֜ 𝑓 1.1, −0.1 ≈ 𝟏
• Verification: 𝑓 𝑥, 𝑦 = 𝑥𝑒 𝑥𝑦 ֜ 𝑓 1.1, −0.1 = 𝟎. 𝟗𝟖𝟓𝟒𝟐
A rectangular box has sides measured as 30𝑚𝑚, 40𝑚𝑚 and 60𝑚𝑚. If
these measurements are liable to be in error of ±0.5𝑚𝑚, ±0.8𝑚𝑚 and

±1.0𝑚𝑚,respectively, calculate the length of a diagonal of the box and

the maximum possible error in the result.
Solution
• The diagonal of the rectangular box 𝐷 = 𝑎2 + 𝑏 2 + 𝑐 2
𝜕𝐷 𝜕𝐷 𝜕𝐷 𝑑𝑎 = ±0.55𝑚𝑚
• The differentials 𝑑𝐷 = 𝑑𝑎 + 𝑑𝑏 + 𝑑𝑐
𝜕𝑎 𝜕𝑏 𝜕𝑐
𝜕𝐷 𝑎 𝜕𝐷 𝑏 𝜕𝐷 𝑐 𝑑𝑏 = ±0.88𝑚𝑚
= , 𝜕𝑏 = ,
2 +𝑏 2 +𝑐 2 𝜕𝑐
= 2 +𝑏 2 +𝑐 2
, 𝑑𝑐 = ±1.0𝑚𝑚,
𝜕𝑎 2 2
𝑎 +𝑏 +𝑐 2 𝑎 𝑎
1
֜ 𝑑𝐷 = (𝑎𝑑𝑎 + 𝑏𝑑𝑏 + 𝑐𝑑𝑐)
𝑎2 + 𝑏2 + 𝑐2
1
֜ 𝑑𝐷 = (30 0.5 + 40 0.8 + 60(1))
302 + 402 + 602
• The maximum error in calculated volume is ±1.37𝑐𝑚3
The base radius and height of a right circular cone are measured
as 10cm and 25cm, respectively, with a possible error of as much as
0.1cm in each dimension. Use differentials to estimate the maximum

error in the volume of the cone.

Solution

𝜋𝑟 2 ℎ 𝑑𝑟 = 0.1 𝑑ℎ = 0.1
• The volume of the cone is given by 𝑉 =
3
𝜕𝑉 𝜕𝑉
• The differentials 𝑑𝑉 = 𝑑𝑟 + 𝑑ℎ
𝜕𝑟 𝜕ℎ
𝜕𝑉 2𝜋𝑟ℎ 𝑑𝑉 𝜋𝑟 2
= , = ,
𝜕𝑟 3 𝑑ℎ 3
2𝜋𝑟ℎ 𝜋𝑟 2 2𝜋 10 25 𝜋 10 2
֜ 𝑑𝑉 = 𝑑𝑟 + 𝑑ℎ = (0.1) + (0.1),
3 3 3 3
50𝜋 10𝜋
֜ 𝑑𝑉 = + = 20𝜋,
3 3

• The maximum error in calculated volume is 20𝜋 ≈ 63𝑐𝑚3