You are on page 1of 22

Unit-4: Interpolation

4.1 Introduction
Given the set of tabular values (𝑥0 , 𝑦0 ), (𝑥1 , 𝑦1 ), (𝑥2 , 𝑦2 ), … , (𝑥𝑛 , 𝑦𝑛 ) satisfying the relation
𝑦 = 𝑓(𝑥) where the explicit nature of 𝑓(𝑥) is not known, it is required to find a simpler
function, say 𝜙(𝑥), such that 𝑓(𝑥) and 𝜙(𝑥) agree at the set of tabulated points. Such a
process is called interpolation. If 𝜙(𝑥) is a polynomial, then the process is called polynomial
interpolation and 𝜙(𝑥) is called the interpolating polynomial. Similarly, different types of
interpolation arise depending on whether 𝜙(𝑥) is a trigonometric function, exponential
functions, etc.
Numerical interpolation approximates functions and we approximate functions for one or
several of the following reasons:

 A large number of important mathematical functions may only be known through tables
of their values.
 Some function may be known to exist but are computationally too complex to
manipulate numerically.
 Some function may be known but the solution of the problem in which they appear may
not have an obvious mathematical expression to work with.
Some of the methods of interpolation that will be considered in this unit include Newton’s
Forward and Backward difference interpolation formulae, Newton’s divided difference
interpolation formula and the Lagrangian interpolation.

1
4.2. Finite Difference
Assume that we have a table of values (𝑥𝑖 , 𝑦𝑖 ), 𝑖 = 0, 1, 2, … , 𝑛 of any function 𝑦 = 𝑓(𝑥), the
valus of 𝑥 being equally spaced, i.e,𝑥𝑖 = 𝑥𝑜 + 𝑖ℎ, 𝑖 = 0, 1, 2, … , 𝑛. Suppose that we are required
to recover the values of 𝑓(𝑥) for some intermediate values of 𝑥, or to obtain the derivative of
𝑓(𝑥) for some 𝑥 in the range 𝑥0 ≤ 𝑥 ≤ 𝑥𝑛 . The methods for the solution to these problems are
based on the concept of the differences of a function which we now proceed to define.
4.2.1. Forward Differences
If 𝒚𝟎 , 𝒚𝟏 , 𝒚𝟐 , … , 𝒚𝒏 denotes a set of values of 𝑦, then 𝑦1 − 𝑦0 , 𝑦2 − 𝑦1 , … , 𝑦𝑛 − 𝑦𝑛−1 are called
the differences of 𝑦. Denoting these differences by ∆𝑦0 , ∆𝑦1 , ∆𝑦2 , … , ∆𝑦𝑛−1 respectively, we
have
∆𝑦0 = 𝑦1 − 𝑦0 , ∆𝑦1 = 𝑦2 − 𝑦1 , … , ∆𝑦𝑛−1 = 𝑦𝑛 − 𝑦𝑛−1
Where ∆ is called the forward difference operator and ∆𝑦0 , ∆𝑦1 , ∆𝑦2 , … are called first ordered
forward differences. The differences of first ordered forward differences are called second
ordered forward differences and are denoted by ∆2 𝑦0 , ∆2 𝑦1 , ∆2 𝑦2 , … . Similarly, one can
defined 3rd , 4th , …, 𝑛𝑡ℎ order forward differences.

∆2 𝑦0 = ∆𝑦1 − ∆𝑦0 = 𝑦2 − 2𝑦1 + 𝑦0


∆3 𝑦0 = ∆2 𝑦1 − ∆2 𝑦0 = 𝑦3 − 3𝑦2 + 3𝑦1 − 𝑦0
∆4 𝑦0 = ∆3 𝑦1 − ∆3 𝑦0 = 𝑦4 − 4𝑦3 + 6𝑦2 − 4𝑦1 + 𝑦0
⋮ = ⋮ = ⋮
∆𝒏 𝒚𝟎 = 𝒚𝒏 − 𝒄𝒏𝟏 𝒚𝒏−𝟏 + 𝒄𝒏𝟐 𝒚𝒏−𝟐 − ⋯ + (−𝟏)𝒏 𝒚𝟎
Table4.1 Forward Difference Table

𝑥 𝑦 = 𝑓(𝑥) ∆ ∆2 ∆3 ∆4 ∆5
𝑥0 𝑦0
∆𝑦0
𝑥1 𝑦1 ∆2 𝑦0
∆𝑦1 ∆3 𝑦0
2
𝑥2 𝑦2 ∆ 𝑦1 ∆4 𝑦0
∆𝑦2 ∆3 𝑦1 ∆5 𝑦0
𝑥3 𝑦3 ∆2 𝑦2 ∆4 𝑦1
∆𝑦3 3
∆ 𝑦2
𝑥4 𝑦4 2
∆ 𝑦3
∆𝑦4
𝑥5 𝑦5

2
Example
Given 𝑓(0) = 3, 𝑓(1) = 12, 𝑓(2) = 81, 𝑓(3) = 200, 𝑓(4) = 100,𝑎𝑛𝑑 𝑓(5) = 8.

a. Find ∆5 𝑓(0) b. ∆2 𝑓(2) c. ∆4 𝑓(1)


Solution

𝑥 𝑦 = 𝑓(𝑥) ∆ ∆2 ∆3 ∆4 ∆5
0 3
9
1 12 60
69 -10
2 81 50 −259
119 -269 755
3 200 −219 496
-100 227
4 100 8
-92
5 8

4.2.2. Backward Differences


The differences 𝑦1 − 𝑦0 , 𝑦2 − 𝑦1 , … , 𝑦𝑛 − 𝑦𝑛−1 are called first backward differences if they are
denoted by ∇𝑦1 , ∇𝑦2 , ∇𝑦3 , … , ∇𝑦𝑛 respectively, so that
∇𝑦1 = 𝑦1 − 𝑦0 , ∇𝑦2 = 𝑦2 − 𝑦1 , … , ∇𝑦𝑛 = 𝑦𝑛 − 𝑦𝑛−1
Where ∇ is called the backward difference operator. In a similar way, one can define backward
differences of higher orders. Thus we obtain:

∇2 𝑦2 = ∇𝑦2 − ∇𝑦1 = 𝑦2 − 2𝑦1 + 𝑦0


∇3 𝑦3 = ∇2 𝑦3 − ∇2 𝑦2 = 𝑦3 − 3𝑦2 + 3𝑦1 − 𝑦0
∇4 𝑦4 = ∇3 𝑦4 − ∇3 𝑦3 = 𝑦4 − 4𝑦3 + 6𝑦2 − 4𝑦1 + 𝑦0
⋮ = ⋮ = ⋮

𝛁 𝒏 𝒚𝒏 = 𝛁 𝒏−𝟏 𝒚𝒏 − 𝛁 𝒏−𝟏 𝒚𝒏−𝟏 = 𝒚𝒏 − 𝒄𝒏𝟏 𝒚𝒏−𝟏 + 𝒄𝒏𝟐 𝒚𝒏−𝟐 − ⋯ + (−𝟏)𝒏 𝒚𝟎

3
Table4.2 Backward Difference Table

𝑥 𝑦 = 𝑓(𝑥) ∇ ∇2 ∇3 ∇4 ∇5
𝑥0 𝑦0
∇𝑦1
𝑥1 𝑦1 ∇2 𝑦2
∇𝑦2 ∇3 𝑦3
𝑥2 𝑦2 ∇2 𝑦3 ∇4 𝑦4
3
∇𝑦3 ∇ 𝑦4 ∇5 𝑦5
𝑥3 𝑦3 ∇2 𝑦4 ∇4 𝑦5
∇𝑦4 3
∇ 𝑦5
𝑥4 𝑦4 ∇2 𝑦5
∇𝑦5
𝑥5 𝑦5

Example
Construct the backward difference table for 𝑦 = log10 𝑥 given below

𝑥 10 20 30 40 50
𝑦 1 1.3010 1.4771 1.6021 1.6990

4.2.3. Central Differences


The central difference operator 𝛿 is defined by the relations:
ℎ ℎ
𝛿𝑓(𝑥) = 𝑓 (𝑥 + 2) − 𝑓(𝑥 − 2) which is equivalent to 𝛿𝑦𝑛−1 = 𝑦𝑛 − 𝑦𝑛−1 .
2

Similarly, higher-order central differences can be defined as

𝛿 2 𝑦1 = 𝛿𝑦3 − 𝛿𝑦1 , 𝛿 3 𝑦3 = 𝛿 2 𝑦2 − 𝛿 2 𝑦1 , 𝛿 4 𝑦2 = 𝛿 3 𝑦5 − 𝛿 3 𝑦3 and so on.


2 2 2 2 2

4
Table4.3 Central Difference Table

𝑥 𝑦 = 𝑓(𝑥) δ δ2 δ3 δ4 δ5
𝑥0 𝑦0

δ𝑦1
2
𝑥1 𝑦1 δ2 𝑦1
δ𝑦3
2 δ3 𝑦3
𝑥2 𝑦2 δ2 𝑦2 2

δ3 𝑦5 δ4 𝑦2
δ𝑦5 2
2
𝑥3 𝑦3 δ2 𝑦3 δ5 𝑦5
δ𝑦7 δ4 𝑦3 2

2 δ3 𝑦7
2
𝑥4 𝑦4 δ2 𝑦4

δ𝑦9
2
𝑥5 𝑦5

Example
1
Construct central difference table for 𝑓(𝑥) = 𝑥 , 𝑥 ∈ [3,3.5] & ℎ = 0.1

𝑥 𝑦 = 𝑓(𝑥) 𝛿 δ2 δ3 δ4 δ5
3 0.3333

δ𝑦1 =-0.0107
2
3.1 0.3225 δ2 𝑦1 =0.0006
δ𝑦3 =-0.0101
2 δ3 𝑦3 =0
2 2
3.2 0.3125 δ 𝑦2 =0.0006
δ4 𝑦2 =0
δ𝑦5 =-0.0095 δ3 𝑦5 =0
2
2
3.3 0.3030 2
δ 𝑦3 =0.0006 δ5 𝑦5 =-0.001
δ𝑦7 =-0.0089 δ4 𝑦3 =-0.0001 2

2 δ3 𝑦7 =-0.0001
2
3.4 0.2941 δ2 𝑦4 =0.0005
δ𝑦9 =-0.0084
2
3.5 0.2857

5
4.2.4. The shift Operator (E)
The operator E is called shift operator or displacement or translation operator. It shows the
operation of increasing the argument value 𝑥 by its interval of differencing ℎ so that:
𝐸𝑓(𝑥) = 𝑓(𝑥 + ℎ) in the case of a continuous variable 𝑥, and
𝐸𝑦𝑥 = 𝑦𝑥+1 in the case of a discrete variable.
Similarly, 𝐸 2 𝑓(𝑥 + ℎ) = 𝑓(𝑥 + 2ℎ)
Powers of the operator (positive or negative) are defined in a similar manner:
𝐸 𝑛 𝑓(𝑥) = 𝑓(𝑥 + 𝑛ℎ), 𝐸 𝑛 𝑦𝑥 = 𝑦𝑥+𝑛ℎ
4.2.5. Averaging Operator (𝝁)
Averaging operator which is denoted by 𝝁 is defined by
ℎ 1 1 1 1
𝑓(𝑥+ )+𝑓(ℎ−ℎ/2) − −
2 𝐸 2 𝑓(𝑥)+𝐸 2 𝑓(𝑥) (𝐸 2 +𝐸 2 )𝑓(𝑥)
𝜇𝑓(𝑥) = = =
2 2 2
𝟏 𝟏

(𝑬𝟐 +𝑬 𝟐 )
Hence 𝜇 = 𝟐

4.2.6. Differential Operator (𝐷)


Differential operator which is denoted by 𝐷 is defined by
ℎ2 𝑓(𝑥) ℎ3 𝑓(𝑥) (ℎ𝐷)2 𝑓(𝑥)
𝐸𝑓(𝑥) = 𝑓(𝑥 + ℎ) = 𝑓(𝑥) + ℎ𝑓 ′ (𝑥) + + + ⋯ += 𝑓(𝑥) + ℎ𝐷𝑓(𝑥) + +⋯
2! 3! 2!

(ℎ𝐷)2 (ℎ𝐷)3
𝐸𝑓(𝑥) = 𝑓(𝑥)[1 + ℎ𝐷 + + +⋯]
2! 3!

(ℎ𝐷)2 (ℎ𝐷)3
𝐸 = 1 + ℎ𝐷 + + +⋯
2! 3!

𝐸 = 𝑒 ℎ𝐷
ln(𝐸)
ln(𝐸) = ℎ𝐷 ⟹ 𝐷 = ℎ

Hence

1. 𝐸 = 1 + ∆= 1 + 𝑒 ℎ𝐷
ln(1+∆) 1 1 1 1
2. 𝐷 = = ℎ [∆ − 2 ∆2 + 3 ∆3 − 4 ∆4 + ⋯ ]

Relationship between the operators


1. ∆𝑓(𝑥) = 𝑓(𝑥 + ℎ) − 𝑓(𝑥)
𝑓(𝑥 + ℎ) − 𝑓(𝑥) = 𝐸𝑓(𝑥) − 𝑓(𝑥) = (𝐸 − 1)𝑓(𝑥)

6
∆𝑓(𝑥) = (𝐸 − 1)𝑓(𝑥) ⟹ ∆= 𝑬 − 𝟏
2. ∇𝑓(𝑥) = 𝑓(𝑥) − 𝑓(𝑥 − ℎ) = 𝑓(𝑥) − 𝐸 −1 𝑓(𝑥)
∇𝑓(𝑥) = (1 − 𝐸 −1 )𝑓(𝑥) ⟹ 𝛁 = 𝟏 − 𝑬−𝟏
𝟏 𝟏 𝟏 𝟏
𝒉 𝒉
3. 𝜹𝒇(𝒙) = 𝒇 (𝒙 + 𝟐) − 𝒇 (𝒙 − 𝟐) = 𝑬𝟐 𝒇(𝒙) − 𝑬−𝟐 𝒇(𝒙) = (𝑬𝟐 − 𝑬−𝟐 ) 𝒇(𝒙)
𝟏 𝟏
𝜹 = 𝑬𝟐 − 𝑬−𝟐
Example
1/2
𝛿2
1. Show that 𝜇 = (1 + )
4

Solution
1/2
𝟏 𝟏 2

1/2 (𝑬𝟐 −𝑬 𝟐 ) 1/2 2 1/2
𝛿2 𝐸+2+𝐸 −1 𝐸 1/2 +𝐸 −1/2
(1 + ) = (1 + ) =( ) = (( ) )
4 4 4 2

1/2 2 1/2
𝛿2 𝐸 1/2 +𝐸 −1/2 𝐸 1/2 +𝐸 −1/2
(1 + ) = (( ) ) = =𝜇
4 2 2

2. Evaluate the following


a. ∆ tan−1 𝑥

b. (3𝑥 − 4)𝐸
c. ∆(𝑒 𝑥 log 2𝑥)
d. ∆(𝑥 2 / cos 2𝑥)
Solution

a. ∆ tan−1 𝑥 = tan−1(𝑥 + ℎ) − tan−1 𝑥


Let tan−1(𝑥 + ℎ) = 𝐴 and tan−1 𝑥 = 𝐵
tan 𝐴−tan 𝐵 𝑥+ℎ−𝑥 ℎ
tan(𝐴 − 𝐵) = 1+tan 𝐴 tan 𝐵 = 1+𝑥(𝑥+ℎ) = 1+𝑥(𝑥+ℎ)


𝐴 − 𝐵 = ∆ tan−1 𝑥 = tan−1 (1+𝑥(𝑥+ℎ) )

∆ 1
b. (3𝑥 − 4) = (∆(3𝑥 − 4)) = 𝐸 −1 (∆(3𝑥 − 4)) = 𝐸 −1 (3(𝑥 + ℎ) − 4 − (3𝑥 − 4))
𝐸 𝐸

(3𝑥 − 4) = 𝐸 −1 (3ℎ) = 3ℎ
𝐸

7
Exercise
Prove that
−𝟏
a. 𝜹 = 𝛁(𝟏 − 𝛁) 𝟐
b. ∆𝛁 = ∆ − 𝛁 = 𝜹𝟐
∆ 𝛁
c. ∆ + 𝛁 = 𝛁 − ∆
𝟏 𝟏
d. 𝜹 (𝑬𝟐 + 𝑬−𝟐 ) = ∆(𝑬−𝟏 + 𝟏)
e. ∆= 𝟏 − 𝒆−𝒉𝑫

8
4.3. Interpolation with Equally Spaced points
4.3.1. Newton’s Forward Interpolation Formula
Let the function 𝑦 = 𝑓(𝑥) take the values 𝑦0 , 𝑦1 , … , 𝑦𝑛 corresponding to the values
𝑥0 , 𝑥1 , … , 𝑥𝑛 of 𝑥. Let these values of 𝑥 be equi-spaced such that 𝑥𝑖 = 𝑥0 + 𝑖ℎ (𝑖 = 1,2,3, …).
Assuming 𝑦(𝑥) to be polynomial of the 𝑛𝑡ℎ degree in 𝑥 such that
𝑦(𝑥0 ) = 𝑦0 , 𝑦(𝑥1 ) = 𝑦1 , … , 𝑦(𝑥𝑛 ) = 𝑦𝑛
Let the polynomial representation of(𝑥0 , 𝑦0 ), (𝑥1 , 𝑦1 ) ,… , (𝑥𝑛 , 𝑦𝑛 ) be

𝑦(𝑥) = 𝑎0 + 𝑎1 (𝑥 − 𝑥0 ) + 𝑎2 (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) + 𝑎3 (𝑥 − 𝑥0 )(𝑥 − 𝑥1 )(𝑥 − 𝑥2 ) + ⋯ +


𝑎𝑛 (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) … (𝑥 − 𝑥𝑛−1 ) ………………………..(*)
Putting 𝒙 = 𝒙𝟎 and solve for 𝑎0 we get 𝒂𝟎 = 𝒚𝟎
Putting 𝒙 = 𝒙𝟏 and solve for 𝑎1 we get

𝑦(𝑥1 ) = 𝑎0 + 𝑎1 (𝑥1 − 𝑥0 )
⟹ 𝑦1 = 𝑦0 + 𝑎1 (ℎ)
𝑦1 −𝑦0
⟹ 𝑎1 = ℎ
∆𝒚𝟎
⟹ 𝒂𝟏 = 𝒉

∆𝟐 𝒚𝟎 ∆𝟑 𝒚𝟎 ∆𝒏 𝒚𝟎
In a similar manner we get 𝒂𝟐 = , 𝒂𝟑 = ,…, 𝒂𝒏 =
𝟐𝒉𝟐 𝟑!𝒉𝟑 𝒏!𝒉𝒏

Substituting these values of 𝑎𝑖 ′𝑠 in (*)we obtain


∆𝒚𝟎 ∆𝟐 𝒚𝟎 ∆𝟑 𝒚𝟎
𝑦(𝑥) = 𝑦0 + (𝑥 − 𝑥0 ) + (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) + (𝑥 − 𝑥0 )(𝑥 − 𝑥1 )(𝑥 − 𝑥2 ) + ⋯ +
𝒉 𝟐𝒉𝟐 𝟑!𝒉𝟑
∆𝒏 𝒚𝟎
𝒂𝒏 = (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) … (𝑥 − 𝑥𝑛−1 )
𝒏!𝒉𝒏
𝑥−𝑥0
Hence let 𝑃 = ℎ
𝒙−𝒙𝒏 𝒙−𝒙𝟎 +𝒙𝟎 −𝒙𝒏 𝒙−𝒙𝟎 𝒙𝟎 −𝒙𝒏 𝒙−𝒙𝟎 𝒙𝒏 −𝒙𝟎 𝒙−𝒙𝟎 𝒏𝒉
= = + = − = − =𝑷−𝒏
𝒉 𝒉 𝒉 𝒉 𝒉 𝒉 𝒉 𝒉

So (*) is written in the form of


𝑷(𝑷−𝟏) 𝑷(𝑷−𝟏)(𝑷−𝟐) 𝑷(𝑷−𝟏)(𝑷−𝟐)...(𝑷−𝒏+𝟏)
𝒚(𝒙) = 𝒚𝟎 + 𝑷∆𝒚𝟎 + ∆ 𝟐 𝒚𝟎 + ∆ 𝟑 𝒚𝟎 + ⋯ + ∆ 𝒏 𝒚𝟎
𝟐! 𝟑! 𝒏!

is called Newton’s Forward Interpolation Formula


We use Newton’s Forward Interpolation Formula at the beginning of the interval.

9
Example
1
a. Construct the difference table for 𝑓(𝑥) = 𝑥 , 𝑥 ∈ [2,2.5] & ℎ = 0.1
b. Using NFIF Find i. 𝑓(2.25) ii. 𝑓(2.35)
Solution
a.

𝑥 𝑦 = 𝑓(𝑥) ∆ ∆2 ∆3 ∆4 ∆5
2 0.5
-0.024
2.1 0.476 0.003
-0.021 -0.002
2.2 0.455 0.001 0.003
-0.020 0.001 −0.005
2.3 0.435 0.002 −0.002
-0.018 -0.001
2.4 0.417 0.001
-0.017
2.5 0.400

𝑥−𝑥0 2.25−2.2
𝑏 𝑖. 𝑃 = for ℎ = 0.1 , 𝑥 = 2.25 𝑎𝑛𝑑 𝑥0 = 2.2 ⟹ 𝑃 = = 0.5
ℎ 0.1
𝑷(𝑷−𝟏) 𝟐 𝑷(𝑷−𝟏)(𝑷−𝟐) 𝟑 𝑷(𝑷−𝟏)(𝑷−𝟐)...(𝑷−𝒏+𝟏) 𝒏
𝑓(𝑥) = 𝑦(𝑥) = 𝒚𝟎 + 𝑷∆𝒚𝟎 + 𝟐!
∆ 𝒚𝟎 + 𝟑!
∆ 𝒚𝟎 + ⋯+ 𝒏!
∆ 𝒚𝟎

𝟎.𝟓(𝟎.𝟓−𝟏) 𝟎.𝟓(𝟎.𝟓−𝟏)(𝟎.𝟓−𝟐)
⟹ 𝑦(2.25) = 0.445 + (0.5)(−0.020) + (𝟎. 𝟎𝟎𝟐) + (−𝟎. 𝟎𝟎𝟏) + ⋯ +
𝟐! 𝟑!

𝑦(2.25) ≅ 0.4446875
1
The exact value is 𝑦(2.25) = 2.25 = 0.44444 ….

Example
Find the 4th degree polynomial which takes the following values

𝒙 0 2 4 6 8 10
𝒚 1 2 1 10 9 ?
Solution
𝒙−𝟎 𝒙
𝑷= =𝟐
𝟐

65 5 𝑥4
𝑦(𝑥) = 1 + 7𝑥 − 12 𝑥 2 + 4 𝑥 3 − 12

10
Example

From the table estimate the number of students who obtained marks b/n 40 and 45.

Mark 30-40 40-50 50-60 60-70 70-80


No.of student 31 42 51 35 31
Solution

First we prepare the cumulative frequency table as follows

Mark less than (x) 40 50 60 70 80


No.of student 31 73 124 159 190
Prepare a forward difference table as follows

𝑥 𝑦𝑥 ∆𝑦𝑥 ∆2 𝑦𝑥 ∆3 𝑦𝑥 ∆4 𝑦𝑥
40 31
42
50 73 9
51 -25
60 124 -16 37
35 12
70 159 -4
31
80 190
We shall find 𝑦45 that is number of students with mark less than 45.
𝑥−𝑥0 5
Taking 𝑥0 = 40, 𝑥 = 45 we have𝑝 = = = 0.5
ℎ 10

Using NFIF we get


𝑷(𝑷 − 𝟏) 𝟐 𝑷(𝑷 − 𝟏)(𝑷 − 𝟐) 𝟑 𝑷(𝑷 − 𝟏)(𝑷 − 𝟐)(𝑷 − 𝟑) 𝟒
𝑦45 = 𝒚𝟒𝟎 + 𝑷∆𝒚𝟎 + ∆ 𝒚𝟒𝟎 + ∆ 𝒚𝟒𝟎 + ∆ 𝒚𝟒𝟎
𝟐! 𝟑! 𝟒!
𝟎.𝟓(𝟎.𝟓−𝟏) 𝟎.𝟓(𝟎.𝟓−𝟏)(𝟎.𝟓−𝟐) 𝟎.𝟓(𝟎.𝟓−𝟏)(𝟎.𝟓−𝟐)(𝟎.𝟓−𝟑)
𝑦45 = 31 + (0.5)(42) + 𝟐!
(𝟗) +
𝟑!
(−𝟐𝟓) +
𝟒!
(𝟑𝟕)

𝑦45 = 47.87
The number of students with marks less than 45 is 47.87 i.e.,48.

But the number of students with marks less than 40 is 31.

Hence the number of students getting marks b/n 40 and 45 is 48-31=17.

Finding missing terms

Example

Determine the missing entry in the table below

11
X 0 1 2 3 4
Y 1 4 17 ? 97

Solution

Let 𝑦 be a polynomial of degree 3.

Hence ∆4 𝑦0 = 0 ⟹ (𝐸 − 1)4 𝑦0 = (𝐸 4 − 4𝐸 3 + 6𝐸 2 − 4𝐸 + 1)𝑦0 = 0

⟹ 𝐸 4 𝑦0 − 4𝐸 3 𝑦0 + 6𝐸 2 𝑦0 − 4𝐸𝑦0 + 𝑦0 = 0
⟹ 𝑦4 − 4𝑦3 + 6𝑦2 − 4𝑦1 + 𝑦0 = 0
⟹ 97 − 4𝑦3 + 6(17) − 4(4) + 1 = 0 ⟹ 𝑦3 = 46
Example

Find the missing entries in the following table

X 0 1 2 3 4 5
Y 1 ? 11 28 ? 116

𝑦0 = 1, 𝑦2 = 11, 𝑦3 = 28, 𝑎𝑛𝑑 𝑦5 = 116 . since four values are known we assume 𝑦 = 𝑓(𝑥)as a
polynomial of degree three.

Hence ∆4 𝑦0 = 0 and ∆5 𝑦0 = 0

⟹ (𝐸 4 − 4𝐸 3 + 6𝐸 2 − 4𝐸 + 1)𝑦0 = 0 and (𝐸 5 − 5𝐸 4 + 10𝐸 3 − 10𝐸 2 + 5𝐸 − 1)𝑦0 = 0

⟹ 𝑦4 − 4𝑦3 + 6𝑦2 − 4𝑦1 + 𝑦0 = 0 and 𝑦5 − 5𝑦4 + 10𝑦3 − 10𝑦2 + 5𝑦1 − 𝑦0 = 0


𝑦4 − 4𝑦1 = 45
{
−5𝑦4 + 5𝑦1 = −285
⟹ 𝑦1 = 4 𝑎𝑛𝑑 𝑦4 = 6

4.3.2. Newton’s Backward Interpolation Formula


To interpolate a value of 𝑓(𝑥) near to the end of the tabular values we use the following
polynomial

𝑦(𝑥) = 𝑎0 + 𝑎1 (𝑥 − 𝑥𝑛 ) + 𝑎2 (𝑥 − 𝑥𝑛 )(𝑥 − 𝑥𝑛−1 ) + 𝑎3 (𝑥 − 𝑥𝑛 )(𝑥 − 𝑥𝑛−1 )(𝑥 − 𝑥𝑛−2 ) + ⋯ +


𝑎𝑛 (𝑥 − 𝑥𝑛 )(𝑥 − 𝑥𝑛−1 ) … (𝑥 − 𝑥1 ) ……………….(**)
Now let us find 𝑎0 , 𝑎1 , … , 𝑎𝑛
Putting 𝒙 = 𝒙𝒏 and solve for 𝑎0 we get 𝒂𝟎 = 𝒚𝒏
Putting 𝒙 = 𝒙𝒏−𝟏 and solve for 𝑎1 we get
𝑦(𝑥𝑛−1 ) = 𝑎0 + 𝑎1 (𝑥𝑛−1 − 𝑥𝑛 )

12
⟹ 𝑦𝑛−1 = 𝑦𝑛 + 𝑎1 (ℎ)
𝑦 −𝑦 𝑦 −𝑦
⟹ 𝑎1 = 𝑥𝑛−1 −𝑥𝑛 = 𝑥𝑛−𝑥𝑛−1
𝑛−1 𝑛 𝑛 𝑛−1

𝛁𝒚𝒏
⟹ 𝒂𝟏 = 𝒉

𝛁𝟐 𝒚𝒏 𝛁𝟑 𝒚𝒏 𝛁𝒏 𝒚𝒏
In a similar manner we get 𝒂𝟐 = , 𝒂𝟑 = ,…, 𝒂𝒏 =
𝟐𝒉𝟐 𝟑!𝒉𝟑 𝒏!𝒉𝒏
𝒙−𝒙𝒏
Let 𝑷 = 𝒉
𝒙−𝒙𝒏−𝒊 𝒙−𝒙𝒏 +𝒙𝒏 −𝒙𝒏−𝒊 𝒙−𝒙𝒏 𝒙𝒏 −𝒙𝒏−𝒊 𝒙−𝒙𝒏 𝒊𝒉
= = + = + =𝑷+𝒊
𝒉 𝒉 𝒉 𝒉 𝒉 𝒉

𝑥−𝑥𝑛 (𝑥−𝑥𝑛 )(𝑥−𝑥𝑛−1 ) (𝑥−𝑥𝑛 )(𝑥−𝑥𝑛−1 )((𝑥−𝑥𝑛−2 ))


⟹ = 𝑃, = 𝑃(𝑝 + 1), = 𝑃(𝑝 + 1)(𝑃 + 2) and so on.
ℎ ℎ2 ℎ3

Substituting these values of 𝑎𝑖 ′𝑠 and 𝑷 + 𝒊 in (**) in their respective place we obtain


𝑷(𝑷+𝟏) 𝑷(𝑷+𝟏)(𝑷+𝟐) 𝑷(𝑷+𝟏)…(𝑷+(𝒏−𝟏))
𝑦(𝑥) = 𝒚𝒏 + 𝑷𝛁𝒚𝒏 + 𝛁 𝟐 𝒚𝒏 + 𝛁 𝟑 𝒚𝒏 + ⋯ + 𝛁 𝒏 𝒚𝒏
𝟐! 𝟑! 𝒏!

Example
The population of a town in the census was as given below.

Year,𝑥 1891 1901 1911 1921 1931


Population,𝑦(in 46 66 81 93 101
thousands)

13
Solution
First we construct the difference table

𝑥 𝑦 = 𝑓(𝑥) ∇ ∇2 ∇3 ∇4
1891 46
∇𝑦1 =20
1901 66 ∇2 𝑦2 =-5
∇𝑦2 =15 ∇3 𝑦3 =2
1911 81 ∇2 𝑦3 =-3 ∇4 𝑦4 =-3
∇𝑦3 =12 3
∇ 𝑦4 =-1
1921 93 2
∇ 𝑦4 =-4
∇𝑦4 =8
1931 101

𝒙−𝒙𝒏 𝟏𝟗𝟐𝟓−𝟏𝟗𝟑𝟏
𝑷= = = −𝟎. 𝟔
𝒉 𝟏𝟎

𝑷(𝑷+𝟏) 𝑷(𝑷+𝟏)(𝑷+𝟐)
𝑦(𝑥) = 𝒚𝒏 + 𝑷𝛁𝒚𝒏 + 𝛁 𝟐 𝒚𝒏 + 𝛁 𝟑 𝒚𝒏 + ⋯
𝟐! 𝟑!
(−𝟎.𝟔)(−𝟎.𝟔+𝟏) (−𝟎.𝟔)(−𝟎.𝟔+𝟏)(−𝟎.𝟔+𝟐)
𝑦(1925) = 𝟏𝟎𝟏 + (−𝟎. 𝟔)(𝟖) + (−𝟒) + (−𝟏) +
𝟐! 𝟑!
(−𝟎.𝟔)(−𝟎.𝟔+𝟏)(−𝟎.𝟔+𝟐)(−𝟎.𝟔+𝟑)
(−𝟑)
𝟒!

𝒚(𝟏𝟗𝟐𝟓) ≅ 𝟗𝟔. 𝟖𝟑𝟔𝟖 𝒕𝒉𝒐𝒖𝒔𝒂𝒏𝒅𝒔 ≅ 𝟗𝟔, 𝟖𝟑𝟕


Exercise
1. The table below gives the value of 𝐭𝐚𝐧 𝒙 for 𝟎. 𝟏𝟎 ≤ 𝒙 ≤ 𝟎. 𝟑𝟎

𝑥 0.10 0.15 0.20 0.25 0.30


𝑦 = 𝑡𝑎𝑛𝑥 0.1003 0.1511 0.2027 0.2553 0.3093
Find
a. 𝑡𝑎𝑛0.12 b. 𝑡𝑎𝑛0.26 c. 𝑡𝑎𝑛0.5
2. Find the number of men getting wages b/n Rs.10 and Rs.15 from the following data:
Wages in Rs. 0-10 10-20 20-30 30-40
Frequency 9 30 35 42

14
4.4.Interpolation With Unequally spaced point
When the values of the argument are not at equally spaced then we use two such formulae for
interpolation.

1. Lagrange’s interpolation formula


2. Newton’s divided difference formula
4.4.1. Lagrange’s Interpolation Formula
Let 𝑦0 = 𝑓(𝑥0 ), 𝑦1 = 𝑓(𝑥1 ), … , 𝑦𝑛 = 𝑓(𝑥𝑛 ) 𝑏𝑒 (𝑛 + 1) entries of a function 𝑦 = 𝑓(𝑥). Let 𝑝(𝑥)
be a polynomial of degree 𝑛 corresponding to the arguments 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 which can be
written as:

𝑃𝑛 (𝑥) = 𝐴0 (𝑥 − 𝑥1 )(𝑥 − 𝑥2 ) … (𝑥 − 𝑥𝑛 ) + 𝐴1 (𝑥 − 𝑥0 )(𝑥 − 𝑥2 ) … (𝑥 − 𝑥𝑛 ) + ⋯ +


𝐴𝑛 (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) … (𝑥 − 𝑥𝑛−1 ) ………………..(1)
Where 𝐴0 , 𝐴1 , … , 𝐴𝑛 are constants to be determined.
The constants𝐴0 , 𝐴1 , … , 𝐴𝑛 , will be determined by considering the tabulated function 𝑦 = 𝑓(𝑥)
and the polynomial function 𝑃(𝑥) agree at the set of tabulated points.
Putting 𝑥 = 𝑥0 , 𝑥1 , 𝑥2 , … , 𝑥𝑛 in(1) successively, we get the following.
For 𝑥 = 𝑥0 , 𝑦0 = 𝐴0 (𝑥0 − 𝑥1 )(𝑥0 − 𝑥2 ) … (𝑥0 − 𝑥𝑛 )
𝑦0
That is 𝐴0 = (𝑥
0 −𝑥1 )(𝑥0 −𝑥2 )…(𝑥0 −𝑥𝑛 )

For 𝑥 = 𝑥1 , 𝑦1 = 𝐴1 (𝑥1 − 𝑥0 )(𝑥1 − 𝑥2 ) … (𝑥1 − 𝑥𝑛 )


𝑦1
That is 𝐴1 = (𝑥
1 −𝑥0 )(𝑥1 −𝑥2 )…(𝑥1 −𝑥𝑛 )

Similarly
For 𝑥 = 𝑥𝑛 , 𝑦𝑛 = 𝐴𝑛 (𝑥𝑛 − 𝑥0 )(𝑥𝑛 − 𝑥2 ) … (𝑥𝑛 − 𝑥𝑛−1 )
𝑦𝑛
That is 𝐴𝑛 = (𝑥
𝑛 −𝑥0 )(𝑥𝑛 −𝑥2 )…(𝑥𝑛 −𝑥𝑛−1 )

Substituting the values of 𝐴0 , 𝐴1 , … , 𝐴𝑛 in equation (1), we get


(𝑥−𝑥1 )(𝑥−𝑥2 )…(𝑥−𝑥𝑛 )
𝑃𝑛 (𝑥) = (𝑥 𝑦0 +
0 −𝑥1 )(𝑥0 −𝑥2 )…(𝑥0 −𝑥𝑛 )

(𝑥−𝑥0 )(𝑥−𝑥2 )…(𝑥−𝑥𝑛 )


(𝑥1 −𝑥0 )(𝑥1 −𝑥2 )…(𝑥1 −𝑥𝑛 )
𝑦1 + ⋯ +

(𝑥−𝑥0 )(𝑥−𝑥1 )…(𝑥−𝑥𝑛−1 )


(𝑥𝑛 −𝑥0 )(𝑥𝑛 −𝑥2 )…(𝑥𝑛 −𝑥𝑛−1 )
𝑦𝑛 …………………….(2)

This is called Lagrange’s interpolation formula and which can be written as a general form:

15
𝑛

𝑃𝑛 (𝑥) = ∑ 𝐿𝑖 (𝑥)𝑓𝑖
𝑖=0

Where
𝑥−𝑥 (𝑥−𝑥0 )…(𝑥−𝑥𝑖−1 )(𝑥−𝑥𝑖+1 )…(𝑥−𝑥𝑛 )
𝐿𝑖 (𝑥) = ∏𝑛𝑗=0,𝑗≠𝑖 (𝑥 −𝑥𝑗 ) = (𝑥 −𝑥 ,( 𝑖 = 0,1,2, … , 𝑛) are
𝑖 𝑗 𝑖 0 )(𝑥𝑖 −𝑥1 )…(𝑥𝑖 −𝑥𝑖−1 )(𝑥𝑖 −𝑥𝑖+1 )…(𝑥𝑖 −𝑥𝑛 )

individually polynomials of degree 𝑛 in 𝑥 and are called the Lagrange’s interpolation


coefficients.
1 𝑖𝑓 𝑖 = 𝑗
Note: 𝐿𝑖 (𝑥𝑗 ) = {
0 𝑖𝑓 𝑖 ≠ 𝑗

Note: The Lagrange’s interpolation formula can also be used to split the given function into
partial fractions.
Dividing both sides of equation (2) by (𝑥 − 𝑥0 )(𝑥 − 𝑥1 )(𝑥 − 𝑥2 ). … . (𝑥 − 𝑥𝑛 ) we get
𝑓(𝑥) 𝑦0 1 𝑦1 1
(𝑥−𝑥0 )(𝑥−𝑥1 )(𝑥−𝑥2 ).….(𝑥−𝑥𝑛 )
= (𝑥 )(𝑥 ).….(𝑥 )
. +(𝑥 )(𝑥 ).….(𝑥 )
. + ⋯+
0 −𝑥1 0 −𝑥2 0 −𝑥𝑛 𝑥−𝑥0 1 −𝑥0 1 −𝑥2 1 −𝑥𝑛 𝑥−𝑥1

𝑦𝑛 1
.
(𝑥𝑛 −𝑥0 )(𝑥𝑛 −𝑥1 )(𝑥𝑛 −𝑥2 ).….(𝑥𝑛 −𝑥𝑛−1 ) 𝑥−𝑥𝑛

Example

Use Lagrange’s method of interpolation to find the unique polynomial 𝑃(𝑥) of degree 2 such
that : 𝑃(1) = 1, 𝑃(3) = 27, 𝑃(4) = 64.
Solution
𝑥0 = 1, 𝑥1 = 3, 𝑥2 = 4 𝑎𝑛𝑑 𝑦0 = 1, 𝑦1 = 27, 𝑦2 = 64
(𝑥−𝑥1 )(𝑥−𝑥2 ) (𝑥−𝑥0 )(𝑥−𝑥2 ) (𝑥−𝑥0 )(𝑥−𝑥1 )
𝑃2 (𝑥) = (𝑥 𝑦0 + (𝑥 𝑦1 + (𝑥 𝑦2
0 −𝑥1 )(𝑥0 −𝑥2 ) 1 −𝑥0 )(𝑥1 −𝑥2 ) 2 −𝑥0 )(𝑥2 −𝑥1 )

(𝑥−3)(𝑥−4) (𝑥−1)(𝑥−4) (𝑥−1)(𝑥−3)


𝑃2 (𝑥) = (1−3)(1−4) (1) + (3−1)(3−4) (27) + (4−1)(4−3) (64) ⟹ 𝑷𝟐 (𝒙) = 𝟖𝒙𝟐 − 𝟏𝟗𝒙 + 𝟏𝟐

Exercise

1. The percentage of criminals for different age group are given below. Determine the
percentage number of criminals under 35 year using the Lagrange’s formula.

16
Age % No. of criminals
Under25 year 52
Under30year 67.3
Under40 year 84.1
Under50 year 94.4
Answer 77.405

2. using Lagrange’s formula express the function


𝑥 2 +6𝑥−1
(𝑥 2 −1)(𝑥−4)(𝑥−6)
as a sum of partial fractions.
[ℎ𝑖𝑛𝑡 𝑡𝑎𝑏𝑢𝑙𝑎𝑡𝑒 𝑡ℎ𝑒 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑓(𝑥) = 𝑥 2 + 6𝑥 − 1 𝑓𝑜𝑟 𝑥 = −1,1,4,6]
Solution
𝑓(𝑥) = 𝑥 2 + 6𝑥 − 1 ,𝑥0 = −1, 𝑥1 = 1, 𝑥2 = 4, 𝑥3 = 6 𝑎𝑛𝑑 𝑦0 = −6, 𝑦1 = 6, 𝑦2 = 39, 𝑦3 = 71
𝑥 2 +6𝑥−1 𝑥 2 +6𝑥−1
(𝑥 2 −1)(𝑥−4)(𝑥−6)
= (𝑥+1)(𝑥−1)(𝑥−4)(𝑥−6)

𝑥 2 +6𝑥−1 𝑦0 1 𝑦1 1
(𝑥+1)(𝑥−1)(𝑥−4)(𝑥−6)
=(𝑥 . + (𝑥 . +
0 −𝑥1 )(𝑥0 −𝑥2 )(𝑥0 −𝑥3 ) 𝑥−𝑥0 1 −𝑥0 )(𝑥1 −𝑥2 )(𝑥1 −𝑥3 ) 𝑥−𝑥1

𝑦2 1 𝑦3 1
.
(𝑥2 −𝑥0 )(𝑥2 −𝑥1 )(𝑥2 −𝑥3 ) 𝑥−𝑥2
+ (𝑥 . 𝑥−𝑥
3 −𝑥0 )(𝑥3 −𝑥1 )(𝑥3 −𝑥2 ) 3

𝑥 2 +6𝑥−1 3 1 1 1 13 1 71 1
(𝑥+1)(𝑥−1)(𝑥−4)(𝑥−6)
= 35 (𝑥+1) + 5 (𝑥−1) − 10 (𝑥−4) + 70 (𝑥−6)

Inverse Interpolation
In interpolation, we estimate the missing value of the function y = f (x) corresponding to a value x intermediate
between two given values. In inverse interpolation, we interpolate the argument x corresponding to an
intermediate value y of the entry.

From
(𝑥−𝑥1 )(𝑥−𝑥2 )…(𝑥−𝑥𝑛 )
𝑦 = (𝑥 𝑦0 +
0 −𝑥1 )(𝑥0 −𝑥2 )…(𝑥0 −𝑥𝑛 )

(𝑥−𝑥0 )(𝑥−𝑥2 )…(𝑥−𝑥𝑛 )


(𝑥1 −𝑥0 )(𝑥1 −𝑥2 )…(𝑥1 −𝑥𝑛 )
𝑦1 + ⋯ +

(𝑥−𝑥0 )(𝑥−𝑥1 )…(𝑥−𝑥𝑛−1 )


(𝑥𝑛 −𝑥0 )(𝑥𝑛 −𝑥2 )…(𝑥𝑛 −𝑥𝑛−1 )
𝑦𝑛

By interchanging 𝑥 and 𝑦 we get


(𝑦−𝑥1 )(𝑦−𝑦2 )…(𝑦−𝑦𝑛 )
𝑥 = (𝑦 𝑥0 +
0 −𝑦1 )(𝑦0 −𝑦2 )…(𝑦0 −𝑦𝑛 )

(𝑦−𝑦0 )(𝑦−𝑦2 )…(𝑦−𝑦𝑛 )


(𝑦1 −𝑦0 )(𝑦1 −𝑦2 )…(𝑦1 −𝑦𝑛 )
𝑥1 + ⋯ +

(𝑦−𝑦0 )(𝑦−𝑦1 )…(𝑦−𝑦𝑛−1 )


(𝑦𝑛 −𝑦0 )(𝑦𝑛 −𝑦2 )…(𝑦𝑛 −𝑦𝑛−1 )
𝑥𝑛

17
Which is called Inverse Interpolation formula.
Example
The following table gives the values of y corresponding to certain values of x. Find the value of x when
y = 167.59789 by applying Lagrange’s inverse interpolation formula.

x 1 2 5 7
y 1 12 117 317
Solution:
Here x0 = 1, x1 = 2, x2 = 5, x3 = 7, y0 = 1, y1 = 12, y2 = 117, y3 = 317 and y = 167.59789.
By using the inverse Lagrange’s formula we can get x = 5.65238 when y = 167.59789

4.4.2. Newton’s General Interpolation Formula with Divided Difference


Lagrange’s interpolation formula has the disadvantage that if any other interpolation point
were added, the interpolation coefficient will have to be recalculated.
This labor of recomputing the interpolation coefficients is saved by using Newton’s divided
differences interpolation formula.
Let the function 𝑦 = 𝑓(𝑥) take the values 𝑦0 , 𝑦1 , … , 𝑦𝑛 corresponding to the values
𝑥0 , 𝑥1 , … , 𝑥𝑛 which are not equally spaced. The difference of the function values with respect to
the difference of the arguments is called divided differences.
𝒚𝟏 −𝒚𝟎 𝒚 −𝒚 𝒚 −𝒚
[𝒙𝟎 , 𝒙𝟏 ] = , [𝒙𝟏 , 𝒙𝟐 ] = 𝒙𝟐 −𝒙𝟏 , … , [𝒙𝒏−𝟏 , 𝒙𝒏 ] = 𝒙𝒏 −𝒙𝒏−𝟏 are called first order divided
𝒙𝟏 −𝒙𝟎 𝟐 𝟏 𝒏 𝒏−𝟏
difference.
𝒚 −𝒚 𝒚 −𝒚
[𝒙𝒏 , 𝒙𝒏+𝟏 ]−[𝒙𝒏−𝟏 , 𝒙𝒏 ] ( 𝒏+𝟏 𝒏 − 𝒏 𝒏−𝟏 )
𝒙𝒏+𝟏 −𝒙𝒏 𝒙𝒏 −𝒙𝒏−𝟏
[𝒙𝒏−𝟏 , 𝒙𝒏 , 𝒙𝒏+𝟏 ] = = is called the second order divided
𝒙𝒏+𝟏 −𝒙𝒏−𝟏 𝒙𝒏+𝟏 −𝒙𝒏−𝟏
difference for the arguments 𝑥𝑛−1 , 𝑥𝑛 , 𝑥𝑛+1

In general the 𝑛𝑡ℎ order divided differences for the arguments 𝑥0 , 𝑥1 , … , 𝑥𝑛 is given by
[𝒙𝟏 ,𝒙𝟐 ,…,𝒙𝒏 ]−[𝒙𝟎 ,𝒙𝟏 ,…,𝒙𝒏−𝟏 ]
[𝒙𝟎 , 𝒙𝟏 , … , 𝒙𝒏 ] =
𝒙𝒏 −𝒙𝟎

Properties of Divided Differences


1. Divided differences are symmetric with respect to the arguments i.e., independent of
the order of arguments.
i.e., [𝒙𝟎 , 𝒙𝟏 ] = [𝒙𝟏 , 𝒙𝟎 ], [𝒙𝒏−𝟏 , 𝒙𝒏 , 𝒙𝒏+𝟏 ] = [𝒙𝒏+𝟏 , 𝒙𝒏−𝟏 , 𝒙𝒏 ] = [𝒙𝒏 , 𝒙𝒏+𝟏 , 𝒙𝒏−𝟏 ].

2. The 𝑛𝑡ℎ order divided differences of a polynomial of 𝑛𝑡ℎ degree are constant

Newton’s Divided Difference Formula


Let 𝑦0 , 𝑦1 , … , 𝑦𝑛 be the values of 𝑦 = 𝑓(𝑥) corresponding to the arguments𝑥0 , 𝑥1 , … , 𝑥𝑛 .

18
Then from the definition of divided differences, we have
𝒚−𝒚𝟎
[𝒙, 𝒙𝟎 ] =
𝒙−𝒙𝟎

⟹ 𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙, 𝒙𝟎 ] …………..(𝒊)
[𝒙, 𝒙𝟎 ]−[𝒙𝟎 , 𝒙𝟏 ]
[𝒙, 𝒙𝟎 , 𝒙𝟏 ] = ⟹ [𝒙, 𝒙𝟎 ] = [𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟏 )[𝒙, 𝒙𝟎 , 𝒙𝟏 ]
𝒙−𝒙𝟏

Substituting this value of [𝒙, 𝒙𝟎 ]in(𝒊),we get


𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )[𝒙, 𝒙𝟎 , 𝒙𝟏 ] …………….(𝒊𝒊)
[𝒙, 𝒙𝟎 , 𝒙𝟏 ]−[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ]
[𝒙, 𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] =
𝒙−𝒙𝟐

⟹ [𝒙, 𝒙𝟎 , 𝒙𝟏 ] = [𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] + (𝒙 − 𝒙𝟐 )[𝒙, 𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ]


Substituting this value of [𝒙, 𝒙𝟎 , 𝒙𝟏 ] in(𝒊𝒊),we get

𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )([𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] + (𝒙 − 𝒙𝟐 )[𝒙, 𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] )


𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )[𝒙, 𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ]

Proceeding in this manner, we get


𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] +
(𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ] +
(𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 ) … (𝒙 − 𝒙𝒏−𝟏 )[𝑥0 , 𝑥1 , … , 𝑥𝑛 ] +

(𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 ) … (𝒙 − 𝒙𝒏 )[𝑥, 𝑥0 , 𝑥1 , … , 𝑥𝑛 ]

Is called Newton’s general interpolation formula with divide differences.

19
Table 4.4: Table of Divided Differences

𝑥 𝑦 1st DD 2nd DD 3rd DD 4th DD 5th DD


= 𝑓(𝑥)
𝑦0
𝑥0 [𝑥0 , 𝑥1 ]
𝑦1 [𝑥0 , 𝑥1 , 𝑥2 ]
𝑥1 [𝑥1 , 𝑥2 ] [𝑥0 , 𝑥1 , 𝑥2 , 𝑥3 ]
𝑦2 [𝑥1 , 𝑥2 , 𝑥3 ] [𝑥0 , 𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 ]
𝑥2 [𝑥2 , 𝑥3 ] [𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 ] [𝑥0 , 𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 ]
𝑦3 [𝑥2 , 𝑥3 , 𝑥4 ] [𝑥1 , 𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 ]
𝑥3 [𝑥3 , 𝑥4 ] [𝑥2 , 𝑥3 , 𝑥4 , 𝑥5 ]
𝑦4 [𝑥3 , 𝑥4 , 𝑥5 ]
𝑥4 [𝑥4 , 𝑥5 ]
𝑦5
𝑥5

Example
Apply Newton’s divided difference formula find 𝑓(8) if
𝑓(1) = 3, 𝑓(3) = 31, 𝑓(6) = 223, 𝑓(10) = 1011, 𝑓(11) = 1343
Solution

𝑥 𝑦 = 𝑓(𝑥) 1st DD 2nd DD 3rd DD 4th DD


1 3
31−3
=14
3−1

31 64−14
3 =10
6−1 19−10
223−31
=64 =1
10−1
6−3
6 223 197−64 1−1
=19 =0
10−3 27−19 11 − 1
1011−223
=197 =1
11−3
10−6 332−197
10 1011 =27
11−6
1343−1011
=332
11−10
11 1343

20
𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ]
+ (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )(𝒙 − 𝒙𝟑 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 , 𝒙𝟒 ]
𝒚 = 𝟑 + (𝒙 − 𝟏)𝟏𝟒 + (𝒙 − 𝟏)(𝒙 − 𝟑)𝟏𝟎 + (𝒙 − 𝟏)(𝒙 − 𝟑)(𝒙 − 𝟔)𝟏 + (𝒙 − 𝟏)(𝒙 − 𝟑)(𝒙 − 𝟔)(𝒙 − 𝟏𝟎)𝟎

𝒚 = 𝟑 + (𝒙 − 𝟏)𝟏𝟒 + (𝒙 − 𝟏)(𝒙 − 𝟑)𝟏𝟎 + (𝒙 − 𝟏)(𝒙 − 𝟑)(𝒙 − 𝟔)𝟏

𝒚(𝟖) = 𝟑 + (𝟖 − 𝟏)𝟏𝟒 + (𝟖 − 𝟏)(𝟖 − 𝟑)𝟏𝟎 + (𝟖 − 𝟏)(𝟖 − 𝟑)(𝟖 − 𝟔)𝟏 = 𝟓𝟐𝟏

Exercise
From the following data estimate the number of persons having incomes b/n 2000 and 2500:

Income Below 500 500-1000 1000-2000 2000-3000 3000-4000


No. of 6000 4250 3600 1500 650
persons

Solution
First we prepare the cumulative frequency table as follows

Mark less than (x) 500 1000 2000 3000 4000


No.of persons 6000 10250 13850 15350 16000
We have to draw a Newton’s divided difference interpolation table.

X y 1st DD 2nd DD 3rd DD 4th DD


500 6000
8.5
1000 10250 -0.00326
3.6 0.00000084
2000 13850 -0.00105 0
1.5 0.000000208
3000 15350 -0.000425
0.65
4000 16000

𝒚 = 𝒚𝟎 + (𝒙 − 𝒙𝟎 )[𝒙𝟎 , 𝒙𝟏 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 ] + (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 ]


+ (𝒙 − 𝒙𝟎 )(𝒙 − 𝒙𝟏 )(𝒙 − 𝒙𝟐 )(𝒙 − 𝒙𝟑 )[𝒙𝟎 , 𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 , 𝒙𝟒 ]
𝒚 = 𝟔𝟎𝟎𝟎 + (𝒙 − 𝟓𝟎𝟎)𝟖. 𝟓 + (𝒙 − 𝟓𝟎𝟎)(𝒙 − 𝟏𝟎𝟎𝟎)(−𝟎. 𝟎𝟎𝟑𝟐𝟔) + (𝒙 − 𝟓𝟎𝟎)(𝒙 − 𝟏𝟎𝟎𝟎)(𝒙 −
𝟐𝟎𝟎𝟎)𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟖𝟒

21
+𝟎
𝒚 = 𝟔𝟎𝟎𝟎 + (𝟐𝟓𝟎𝟎 − 𝟓𝟎𝟎)𝟖. 𝟓 + (𝟐𝟓𝟎𝟎 − 𝟓𝟎𝟎)(𝟐𝟓𝟎𝟎 − 𝟏𝟎𝟎𝟎)(−𝟎. 𝟎𝟎𝟑𝟐𝟔) + (𝟐𝟓𝟎𝟎 −
𝟓𝟎𝟎)(𝟐𝟓𝟎𝟎 − 𝟏𝟎𝟎𝟎)(𝟐𝟓𝟎𝟎 − 𝟐𝟎𝟎𝟎)𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟖𝟒
+𝟎
𝒚 = 𝟔𝟎𝟎𝟎 + (𝟐𝟎𝟎𝟎)𝟖. 𝟓 + (𝟐𝟎𝟎𝟎)(𝟏𝟓𝟎𝟎)(−𝟎. 𝟎𝟎𝟑𝟐𝟔) + (𝟐𝟎𝟎𝟎)(𝟏𝟓𝟎𝟎)(𝟓𝟎𝟎)𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟖𝟒 + 𝟎
𝒚𝟐𝟓𝟎𝟎 = 𝟏𝟒𝟒𝟖𝟎
That means the number of persons having income less than 2500 is 𝟏𝟒𝟔𝟕𝟓

The number of persons having income less than 2000 is 13850

The number of persons having income b/n 2000 and 2500 is 𝟏𝟒𝟔𝟕𝟓 − 13850 = 825

22

You might also like