Interpolation

Difference
operators
Dr. Hina Dutt
[email protected]
SEECS-NUST
Introduction
For interpolation of a tabulated function, the concept
of finite differences is important. The knowledge
about various finite difference operators and their
symbolic relations are very much needed to establish
various interpolation formulae.
Finite Difference Operators

Forward
difference

Backward
difference

Central difference
Forward Difference
Operator
Forward Difference Operator
For a given table of values 𝑥𝑖 , 𝑦𝑖 , 𝑖 = 0,1, … , 𝑛 with
equally spaced abscissas of a function 𝑦 = 𝑓 𝑥 , we
define the forward difference operator as follows:
First Forward Difference Operator
The first forward differences of 𝑦 = 𝑓 𝑥 , represented by ∆, are
defined as
∆ 𝑓 𝑥𝑖 = 𝑓 𝑥𝑖 + ℎ − 𝑓(𝑥𝑖 )
∆𝑦𝑖 = 𝑦𝑖+1 − 𝑦𝑖 , 𝑖 = 0,1, … , 𝑛 − 1
i.e,
∆𝑦0 = 𝑦1 − 𝑦0
∆𝑦1 = 𝑦2 − 𝑦1
. . .
∆𝑦𝑛−1 = 𝑦𝑛 − 𝑦𝑛−1
Second Forward Difference
Operator
Similarly, the differences of the first differences are called second
differences, represented by ∆2 , and defined by
∆2 𝑦𝑖 = ∆𝑦𝑖+1 − ∆𝑦𝑖 , 𝑖 = 0,1, … , 𝑛 − 2
i.e,
∆2 𝑦0 = ∆𝑦1 − ∆𝑦0
∆2 𝑦1 = ∆𝑦2 − ∆𝑦1
. . .
∆2 𝑦𝑛−2 = ∆𝑦𝑛−1 − ∆𝑦𝑛−2
 𝒕𝒉
𝒌 Forward Difference Operator
In general, 𝒌𝒕𝒉 forward differences of 𝑦 = 𝑓 𝑥 ,
represented by ∆𝑘 , are defined as

∆𝑘 𝑦𝑖 = ∆𝑘−1 𝑦𝑖+1 − ∆𝑘−1 𝑦𝑖 , 𝑖 = 0,1, … , 𝑛 − 𝑘

Forward Difference Table
By defining a difference table as a convenient
device for displaying various differences, the above
defined differences can be written down
systematically by constructing a difference table
for values 𝑥𝑖 , 𝑦𝑖 ; 𝑖 = 0, 1, 2, 3.
Forward Difference Table
𝒙 𝒚 ∆𝒚 ∆𝟐 𝒚 ∆𝟑 𝒚
∆𝟐 𝑦1 − ∆𝟐 𝑦0
𝑥0 𝑦0 𝑦1 − 𝑦0 = ∆𝑦0 ∆𝑦1 − ∆𝑦0 = ∆𝟐 𝑦0 = ∆𝟑 𝑦0

∆𝑦2 − ∆𝑦1 = ∆𝟐 𝑦1
𝑥1 𝑦1 𝑦2 − 𝑦1 = ∆𝑦1

𝑥2 𝑦2 𝑦3 − 𝑦2 = ∆𝑦2
𝑥3 𝑦3
Backward Difference
Operator
Backward Difference Operator
For a given table of values 𝑥𝑖 , 𝑦𝑖 , 𝑖 = 0,1, … , 𝑛 with
equally spaced abscissas of a function 𝑦 = 𝑓 𝑥 , we
define the backward difference operator as follows:
First Backward Difference Operator
The first backward differences of 𝑦 = 𝑓 𝑥 , represented
by 𝛻, are defined as
𝛻 𝑓 𝑥𝑖 = 𝑓 𝑥𝑖 − 𝑓(𝑥𝑖 − ℎ)
𝛻𝑦𝑖 = 𝑦𝑖 − 𝑦𝑖−1 , 𝑖 = 1, … , 𝑛
i.e,
𝛻𝑦1 = 𝑦1 − 𝑦0
𝛻𝑦2 = 𝑦2 − 𝑦1
. . .
𝛻𝑦𝑛 = 𝑦𝑛 − 𝑦𝑛−1
Second Backward Difference
Operator
The second backward differences of 𝑦 = 𝑓(𝑥) are defined as
𝛻 2 𝑦𝑖 = 𝛻𝑦𝑖 − 𝛻𝑦𝑖−1 , 𝑖 = 2, … , 𝑛
i.e,

𝛻 2 𝑦2 = 𝛻𝑦2 − 𝛻𝑦1
. . .
𝛻 2 𝑦𝑛 = 𝛻𝑦𝑛 − 𝛻𝑦𝑛−1
 𝒕𝒉
𝒌 Backward Difference Operator
In general, 𝒌𝒕𝒉 backward differences of 𝑦 = 𝑓(𝑥) are
defined as
𝛻𝑘 𝑦𝑖 = 𝛻𝑘−1 𝑦𝑖 − 𝛻𝑘−1 𝑦𝑖−1 , 𝑖 = 𝑘, … , 𝑛
Backward Difference Table
𝒙 𝒚 𝛁𝒚 𝛁𝟐 𝒚 𝛁𝟑 𝒚
𝑥0 𝑦0

𝑥1 𝑦1 𝑦1 − 𝑦0 = 𝛻𝑦1

𝑥2 𝑦2 𝑦2 − 𝑦1 = 𝛻𝑦2 𝛻𝑦2 − 𝛻𝑦1 = 𝛻 𝟐 𝑦2

𝑥3 𝑦3 𝑦3 − 𝑦2 = 𝛻𝑦3 𝛻𝑦3 − 𝛻𝑦2 = 𝛻 𝟐 𝑦3 𝛻 𝟐 𝑦3 − 𝛻 𝟐 𝑦2

= 𝛻 𝟑 𝑦3
Note
It is observed from the forward and backward
difference tables that for a given table of values
both the tables are same. Practically, there are
no differences among the values of the tables,
but, theoretically they have separate significant.
Relation Between Forward Difference and
Divided Difference
𝑦1 − 𝑦0
𝑦 𝑥0 , 𝑥1 =
𝑥1 − 𝑥0
Since data is equally spaced therefore 𝒙𝟏 − 𝒙𝟎 = 𝒉.
1
𝑦 𝑥0 , 𝑥1 = ∆𝑦0
ℎ
1 1
∆𝑦1 − ∆𝑦0 1 2
𝑦 𝑥0 , 𝑥1 , 𝑥2 = ℎ ℎ = ∆ 𝑦0
2ℎ 2! ℎ 2

Thus
1 𝑛𝑦
𝑦 𝑥0 , 𝑥1 , … , 𝑥𝑛 = ∆ 0
𝑛! ℎ𝑛
Relation Between Backward Difference and
Divided Difference
𝑦1 − 𝑦0
𝑦 𝑥0 , 𝑥1 =
𝑥1 − 𝑥0
Since data is equally spaced therefore 𝒙𝟏 − 𝒙𝟎 = 𝒉.
1
𝑦 𝑥0 , 𝑥1 = 𝛻𝑦1
ℎ
1 1
𝛻𝑦2 − 𝛻𝑦1 1
𝑦 𝑥0 , 𝑥1 , 𝑥2 = ℎ ℎ = 𝛻 2𝑦
2
2ℎ 2! ℎ2
Thus
1 𝑛𝑦
𝑦 𝑥0 , 𝑥1 , … , 𝑥𝑛 = 𝑦 𝑥𝑛 , 𝑥𝑛−1 , … , 𝑥0 = 𝛻 𝑛
𝑛! ℎ𝑛
Example 1
Show that any value of 𝑦 can be expressed
in terms of 𝑦𝑛 and its backward differences.
Example 1
From yi  yi  yi 1i  n, (n  1), ,1

We get yn1  yn  yn yn2  yn1  yn1

From 2 yi  yi  yi 1 , i  n, (n  1),..., 2

We get yn 1  yn   yn 2

Hence yn  2  yn  2yn   yn 2
Example 1
Similarly, we can show that
yn 3  yn  3yn  3 2 yn  3 yn
Symbolically, these equations can be rewritten as
yn 1  (1   ) yn
yn  2  (1   ) 2 yn
yn  3  (1   ) yn 3

.......
yn  r  (1   ) r yn

ynr  yn  C1yn  C2 yn 

n n 2
 (1)  yn
r r
Central Difference
Operator
Central Difference Operator
In some applications, central difference notation
is found to be more convenient to represent the
successive differences of a function.
Here, we use the symbol 𝛿 to represent central
difference operator and the subscript of 𝛿𝑦 for
any difference as the average of the subscripts.
Central Difference Operator
We define central difference operator as
 y1 2  y1  y0 ,  y3 2  y2  y1 ,
In general
 yi  yi  (1 2)  yi (1 2)
Central Difference Operator
Higher order central difference operators are defined as

 yi   yi  (1 2)   yi (1 2)
2

 yi  
n n 1
yi  (1 2)   n 1
yi (1 2)
Central Difference Table
Note: It may be observed that all
odd (even) order differences
have fraction suffices (integral
suffices)
Shift Operator
Shift Operator
The shift operator, represented by 𝐸, is defined as
E 𝑓 𝑥𝑖 = 𝑓 𝑥𝑖 + ℎ
𝐸𝑦𝑖 = 𝑦𝑖+1 , 𝑖 = 0,1, … , 𝑛 − 1,
i.e,
𝐸𝑦0 = 𝑦1
𝐸𝑦1 = 𝑦2
. . .
𝐸𝑦𝑛−1 = 𝑦𝑛
Shift Operator
Note that shift operator increases subscript of 𝑦 by one. When the
shift operator is applied twice on the function 𝑓(𝑥), then the
subscript of 𝑦 is increased by 2, i. e.
𝐸 2 𝑓 𝑥 = 𝐸. 𝐸𝑓 𝑥 = 𝐸𝑓 𝑥 + ℎ = 𝑓 𝑥 + 2ℎ
In general,
𝐸 𝑛 𝑓 𝑥 = 𝑓(𝑥 + 𝑛ℎ)
Shift Operator
The inverse can also be find in similar manner. It is denoted by 𝐸 −1
and is defined by
𝐸 −1 𝑓 𝑥 = 𝐸𝑓 𝑥 − ℎ
In general,
𝐸 −𝑛 𝑓 𝑥 = 𝑓(𝑥 − 𝑛ℎ)
Relations between the Operators
1 1
−
∆= 𝐸 − 1, 𝛻 = 1 − 𝐸 −1 , 𝛿= 𝐸2 −𝐸 2
1
∆= 𝛻𝐸 = 𝛿𝐸 2
1+∆ 1−𝛻 =1