CONCEPTS COVERED

MULTIVARIABLE CALCULUS

❑ Taylor’s Theorem

❑ Worked Problem

1
Taylor’s Theorem for a Function of Single Variables (Recall)

Assume that the function 𝑓 has all derivatives up to the order 𝑛 + 1 in some
interval containing the point 𝑥 = 𝑥0 .

ℎ2 ′′ ℎ𝑛
𝑓 𝑥0 + ℎ = 𝑓 𝑥0 + ℎ𝑓 ′ 𝑥0 + 𝑓 𝑥0 + ⋯ + 𝑓 𝑛 𝑥0 + 𝑅𝑛
2! 𝑛!

ℎ𝑛+1 𝑛+1
𝑅𝑛 = 𝑓 𝜉 , 𝑥0 < 𝜉 < 𝑥0 + ℎ
𝑛+1 !

2
Taylor’s Theorem for a Function of Two Variables
Let a function be defined in some domain 𝐷 in ℝ2 and have continuous partial
derivatives up to 𝑛 + 1 th order in some neighborhood of a point 𝑃 𝑥0 , 𝑦0 in 𝐷. Then
2
𝜕 𝜕 1 𝜕 𝜕
𝑓 𝑥0 + ℎ, 𝑦0 + 𝑘 = 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 +
𝜕𝑥 𝜕𝑦 2! 𝜕𝑥 𝜕𝑦
𝑛
1 𝜕 𝜕
…+ ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + 𝑅𝑛
𝑛! 𝜕𝑥 𝜕𝑦
where the remainder is given by
𝑛+1
1 𝜕 𝜕
𝑅𝑛 = ℎ +𝑘 𝑓 𝑥0 + 𝜃ℎ, 𝑦0 + 𝜃𝑘 , 0<𝜃<1
(𝑛 + 1)! 𝜕𝑥 𝜕𝑦

3
Taylor’s Theorem for a Function of Two Variables
Proof: For Simplicity, we take 𝑛 = 2 (terms up to order 3)
Let 𝑥 = 𝑥0 + 𝑡ℎ, 𝑦 = 𝑦0 + 𝑡𝑘, where the parameter 𝑡 ∈ 0, 1 .

Define 𝜙 𝑡 = 𝑓(𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘)

𝜕𝑓 𝑑𝑥 𝜕𝑓 𝑑𝑦 𝜕 𝜕
𝜙′ 𝑡 = + = ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘
𝜕𝑥 𝑑𝑡 𝜕𝑦 𝑑𝑡 𝜕𝑥 𝜕𝑦

𝜕2𝑓 𝜕2𝑓 𝜕2𝑓 𝜕2𝑓

𝜙 ′′ 𝑡 =ℎ 2
ℎ+ 𝑘 +𝑘 ℎ+ 2𝑘
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦
2 2 2 2
2
𝜕 𝑓 𝜕 𝑓 2
𝜕 𝑓 𝜕 𝜕
=ℎ 2
+ 2ℎ𝑘 +𝑘 2
= ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘
𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦

4
 𝜕3𝑓 𝜕3𝑓 𝜕3𝑓 𝜕3𝑓 𝜕 3𝑓 𝜕 3𝑓
𝜙 ′′ ′ 𝑡 = ℎ2 3
ℎ + 2
𝑘 + 2ℎ𝑘 2
ℎ + 2
𝑘 + 𝑘 2
2
ℎ+ 3𝑘
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦 𝜕𝑦

3
𝜕3𝑓 𝜕3𝑓 𝜕3𝑓 𝜕3𝑓 𝜕 𝜕
= ℎ3 + 3ℎ2 𝑘 + 3ℎ𝑘 2 + 𝑘3 = ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘
𝜕𝑥 3 𝜕𝑥 2 𝜕𝑦 𝜕𝑥 𝜕𝑦 2 𝜕𝑦 3 𝜕𝑥 𝜕𝑦

Using Taylor’s Theorem for 𝜙(𝑡) about the point 0 as

𝑡 2 𝑡 3
𝜙 𝑡 = 𝜙 0 + 𝑡 𝜙 ′ (0) + 𝜙 ′′ (0) + 𝜙 ′′′ 𝜃𝑡 , 0<𝜃<1
2! 3!

1 ′′ 1 ′′′
′
𝜙 1 = 𝜙 0 + 𝜙 (0) + 𝜙 (0) + 𝜙 𝜃 , 0<𝜃<1
2! 3!

5
 𝜕 𝜕
𝜙 𝑡 = 𝑓(𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘) ′
𝜙 𝑡 = ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘
𝜕𝑥 𝜕𝑦
2 3
𝜕 𝜕 𝜕 𝜕
𝜙 ′′ 𝑡 = ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘 𝜙 ′′′ 𝑡 = ℎ +𝑘 𝑓 𝑥0 + 𝑡ℎ, 𝑦0 + 𝑡𝑘
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦

1 ′′ 1 ′′′
𝜙 1 =𝜙 0 + 𝜙 ′ (0) + 𝜙 (0) + 𝜙 𝜃 , 0<𝜃<1
2! 3!

𝜕 𝜕
𝑓 𝑥0 + ℎ, 𝑦0 + 𝑘 = 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 +
𝜕𝑥 𝜕𝑦
2 3
1 𝜕 𝜕 1 𝜕 𝜕
ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 + 𝜃ℎ, 𝑦0 + 𝜃𝑘
2! 𝜕𝑥 𝜕𝑦 3! 𝜕𝑥 𝜕𝑦

6
General Case:
2
𝜕 𝜕 1 𝜕 𝜕
𝑓 𝑥0 + ℎ, 𝑦0 + 𝑘 = 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 +
𝜕𝑥 𝜕𝑦 2! 𝜕𝑥 𝜕𝑦
𝑛 𝑛+1
1 𝜕 𝜕 1 𝜕 𝜕
…+ ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 + 𝜃ℎ, 𝑦0 + 𝜃𝑘
𝑛! 𝜕𝑥 𝜕𝑦 (𝑛 + 1)! 𝜕𝑥 𝜕𝑦

Alternatively,
𝜕 𝜕
𝑓 𝑥, 𝑦 = 𝑓 𝑥0 , 𝑦0 + 𝑥 − 𝑥0 + (𝑦 − 𝑦0 ) 𝑓 𝑥0 , 𝑦0 + ⋯
𝜕𝑥 𝜕𝑦
𝑛+1
1 𝜕 𝜕
+ 𝑥 − 𝑥0 + 𝑥 − 𝑥0 𝑓 𝑥0 + 𝜃 𝑥 − 𝑥0 , 𝑦0 + 𝜃 𝑦 − 𝑦0
(𝑛 + 1)! 𝜕𝑥 𝜕𝑦

7
Problem - 1 Find the quadratic polynomial approximation to the function
𝑥−𝑦
𝑓 𝑥, 𝑦 = about the point (1, 1)
𝑥+𝑦

𝑥 + 𝑦 − (𝑥 − 𝑦) 2𝑦 1
𝑓𝑥 (𝑥, 𝑦) = 2
= 2 ⟹ 𝑓𝑥 1, 1 =
𝑥+𝑦 𝑥+𝑦 2
− 𝑥 + 𝑦 − (𝑥 − 𝑦) −2𝑥 1
𝑓𝑦 (𝑥, 𝑦) = 2
= 2
⟹ 𝑓𝑦 1, 1 = −
𝑥+𝑦 𝑥+𝑦 2
−4𝑦 1 4𝑥
𝑓𝑥𝑥 (𝑥, 𝑦) = ⟹ 𝑓𝑥𝑥 1, 1 = − 𝑓𝑦𝑦 (𝑥, 𝑦) =
𝑥+𝑦 3 2 𝑥+𝑦 3

1 2𝑥 − 2𝑦
⟹ 𝑓𝑦𝑦 1, 1 = 𝑓𝑥𝑦 (𝑥, 𝑦) = 3
⟹ 𝑓𝑥𝑦 1, 1 = 0
2 𝑥+𝑦

8
 1 1 1 1
𝑓𝑥 1, 1 = 𝑓𝑦 1, 1 = − 𝑓𝑥𝑥 1, 1 = − 𝑓𝑦𝑦 1, 1 = 𝑓𝑥𝑦 1, 1 = 0
2 2 2 2

1
𝑃2 𝑥, 𝑦 = 𝑓 1, 1 + 𝑓𝑥 1, 1 𝑥 − 1 + 𝑓𝑦 1, 1 𝑦 − 1 + 𝑓𝑥𝑥 1, 1 𝑥 − 1 2
2
1 2
+𝑓𝑥𝑦 1, 1 (𝑥 − 1)(𝑦 − 1) + 𝑓𝑦𝑦 1,1 𝑦 − 1
2

1 1 1 2
1 2
𝑃2 𝑥, 𝑦 = 𝑥 − 1 − 𝑦 − 1 − 𝑥 − 1 + 𝑦−1
2 2 4 4

9
Problem - 2 Let 𝑓 𝑥, 𝑦 = 𝑥 2 + 𝑥𝑦 + 𝑦 2 be linearly approximated by the Taylor’s polynomial
about the point 1, 1 . Find out the maximum error in this approximation at a
point in the square 𝑥 − 1 ≤ 0.1, 𝑦 − 1 ≤ 0.1.

𝑓𝑥 𝑥, 𝑦 = 2𝑥 + 𝑦 𝑓𝑥𝑥 𝑥, 𝑦 = 2 𝑓𝑥𝑦 𝑥, 𝑦 = 1

𝑓𝑦 𝑥, 𝑦 = 𝑥 + 2𝑦 𝑓𝑦𝑦 𝑥, 𝑦 = 2
2
1 𝜕 𝜕
Remainder: 𝑅1 = 𝑥−1 + 𝑦−1 𝑓 1 + 𝜃 𝑥 − 1 ,1 + 𝜃 𝑦 − 1
2 𝜕𝑥 𝜕𝑦
1
𝑅1 = 𝑥 − 1 2 𝑓𝑥𝑥 + 2 𝑥 − 1 𝑦 − 1 𝑓𝑥𝑦 + 𝑦 − 1 2 𝑓𝑦𝑦
2
𝑅1 = 𝑥 − 1 2 + 𝑥−1 𝑦−1 + 𝑦−1 2

Maximum Error: 𝑅1 ≤ 0.1 2 + 0.1 2 + 0.1 2

= 0.03

10
Problem - 3 Obtain Taylor’s formula about the point 0, 0 involving derivatives up to 3rd
order for the function 𝑓 𝑥, 𝑦 = cos(𝑥 + 𝑦).
Taylor’s theorem:
2 3
𝜕 𝜕 1 𝜕 𝜕 1 𝜕 𝜕
𝑓 𝑥, 𝑦 = 𝑓 0,0 + 𝑥 +𝑦 𝑓 0, 0 + 𝑥 +𝑦 𝑓 0, 0 + 𝑥 +𝑦 𝑓 𝜃𝑥, 𝜃𝑦
𝜕𝑥 𝜕𝑦 2! 𝜕𝑥 𝜕𝑦 3! 𝜕𝑥 𝜕𝑦

• 𝑓 0, 0 = 1 0<𝜃<1

• First order derivatives: 𝑓𝑥 = − sin 𝑥 + 𝑦 ⟹ 𝑓𝑥 (0, 0) = 0

𝑓𝑦 = − sin 𝑥 + 𝑦 ⟹ 𝑓𝑦 (0, 0) = 0
• Second order derivatives: 𝑓𝑥𝑥 = 𝑓𝑦𝑦 = 𝑓𝑥𝑦 = − cos(𝑥 + 𝑦)

⟹ 𝑓𝑥𝑥 0, 0 = 𝑓𝑦𝑦 0,0 = 𝑓𝑥𝑦 0,0 = −1

11
 2 3
𝜕 𝜕 1 𝜕 𝜕 1 𝜕 𝜕
𝑓 𝑥, 𝑦 = 𝑓 0,0 + 𝑥 +𝑦 𝑓 0, 0 + 𝑥 +𝑦 𝑓 0, 0 + 𝑥 +𝑦 𝑓 𝜃𝑥, 𝜃𝑦
𝜕𝑥 𝜕𝑦 2! 𝜕𝑥 𝜕𝑦 3! 𝜕𝑥 𝜕𝑦
0<𝜃<1
• Third order derivatives: 𝑓𝑥𝑥𝑥 = 𝑓𝑦𝑦𝑦 = 𝑓𝑥𝑥𝑦 = 𝑓𝑥𝑦𝑦 = sin(𝑥 + 𝑦)

𝑓𝑥𝑥𝑥 (𝜃𝑥, 𝜃𝑦) = 𝑓𝑦𝑦𝑦 (𝜃𝑥, 𝜃𝑦) = 𝑓𝑥𝑥𝑦 (𝜃𝑥, 𝜃𝑦) = 𝑓𝑥𝑦𝑦 (𝜃𝑥, 𝜃𝑦) = sin(𝜃𝑥 + 𝜃𝑦)

1 2 2
1 3
𝑓 𝑥, 𝑦 = 1 + 0 − 𝑥 + 2𝑥𝑦 + 𝑦 + 𝑥 + 3𝑥 2 𝑦 + 3𝑥𝑦 2 + 𝑦 3 sin(𝜃𝑥 + 𝜃𝑦)
2! 3!

1 2
1 3 sin(𝜃𝑥
𝑓 𝑥, 𝑦 = 1 − 𝑥+𝑦 + 𝑥+𝑦 + 𝜃𝑦)
2! 3!

12
 CONCLUSIONS

Taylor’s Theorem for a Function of Two Variables

2
𝜕 𝜕 1 𝜕 𝜕
𝑓 𝑥0 + ℎ, 𝑦0 + 𝑘 = 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ⋯
𝜕𝑥 𝜕𝑦 2! 𝜕𝑥 𝜕𝑦

𝑛 𝑛+1
1 𝜕 𝜕 1 𝜕 𝜕
+ ℎ +𝑘 𝑓 𝑥0 , 𝑦0 + ℎ +𝑘 𝑓 𝑥0 + 𝜃ℎ, 𝑦0 + 𝜃𝑘
𝑛! 𝜕𝑥 𝜕𝑦 (𝑛 + 1)! 𝜕𝑥 𝜕𝑦

13