Miladur Rahaman

Roll= 170501

The calculus of finite difference:

The calculus of difference is the study of change that take place in the value of the function or
the dependent variable say y= 𝑓(𝑥), with respect to finite changes in the independent variable x.

Let y= 𝑓(𝑥) = 𝑥 2 be the function of the independent variable x. if there is finite increase , in x , say ∆ x,
then 𝑓(𝑥) = 𝑥 2 will be increased to 𝑓(𝑥 + ∆ x)=(𝑥 + ∆ x)2 , bearing the increment.

𝑓(𝑥 + ∆ x)-𝑓(𝑥) = (𝑥 + ∆ x)2 +𝑥 2

Such finite change are studied in the calculus of finite difference.

Shift Operator:

The operators E, let y= 𝑓(𝑥) be any function of x. By operating E on 𝑓(𝑥) we mean to simply
give an increment to the value of the x in the function 𝑓(𝑥) if this increment be denoted by h,

Then the operation of E on 𝑓(𝑥) means that put x+h in the function 𝑓(𝑥) whenever there is x I e.

Here we should note that E 𝑓(𝑥) does not imply the multiplication of E and 𝑓(𝑥) but it implies that E is
operated on 𝑓(𝑥). The operator is known as the shift operator. By 𝐸 2 𝑓(𝑥) we mean that the operator E
is applied twice on the function 𝑓(𝑥).

𝐸 2 𝑓(𝑥) = 𝐸1 𝐸 𝑓(𝑥)

=E{ 𝑓(𝑥 + ℎ)} by def of E

= 𝑓(𝑥 + ℎ + ℎ) by def of E

=𝑓(𝑥 + 2ℎ).

Similarly 𝐸 𝑛 𝑓(𝑥) means that the operator E is applied n times on the function 𝑓(𝑥).

𝐸 𝑛 𝑓(𝑥) = 𝐸𝐸 … … . . 𝐸 (𝑛 𝑡𝑖𝑚𝑒𝑠 )𝑓(𝑥)

=𝐸 𝑛−1 {𝐸 𝑓(𝑥)}=𝐸 𝑛−1 {𝐸 𝑓(𝑥 = ℎ)}

=𝐸 𝑛−2 {𝐸 𝑓(𝑥 + ℎ)}=𝐸 𝑛−2 {𝐸 𝑓(𝑥 + 2ℎ)}

=𝐸 𝑛−3 {𝐸 𝑓(𝑥 + 3ℎ)} = ⋯ … … ..=𝐸 𝑓(𝑥 + 𝑛 − 1 . ℎ)

= 𝑓(𝑥 + 𝑛ℎ).
Note: The operator E – is inverse operator of the operator E and is defined as

𝐸 −1 𝑓(𝑥) = 𝑓{𝑥 + (−1)ℎ} = 𝑓(𝑥 − ℎ)

The operator ∆. Let y=𝑓(𝑥) be a function of x. Let the consecutive values of x be a,a+h,a+2h,…………a+nh.
Differing by h .Than corresponding value of y are

𝑓(𝑎). 𝑓(𝑎 + ℎ). 𝑓(𝑎 + 2ℎ) … … … … … … … . . 𝑓(𝑎 + 𝑛ℎ)

The independent variable x is known as argument and the dependent variable is y as known as entry.
Thus we are given a set of values of argument and entry.

The difference 𝑓(𝑎 + ℎ) − 𝑓(𝑎) is called the first forward difference of the function 𝑓(𝑥) at the point
x=a and we denoted it by , ∆ 𝑓(𝑎).

∆ 𝑓(𝑎) = ∆𝑓(𝑎 + ℎ) − ∆𝑓(𝑎)

Again the difference 𝑓(𝑎 + 2ℎ), 𝑓(𝑎 + ℎ) is called the first forward difference of the function --- at the
point x=a+h and is denoted by ∆𝑓(𝑎 + ℎ).

∆𝑓(𝑎 + ℎ) = 𝑓(𝑎 + 2ℎ) − 𝑓(𝑎 + ℎ)

Continuing in a similar manner, we ultimately have

∆𝑓(𝑎 + 𝑛 − 1. ℎ) = 𝑓(𝑎 + 𝑛ℎ) − 𝑓(𝑎 + 𝑛 − 1. ℎ)

The operator ∆ is called the forward or descending difference operator. The difference
∆ 𝑓(𝑎), ∆𝑓(𝑎 + ℎ) etc are called first forward difference.

Moinul Islam Ripon

Roll= 170502

Drive the general form of 2nd order forward difference at point x=a+h

Thus the frist forward difference of 𝑓(𝑥) is defined as,

∆ 𝑓(𝑥) = 𝑓(𝑥 + ℎ) − 𝑓(𝑥)

The difference ∆𝑓(𝑎 + ℎ) − ∆𝑓(𝑎) is known as the scond forward deiierence of 𝑓(𝑥) at the point of
𝑥 = 𝑎 and it denoted by ∆2 𝑓(𝑎) 𝑖. 𝑒

∆2 𝑓(𝑎) = ∆𝑓(𝑎 + ℎ) − ∆𝑓(𝑎)

= {𝑓(𝑎 + 2ℎ) − 𝑓(𝑎 + ℎ)} − {𝑓(𝑎 + ℎ) − 𝑓(𝑎)}

= 𝑓(𝑎 + 2ℎ) − 2𝑓(𝑎 + ℎ) + 𝑓(𝑎)

The difference ∆𝑓(𝑎 + 2ℎ) − ∆𝑓(𝑎 + ℎ) is called the 2nd forward difference of 𝑓(𝑥) at the point of 𝑥 =
𝑎 + ℎ and it denoted by

∆2 𝑓(𝑎 + ℎ)

Thus

= ∆2 𝑓(𝑎 + ℎ) = ∆𝑓(𝑎 + 2ℎ) − ∆𝑓(𝑎 + ℎ)

= {𝑓(𝑎 + 3ℎ) − 𝑓(𝑎 − 2ℎ)} − {𝑓(𝑎 − 2ℎ) − 𝑓(𝑎 − ℎ)}

= 𝑓(𝑎 + 3ℎ) − 2𝑓(2 + 2ℎ) + 𝑓(𝑎 + ℎ)

In general, 2nd forward difference of 𝑓(𝑥) is given by

∆2 𝑓(𝑎) = ∆[∆𝑓(𝑥)] = ∆[𝑓(𝑥 + ℎ) − 𝑓(𝑥)]

= ∆𝑓(𝑥 + ℎ) − 𝑓(𝑥)

= {𝑓(𝑥 + 2ℎ) − 𝑓(𝑥 + ℎ)} − {𝑓(𝑥 + ℎ) − 𝑓(𝑥)

= 𝑓(𝑥 + 2ℎ) − 2𝑓(𝑥 + ℎ) + 𝑓(𝑥).

The difference of the second forward difference are called third forward difference and are denoted by,

∆3 𝑓(𝑎), ∆3 𝑓(𝑎 + ℎ) 𝑒𝑡𝑐

Thus ∆3 𝑓(𝑎) = ∆3 𝑓(𝑎 ) − ∆3 𝑓(𝑎)

= {𝑓(𝑎 + 3ℎ) − 2𝑓(𝑎 + 2ℎ) + 𝑓(𝑎 + ℎ)} − {𝑓(𝑎 + 2ℎ) − 2𝑓(𝑎 + ℎ) + 𝑓(𝑎)}

= 𝑓(𝑎 + 3ℎ) − 3𝑓(𝑎 + 2ℎ) + 3𝑓(𝑎 + ℎ) − 𝑓(𝑎).

In general the nth forward difference of 𝑓(𝑥) is given by

∆𝑛 𝑓(𝑥) = ∆𝑛+1 𝑓(𝑥 + ℎ) − ∆𝑛−1 𝑓(𝑥).

 Md S B Sanaullah Sakib

Roll= 170503

The operator ∇. The difference 𝑓(𝑎 + ℎ) − 𝑓(𝑎) is called the first backward difference of 𝑓(𝑥)
at 𝑥 = 𝑎 + ℎ and is denoted by ∇𝑓(𝑎 + ℎ).

∇𝑓(𝑎 + ℎ) = 𝑓(𝑎 + ℎ) − 𝑓(𝑎)

Similarly ∇𝑓(𝑎 + ℎ) = 𝑓(𝑎 + 2ℎ) − 𝑓(𝑎 + ℎ)

... ... ... ... ... ... ... ... ... ... ... ...

∇𝑓(𝑎 + 𝑛ℎ) = 𝑓(𝑎 + 𝑛ℎ) − 𝑓(𝑎 + 𝑛 − 1ℎ)

The operator ∇ is called the backward or ascending difference operator. The differences of first
backward differences are called second backward differences and denoted by

∇2 𝑓(𝑎 + 2ℎ), ∇2 𝑓(𝑎 + 3ℎ) etc.

Thus ∇2 𝑓(𝑎 + 2ℎ) = ∇𝑓(𝑎 + 2ℎ) − ∇𝑓(𝑎 + ℎ)

= {𝑓(𝑎 + 2ℎ) − ∇𝑓(𝑎 + ℎ)} − {𝑓(𝑎 + ℎ) − 𝑓(𝑎)}

= 𝑓(𝑎 + 2ℎ) − 2𝑓(𝑎 + ℎ) + 𝑓(𝑎)

In general, the first backward difference of 𝑓(𝑥) is defined as,

∇𝑓(𝑥) = 𝑓(𝑥) − 𝑓(𝑥 − ℎ).

The second backward difference of 𝑓(𝑥) is defined as

∇2 𝑓(𝑥) = ∇{∇𝑓(𝑥)} = ∇{𝑓(𝑥) − 𝑓(𝑥 − ℎ)}

= ∇𝑓(𝑥) − ∇𝑓(𝑥 − ℎ)

= {𝑓(𝑥) − 𝑓(𝑥 − ℎ)} − {𝑓(𝑥 − ℎ) − 𝑓(𝑥 − 2ℎ)}

= 𝑓(𝑥) − 2𝑓(𝑥 − ℎ) + 𝑓(𝑥 − 2ℎ).

Similarly the nth backward difference of 𝑓(𝑥) is defined as,

∇𝑛 𝑓(𝑥) = ∇{∇n−1 𝑓(𝑥)} = ∇𝑛−1 𝑓(𝑥) − ∇𝑛−1 𝑓(𝑥 − ℎ).

Relations between the operators.

(a) E≡ 1 + ∆ or ∆≡ 𝐸 − 1

We have ∆𝑓(𝑥) = 𝑓(𝑥 + ℎ) − 𝑓(𝑥)

= 𝐸𝑓(𝑥) − 𝑓(𝑥)
= (𝐸 − 1)𝑓(𝑥).
Thus ∆𝑓(𝑥) = (𝐸 − 1)𝑓(𝑥), for any function 𝑓(𝑥)

∴ ∆≡ 𝐸 − 1

or , 𝐸 ≡ 1+∆

(b) ∇≡ 1 − 𝐸 −1 or 𝐸 −1 ≡ 1 − ∇

We have ∇𝑓(𝑥) = 𝑓(𝑥) − 𝑓(𝑥 − ℎ)

= 𝑓(𝑥) − 𝐸 −1 𝑓(𝑥)

= (1 − 𝐸 −1 )𝑓(𝑥).

Thus ∇𝑓(𝑥) = (1 − 𝐸 −1 )𝑓(𝑥), for any function 𝑓 (𝑥)

∴ ∇≡ 1 − 𝐸 −1 𝑜𝑟 𝐸 −1 ≡ 1 − ∇.

(c) 𝐸∇≡ ∇𝐸 ≡ ∆

We have (𝐸∇)𝑓(𝑥) = 𝐸{∇𝑓(𝑥)}

= 𝐸{𝑓(𝑥) − 𝑓(𝑥 − ℎ)}

= 𝐸𝑓(𝑥) − 𝐸𝑓(𝑥 − ℎ)

= 𝑓(𝑥 + ℎ) − 𝑓(𝑥)

= ∆𝑓(𝑥). ...(1)

Also (∇𝐸)𝑓(𝑥) = ∇{𝐸𝑓(𝑥)} = ∇𝑓(𝑥 + ℎ)

= 𝑓(𝑥 + ℎ) − 𝑓(𝑥)

= ∆𝑓(𝑥) ...(2)

From (1) and (2), we have

𝐸∇≡ ∆ and ∇𝐸 ≡ ∆.
Thus 𝐸∇≡ ∇𝐸 ≡ ∆.
 Ismath Rahaman

Roll= 170505

The Calculus of Finite Differences :

(d) ∆ − ∇≡ ∆∇

We have f(x)= 𝑓(𝑥 + ℎ) − 𝑓(𝑥)

And ∇𝑓(𝑥) = 𝑓(𝑥) − 𝑓(𝑥 − ℎ)

Where h is the interval of differencing.

Now (∆∇)𝑓(𝑥) = ∆{∇𝑓(𝑥)} = ∆{𝑓(𝑥) − 𝑓(𝑥 − ℎ)}

= ∆𝑓(𝑥) − ∆𝑓(𝑥 − ℎ)

= {𝑓(𝑥 + ℎ) − 𝑓(𝑥)} − {𝑓(𝑥) − 𝑓(𝑥 − ℎ)}

= ∆𝑓(𝑥) − ∇𝑓(𝑥)

= (∆ − ∇)𝑓(𝑥)

Thus (∆∇)𝑓(𝑥) = (∆ − ∇)𝑓(𝑥), 𝑓𝑜𝑟 𝑎𝑛𝑦 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑓(𝑥).

∴ ∆∇≡ ∆ − ∇.

(e) (1+∆)(1 − ∇) ≡ 1

We have (1+∆)(1 − ∇)𝑓(𝑥) = (1 + ∆){(1 − ∇)𝑓(𝑥)}

= (1 + ∆)[𝑓(𝑥) − {𝑓(𝑥) − 𝑓(𝑥 − ℎ)}]

= (1 + ∆)𝑓(𝑥 − ℎ)

= 𝐸𝑓(𝑥 − ℎ) [∴ 𝐸 ≡ 1 + ∆]

= 𝑓(𝑥) = 1. 𝑓(𝑥).

Thus (1+∆)(1 − ∇)𝑓(𝑥) = 1. 𝑓(𝑥),for any function f(x).

∴ (1 + ∆)(1 + ∇) ≡ 1.
(f) E≡ 𝑒 ⋌𝐷 ≡ 1 + ∆,where D is the differential operator of differential calculus.

We have Ef(x)= 𝑓(𝑥 + ℎ)

ℎ2
= 𝑓(𝑥) + ℎ𝑓 , (𝑥) + 𝑓 , (x)+…….
2!

[by Taylor’s theorem of differential calculus]

ℎ2
= 1. 𝑓(𝑥) + ℎ 𝐷 𝑓(𝑥) + 𝐷2 𝑓(𝑥) + ⋯
2!

ℎ2 𝐷2
= {1 + ℎ 𝐷 + + ⋯ } 𝑓(𝑥)
2!

= 𝑒 ⋌𝐷 𝑓(𝑥).

Thus E𝑓(𝑥) = 𝑒 ⋌𝐷 𝑓(𝑥),for any function f(x).

∴ 𝐸 ≡ 𝑒 ⋌𝐷 .

Again we know that 𝐸 ≡ 1 + ∆.

Therefore E≡ 𝑒 ⋌𝐷 ≡ 1 + ∆.

Remark. We have 𝑒 ⋌𝐷 ≡ 1 + ∆

Or, ℎ𝐷 ≡ log(1 + ∆)
∆2 ∆3 ∆4
≡ (∆ − + − +⋯)
2 3 4

1 ∆2 ∆3 ∆4
Or D= ℎ(∆ − + − + ⋯)
2 3 4

Nazmus sakib

Roll= 170506

EX.12: construct a forward difference table for the following ing value.

X=0 5 10 15 20 25

𝒇(𝒙) = 7 11 14 18 24 32
Sol:

x y ∆2 𝑦 ∆3 𝑦 ∆4 𝑦 ∆5 𝑦

0 7

5 11 -1

3 2

10 14 1 -1

4 1

15 18 2 -1

6 0

20 24 2

25 32

Ex 14: Obtain the missing terms in the following table:

x 1 2 3 4 5 6 7 8

f(x) 1 8 ? 64 ? 216 343 512

Sol. Here we are given six values, so a polynomial of degree 5 may be fitted which will have its
6th difference as zero ,

i.e. ∆6 𝑓(𝑥) = 0 for all values of x

i.e. (𝐸 − 1)6 𝑓(𝑥) = 0 ∀ x [since ∆≡ 𝐸 − 1]

i.e. (𝐸 6 − 6𝐸 5 + 15𝐸 4 − 20𝐸 3 + 15𝐸 2 − 6𝐸 + 1)

𝑓(𝑥) = 0 ∀ x

i.e. 𝐸 6 𝑓(𝑥) − 6𝐸 5 𝑓(𝑥) + 15𝐸 4 𝑓(𝑥) − 20𝐸 3 𝑓(𝑥)

+15𝐸 2 𝑓(𝑥) − 6𝐸𝑓(𝑥) + 𝑓(𝑥) = 0 ∀ x

i.e. 𝑓(𝑥 + 6) − 6𝑓(𝑥 + 5) + 15𝑓(𝑥 + 4) − 20𝑓(𝑥 + 3)

+15𝑓(𝑥 + 2) − 6𝑓(𝑥 + 1) + 𝑓(𝑥) = 0 ∀ x

________(1)

here interval of differencing is unity.

Putting x=1 and x=2 in (1), we get

𝑓(7) − 6𝑓(6) + 15𝑓(5) − 20𝑓(4) + 15𝑓(3) − 6𝑓(2)

+𝑓(1) = 0 _______ (2)

and

𝑓(8) − 6𝑓(7) + 15𝑓(6) − 20𝑓(5) + 15𝑓(4) − 6𝑓(3)

+𝑓(2) = 0 _______ (3)

Putting the values of f(8),f(7),f(6),f(4),f(2),f(1) in (1) and (2), we get

343 − 6 × 216 + 15𝑓(5) − 20 × 64 + 15𝑓(3) − 6 × 8 + 1 = 0

and

512 − 6 × 343 + 15 × 216 − 20𝑓(5) + 15 × 64

−6𝑓(3) + 8 = 0

i.e. 15𝑓(5) + 15𝑓(3) = 2280

and 20𝑓(5) + 6𝑓(3) = 2662

i.e. 𝑓(5) + 𝑓(3) = 152 _______ (4)

and 10𝑓(5) + 3𝑓(3) = 1331 _______ (5)

Solving (4) and (5), we get, f(3)=27, f(5)=125.

Shasoti Nowaz Angona

Roll= 170507

Ex 15: Obtain the missing terms in the following table:

x 2.0 2.1 2.2 2.3 2.4 2.5 2.6

f(x) 0.135 ? 0.111 0.100 ? 0.082 0.074

Sol. Since we are given five values so ∆5 𝑓(𝑥) = 0 ∀ x

i.e. (𝐸 − 1)5 𝑓(𝑥) = 0 for every value of x

i.e. (𝐸 5 − 5𝐸 4 + 10𝐸 3 − 10𝐸 2 + 5𝐸 − 1)𝑓(𝑥) = 0

i.e. 𝐸 5 𝑓(𝑥) − 5𝐸 4 𝑓(𝑥) + 10𝐸 3 𝑓(𝑥) − 10𝐸 2 𝑓(𝑥)

+5𝐸𝑓(𝑥) − 𝑓(𝑥) = 0 ∀ x

i.e. 𝑓(𝑥 + 5 × .1) − 5𝑓(𝑥 + 4 × .1) + 10𝑓(𝑥 + 3 × .1)

−10𝑓(𝑥 + 2 × .1) + 5𝑓(𝑥 + .1) − 𝑓(𝑥) = 0 ∀ x

here interval of differencing is 0.1

Putting x=2.0 and x=2.1,we get

𝑓(2.5) − 5𝑓(2.4) + 10𝑓(2.3) − 10𝑓(2.2) + 5𝑓(2.1) −𝑓(2.0) = 0

and

𝑓(2.6) − 5𝑓(2.5) + 10𝑓(2.4) − 10𝑓(2.3) + 5𝑓(2.2)

−𝑓(2.1) = 0

Substituting he values of f(2.0),f(2.2),f(2.3),f(2.5),f(2.6) and then solving the equations

simultaneously , we shall have

f(2.1)=0.123 and f(2.4)=0.090

 Jannatul Haider Konok

Roll= 170508

Example 20: Find 𝒇(𝟔) given that 𝒇(𝟎) = −𝟑, 𝒇(𝟏) = 𝟔, 𝒇(𝟐) = 𝟖, 𝒇(𝟑) = 𝟏𝟐, the third
difference being constant:

Solution: We construct the following difference table:

𝒙 𝒇(𝒙) ∆𝒇(𝒙) ∆𝟐 𝒇(𝒙) ∆𝟑 𝒇(𝟎𝒙)

0 -3
9
1 6 -7
2 9
2 8 2
4
3 12

We have
𝑓(6) = 𝑓(0 + 6) = 𝐸 0 𝑓(0) = (1 + ∆)0 𝑓(0)

= (1 + 6∆ + 15∆2 + 20∆3 )𝑓(0)

[higher difference being zero]
= 𝑓(0) + 6∆𝑓(0) + 15∆2 𝑓(0) + 20∆3 𝑓(0)
= −3 + 6 × 9 + 15 × (−7) + 20 × 9
= −3+54-105+180
= 126 [Ans]

Anowara Khatun

Roll= 170509

Example 26: Find the lowest degree polynomial which takes the following values..
 𝒙 0 1 2 3 4 5

𝒇(𝒙) 0 3 8 15 24 35

Solution:
We know that
𝑓(𝑎 + 𝑛ℎ) = 𝑓(𝑎) + 𝑛𝐶1 ∆𝑓(𝑎) + 𝑛𝐶2 ∆2 𝑓(𝑎) + ⋯ + 𝑛𝐶𝑛 ∆𝑛 𝑓(𝑎) … (1)
Putting 𝑎 = 0, ℎ = 1, 𝑛 = 𝑥 we get,
𝑓(𝑥) = 𝑓(0) + 𝑥𝐶1 ∆𝑓(0) + 𝑥𝐶2 ∆2 𝑓(0) + 𝑥𝐶3 ∆3 𝑓(0) + ⋯
𝑥 (2)
= 𝑓(0) + 𝑥 (1) ∆𝑓(0) + ∆2 𝑓(0) + ⋯ (2)
2!
Now we are prepare the difference table for the given data to find
∆𝑓(0), ∆2 𝑓(0), ∆3 𝑓(0) etc.

𝒙 𝒇(𝒙) ∆𝒇(𝒙) ∆𝟐 𝒇(𝒙) ∆𝟑 𝒇(𝟎𝒙)

0 0
3
1 3 2
5 0
2 8 2
7 0
3 15 2
9 0
4 24 2
11
5 35

Putting the values of 𝑓(0), ∆𝑓(0), ∆2 𝑓(0), ∆3 𝑓(0) in (2), we get

𝑥 (2)
𝑓(𝑥) = 0 + 3𝑥 (1) + 2 +0
2!
= 3𝑥 + 𝑥(𝑥 − 1) = 𝑥 2 + 2𝑥. (Ans
Interpolation

The interpolation may be defined as the Technique of obtaining the most likely estimate of a certain
quantity under certain assumption.

The various methods of interpolation are as follows:

i. Graphic method
 ii. Method of curve fitting

iii. use of the calculus of finite difference formula

Nahida Akther

Roll= 170511

Newton-Gregory Formula for Forward Interpolation:

Let 𝑦 = 𝑓(𝑥) be a function which takes the values 𝑓(𝑎), 𝑓(𝑎 + ℎ), 𝑓(𝑎 + 2ℎ) … 𝑓(𝑎 + 𝑛ℎ) , ) for
the (𝑛 + 1) equidistance values 𝑎, 𝑎 + ℎ, 𝑎 + 2ℎ, … , 𝑎 + 𝑛ℎ of the independent variable of 𝑥 and
let 𝑝𝑛 (𝑥) be a polynomial in 𝑥 of degree n.

Let

𝑃𝑛 (𝑥) = 𝐴0 + 𝐴1 (𝑥 − 𝑎) + 𝐴2 (𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ) + 𝐴3 (𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ)

(𝑥 − 𝑎 − 2ℎ)+. .. +𝐴𝑛 (𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ) … (𝑥 − 𝑎 − ̅̅̅̅̅̅̅

𝑛 − 1ℎ) …….. (1)

We choose the coefficient 𝐴0 , 𝐴1 , … 𝐴𝑛 such that,

𝑃𝑛 (𝑎) = 𝑓(𝑎), 𝑃𝑛 (𝑎 + ℎ) = 𝑓(𝑎 + ℎ) … , 𝑃𝑛 (𝑎 + 𝑛ℎ) = 𝑓(𝑎 + 𝑛ℎ).

Putting 𝑥 = 𝑎, 𝑎 + ℎ, … , 𝑎 + 𝑛ℎ in 1 and then also putting the values of 𝑃𝑛 (𝑎), 𝑃𝑛 (𝑎 + ℎ),

𝑃𝑛 (𝑎 + 𝑛ℎ), we get,

𝑓(𝑎) = 𝐴0 ⇒ 𝐴0 = 𝑓(𝑎)

𝑓(𝑎 + ℎ) − 𝑓(𝑎) ∆𝑓(𝑎)

𝑓(𝑎 + ℎ) = 𝐴0 + 𝐴1 ℎ ⇒ 𝐴1 = = ,
ℎ ℎ
𝑓(𝑎 + 2ℎ) = 𝐴0 + 2𝐴1 ℎ + 2ℎ. ℎ𝐴2

𝑓(𝑎 + 2ℎ) − 2[𝑓(𝑎 + ℎ) − 𝑓(𝑎)] − 𝑓(𝑎)

⇒ 𝐴2 =
2ℎ2

𝑓(𝑎 + 2ℎ) − 2𝑓(𝑎 + ℎ) + 𝑓(𝑎)

=
2ℎ2
1
= 2ℎ2 ∆2 𝑓(𝑎)
 1
Similarly 𝐴3 = 3!ℎ2 ∆3 𝑓(𝑎)

… … … … … …
1
𝐴𝑛 = 𝑛!ℎ2 ∆𝑛 𝑓(𝑎)

Putting the values of 𝐴0 , 𝐴1 , … 𝐴𝑛 found above in(1) we get,

∆𝑓(𝑎) 1 1
𝑃𝑛 (𝑥) = 𝑓(𝑎) + (𝑥 − 𝑎) + 2 ∆2 𝑓(𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ) + 2
∆3 (𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ)
ℎ 2ℎ 3! ℎ
1
(𝑥 − 𝑎 − 2ℎ)+. .. + ∆𝑛 (𝑥 − 𝑎)(𝑥 − 𝑎 − ℎ) … (𝑥 − 𝑎 − ̅̅̅̅̅̅̅
𝑛 − 1ℎ)
𝑛!ℎ2

This is known Newton-Gregory Formula for forward interpolation .

𝑥−𝑎
Putting = 𝑢 𝑜𝑟 𝑥 = 𝑎 + ℎ𝑢 the formula takes the form,
ℎ

𝑢(𝑢−1) 𝑢(𝑢−1)(𝑢−2) 3
𝑃𝑛 (𝑥) = 𝑃𝑛 (𝑎 + ℎ𝑢) = 𝑓(𝑎) + 𝑢∆𝑓(𝑎) + ∆2 𝑓(𝑎) + ∆ 𝑓(𝑎) +…
2! 3!
𝑢(𝑢−1)(𝑢−2)…(𝑢−𝑛+1) 𝑛
+ ∆ 𝑓(𝑎)……… (2)
𝑛!

The result can be written as

1 1
𝑃𝑛 (𝑥) = 𝑃𝑛 (𝑎 + ℎ𝑢) = 𝑓(𝑎) + 𝑢(1) ∆𝑓(𝑎) + 2! 𝑢(2) ∆2 𝑓(𝑎) + 3! 𝑢(3) ∆3 𝑓(𝑎) +…
1
+ 𝑢(𝑛) ∆𝑛 𝑓(𝑎)
𝑛!

Here 𝑢(𝑛) = 𝑢(𝑢 − 1)(𝑢 − 2) … (𝑢 − 𝑛 + 1)

This is the form of in which Newton-Gregory Formula for Forward Interpolation is often written.

Arifa Khanom

Roll= 170512

❖ Newton-Gregory Formula for Backward Interpolation.

To drive this formula, we write

𝑃𝑛 (𝑥) = 𝐴0 + 𝐴1 (𝑥 − 𝑎 − 𝑛ℎ) + 𝐴2 (𝑥 − 𝑎 − 𝑛ℎ)(𝑥 − 𝑎 − 𝑛ℎ + ℎ) + 𝐴3 (𝑥 − 𝑎 − 𝑛ℎ)

(𝑥 − 𝑎 − 𝑛ℎ + ℎ)(𝑥 − 𝑎 − 𝑛ℎ + 2ℎ)+. . . +𝐴𝑛 (𝑥 − 𝑎 − 𝑛ℎ)(𝑥 − 𝑎 − 𝑛ℎ + ℎ). . . (𝑥 − 𝑎
− ℎ), . . . . . . . (1)
Where A0 , A1 ,..., An are constants which are to be determined so as to make

𝑝𝑛 (𝑎 + 𝑛ℎ) = 𝑓(𝑎 + 𝑛ℎ), . . . , 𝑝𝑛 (𝑎) = 𝑓(𝑎).

Putting in (1), 𝑥 = 𝑎 + 𝑛ℎ, 𝑥 = 𝑎 + 𝑛ℎ − ℎ, . .. and also putting𝑝𝑛 (𝑎 + 𝑛ℎ) = 𝑓(𝑎 + 𝑛ℎ), . . ., we

get

𝑓(𝑎 + 𝑛ℎ) = 𝐴0 or 𝐴0 = 𝑓(𝑎 + 𝑛ℎ),

𝑓(𝑎+𝑛ℎ)−𝑓(𝑎+𝑛ℎ−ℎ)
𝑓(𝑎 + 𝑛ℎ − 2ℎ) = 𝐴0 + 𝐴1 (−ℎ) or A1 = ℎ

1
= 𝛻𝑓(𝑎 + 𝑛ℎ),
h
𝑓(𝑎 + 𝑛ℎ − 2ℎ) = 𝐴0 + 𝐴1 (−2ℎ) + 𝐴2 (−2ℎ)(−ℎ)
−𝐴0 +2𝐴1 ℎ+𝑓(𝑎+𝑛−2ℎ)
𝑜𝑟, 𝐴2 = 2ℎ2
−𝑓(𝑎 + 𝑛ℎ) + 2[𝑓(𝑎 + 𝑛ℎ) − 𝑓(𝑎 + 𝑛ℎ − ℎ)] + 𝑓(𝑎 + 𝑛 − 2ℎ)
=
2ℎ2
𝑓(𝑎 + 𝑛ℎ) − 2𝑓(𝑎 + 𝑛 − 1ℎ) + 𝑓(𝑎 + 𝑛 − 2ℎ)
=
2ℎ2
1
= 2!ℎ2 𝛻 2 𝑓(𝑎 + 𝑛ℎ).

Similarly,

 3 f (a + nh ),...,
1
A3 = 3
3!h
 n f (a + nh ).
1
A0 = n
n!h

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

𝛻𝑓(𝑎 + 𝑛ℎ)
𝑝𝑛 (𝑥) = 𝑓(𝑎 + 𝑛ℎ) + (𝑥 − 𝑎 + 𝑛ℎ)
ℎ
𝛻 2 (𝑎 + 𝑛ℎ)
+ (𝑥 − 𝑎 + 𝑛ℎ)(𝑥 − 𝑎 + 𝑛ℎ − ℎ) +
2! ℎ2
𝛻 𝑛 𝑓(𝑎 + 𝑛ℎ)
...+ (𝑥 − 𝑎 + 𝑛ℎ)(𝑥 − 𝑎 + 𝑛ℎ − ℎ). . . (𝑥 − 𝑎 − ℎ).
𝑛! ℎ𝑛
This is Newton-Gregory formula for backward interpolation.
 x − (a + nh)
Putting u = or x = a + nh + uh, we get
h
𝑢(𝑢+1)
𝑝𝑛 (𝑥) = 𝑝𝑛 (𝑎 + 𝑛ℎ + ℎ𝑢) = 𝑓(𝑎 + 𝑛ℎ) + 𝑢𝛻𝑓(𝑎 + 𝑛ℎ) + 2! 𝛻 2 𝑓(𝑎 + 𝑛ℎ)
𝑢(𝑢 + 1)(𝑢 + 2) 3 𝑢(𝑢 + 1). . . (𝑢 + 𝑛 − 1) 𝑛
+ 𝛻 𝑓(𝑎 + 𝑛ℎ)+. . . + 𝛻 𝑓(𝑎 + 𝑛ℎ). . . . . . . . (2)
3! 𝑛!

This is the form in which Newton-Gregory formula for backward interpolation is often
written.

Mafia Aktari

Roll= 170513

Example 1: From the following table of yearly premiums for policies maturing at
quinquennial ages, estimate the premiums for policies maturing at the ages of 46 years

Age x : 45 50 55 60 65
Premium f(x) : 2.871 2.404 2.083 1.862 1.712

Sol :The Difference table of the following data is given below

Age x Premium f(x) ∆f(x) ∆2 f(x) ∆3f(x) ∆4 f(x)

45 2.871
-0.467
50 2.404 -0.046
-0.321 0.146
55 2.083 0.017
-0.221
60 1.862 0.1 -0.029
-0.15
65 1.712 0.071
We want , f(x)= f(f(a+uh)
 where a=45
h=5
∴ a+uh=46
45+u*5=46
u=1/5=0.2

Substituting u=1/5 in newton`s formula for forward interpolation, we get

f(46)=f(45)+1/5 ∆ f(45)+1/5(-4/5)1/2! ∆2f(45) +1/5(-4/5)(-9/5)1/3!∆3f(45)

+1/5(-4/5)(-9/5)(-14/5)1/4!∆4f(45)
=2.871+1/5(-0.467)-2/25(.0146)+6/125(-0.046)-
21/625(0.017) =2.763 (Approx.)

Ex-2. From the following table estimate the number of student who obtained marks
between 40 & 45

Marks No of student
30-40 31
40-50 42
50-60 51
60-70 35
70-80 31

Sol. Cumulative frequency table is given below

Marks Above 30 but less than31 No of student x f(x) 40

50 73
60 124
70 159
80 190
Difference table given below
 x f(x) ∆f(x) ∆2 f(x) ∆3f(x) ∆4 f(x)

40 31
42
9
25
50 73

51
60 124 37
-16

70 159 35 12
-4
80 190 31

We shall find: f(45)= Number of student with marks less

then 45.taking a+uh=45, we get
40+u x 10 =45
u=1/2
Substituting u=1/2 in newton`s formula for forward interpolation, we get f(45)=f(40) +1/2 ∆f(40)
+1/2 (-1/2) ∆2f(40) /2!+1/2(-1/2)(-3/2)∆3f(40) /3!+1/2(-1/2)(-3/2)(-5/2)∆4f(40)/4! = 31+1/2
x42 +1/2(-1/2) 1/2! X 9 +1/2(-1/2)(-3/2) 1/3!(-25) +1/2(-1/2)(-3/2)(-5/2) 1/4! X37 = 47.868
( NO Simplification)
Hence the number of student with marks less than 45 is 47.868 i.
e. 48 But the number of students with marks less than 40 is 31
hence the required number of students getting marks between 40 & 45 is =48-31=17

Souvik Biswas Soumma

170514

Example 03: The following table gives the population of a town during the last six censuses.
Estimate using any suitable interpolation formula, the increase in the population during the
period from 1946 to 1948.
Year 1911 1921 1931 1941 1951 1961
Population 12 15 20 27 39 52
(in thousands)
Solution: the difference table for the given data is as follows;

Year Population
𝛁𝒚 𝛁𝟐𝒚 𝛁𝟑 𝒚 𝛁𝟒 𝒚 𝛁𝟓𝒚
X y
1911 12
3
1921 15 2
5 0
1931 20 2 3
7 3 -10
1941 27 -5 -7
12 -4
1951 39 1
13
1961 52
We have to find f (1946) and f (1948)
1946−1911
To find f (1946), u = =3.5
10
Applying Newtons formula for forward interpolation, we get
(3.5)(2.5) (3.5)(2.5)(1.5)(0.5) (3.5)(2.5)(1.5)(0.5)(−.5)
f (1946)=12+3.5×3+ ×2+0+ ×3+ ×(-10)
2! 4! 5!
=12+10.5+8.75+0.8203+0.2734
=32.47
1948−1911
To find f (1948), u= =3.7
10
(3.7)(2.7) (3.7)(2.7)(1.7)(0.7) (3.7)(2.7)(1.7)(0.7)(−0.3)
f (1948)=12+3.7×3+ ×2+0+ ×3+ ×(-10)
2! 4! 5!
=12+11.11+9.99+1.4860+0.2972
=34.8732
therefore, increase in the population during the period from
1946 to 1948 =f(1948)-f(1946)
=34.8732
=2.5295 thousand
=2.53 thousand approximately.

Md. Jonaed

ID: 170515

Example: 4

Given,

sin 45⁰ = 0.7071

sin 50⁰ = 0.7660

sin 55⁰ = 0.8192

sin 60⁰ = 0.8660

Find sin 52⁰ by using any method of interpolation.

Solution:

Here, we have

x 45⁰ 50⁰ 55⁰ 60⁰

f(x) 0.7071 0.7660 0.8192 0.8660

The difference table for the given data is as follows:

x f(x) △ f(x) △2 f(x) △3 f(x)

45⁰ 0.7071
0.0589
50⁰ 0.7660 -0.0057
0.0532 -0.0007
55⁰ 0.8192 -0.0064
0.0468
60⁰ 0.8660

We want 𝑓(52) = 𝑓(𝑎 + 𝑢ℎ), say.

∴ 520 = 𝑎 + 𝑢ℎ

⇒ 520 = 450 + u × 50

7
⇒u=
5
⇒ u = 1.4

By newton’s forward interpolation formula, we get

u(u − 1) 𝟐 u(u − 1)(u − 2) 𝟑

𝑓(𝑎 + 𝑢ℎ) = 𝑓(𝑎) + 𝑢 △ f(a) + △ f(a) + △ f(a)
2! 3!
 (1.4)(0.4) 2 (1.4)(0.4)(−0.6) 3
∴ 𝑓(52°) = 𝑓(45°) + 1.4. ∆𝑓(45°) + ∆ 𝑓(45°) + ∆ 𝑓(45°)
2! 3!
(1.4)(0.4) (1.4)(0.4)(−0.6)
⇒ f(52°) = 0.7071 + 1.4 × 0.0589 + (−0.0057) + × (−0.0007)
2! 3!
⇒ f(52°) = 0.7071 + 0.08246 − 0.001596 + 0.0000392

∴ 𝑓(52°) = 0.788032

Thus sin 52⁰ = 0.7880032 or 0.7880 (approx.).

Name: Cosmic Kapuria

ID: 170516

Example - 9

Find the number of men getting wages between Rs. 10 and Rs. 15 from the following table:

Wages in Rs. Frequency

0 – 10 9

10 – 20 30

20 – 30 35

30 – 40 42

Solution:

At first, we prepare the cumulative frequency table as given below:

Wages less than Rupees No. of men

X 𝒇 (x)

10 9

20 39
 30 74

40 116

The difference table is as given below:

x 𝒇 (x) 𝜟𝒇 (x) 𝜟2 𝒇 (x) 𝜟3 𝒇 (x)

10 9

30

20 39 5

35 2

30 74 7

42

40 116

Now 𝑓 (15) = No. of men getting wages less than Rs.15

= 𝑓 (a + uh), (say)
1
Hence, 15 = a + uh ⇒ 15 = 10 + u ✕ 10 ⇒ u = 2

Hence, By Newton’s forward interpolation formula:

1 1
1 ( −1)
𝑓 (15) = 𝑓 (10) + 𝛥𝑓 ′ (10) + 2 2
𝛥2 𝑓(10)
2 2!

1 1 1
( −1) ( −2)
2 2 2
+ 𝛥3 𝑓(10)
3!

1 1 1 1 1 1 3 1
=9 + 2 ✕ 30 + 2 (− 2) (2!) ✕ 5 + 2 (− 2) (− 2 ) 3! ✕ 2
 =9 + 15 – 0.625 + 0.125 = 23.5

=24 nearly

Therefore, no. of persons getting wages between Rs. 10 and Rs. 15 = 24 – 9 = 15.

Answer: 15

Moumita Yesmin

ID: 170517

Ex.10.Use Newton formula for interpolation to find the net premium at age 25 from the
table given below:

Age Annual net premium

20 0.01427

24 0.01581

28 0.01772

32 0.01996

Sol. The difference table for given data is as follows:

Age x Premium 𝒇(𝒙) ∆𝒇(𝒙) ∆𝟐 𝒇(𝒙) ∆𝟑 𝒇(𝒙)

20 0.01427

0.00154

24 0.01581 0.00037

0.00191 -0.00004

28 0.01772 0.00033

0.00224

32 0.01996
We have to find 𝑓(𝑥) = 𝑓(𝑎 + 𝑢ℎ)

∴ 25 = 𝑎 + 𝑢ℎ

∴ 25 = 20 + 𝑢 × 4

𝑢 = 1.25

By newtons formula we get

𝑢(𝑢 − 1) 2 𝑢(𝑢 − 1)(𝑢 − 2) 3

𝑓(𝑢 + 𝑎ℎ) = 𝑓(𝑎) + 𝑢 △ 𝑓(𝑎) + ∆ 𝑓(𝑎) + ∆ 𝑓(𝑎)
2! 3!
(1.25)(1.25 − 1) 2
∴ 𝑓(25) = 𝑓(20) + (1.25) × ∆𝑓(20) + ∆ 𝑓(20)
2!
(1.25)(1.25 − 1)(1.25 − 2) 3
+ ∆ 𝑓(20)
3!
= 0.01427 + 0.001925 + 0.0000578 + 0.0000016

= 0.0162544 = 0.01625 𝑎𝑝𝑝𝑟𝑜𝑥

Prity Lata Mozumder

Roll: 170518

Ex. 12. Given

x : 1 2 3 4 5 6 7 8

ƒ(x) : 1 8 27 64 125 216 343 512

Find ƒ(7.5)

Sol. The value to be interpolated lies at the end of the given observations i.e. near x=8. So in this case
Newton’s backward formula will be more suitable.

𝑥−(𝑎+𝑛ℎ) 7.5−8
Here u= = =-0.5.
ℎ 1

To calculate backward differences ƒ(a+nh), 2ƒ(a+nh),………,we prepare the following difference table.
2 3
x ƒ(x) ƒ(x) ƒ(x) ƒ(x)

1 1
 2 8 7

3 27 19 12

4 64 37 18 6

5 125 61 24 6

6 216 91 30 6

7 343 127 36 6

8 512 169 42 6

3
Since ƒ(x) is constant ,we can leave higher order differences .

By Newton’s backward interpolation formula,

𝑢(𝑢+1) 2
𝑢(𝑢+1)(𝑢+2) 3
ƒ(a+nh+uh)=ƒ(a+nh)+u ƒ (a+nh) + ƒ(a+nh)+ ƒ(a+nh)
2! 3!

(−0.5)(−0.5+1) (−0.5)(−0.5+1)(−0.5+2)
∴ƒ(7.5)=ƒ(8)+(-0.5) ƒ(8)+ 2 ƒ(8)+ 3 ƒ(8)
2 6

=512-84.5-5.25-.375

=421.875.

Zhumpa Rani Saha

Roll : 170519
Ex. 13: A second degree polynomial passes through (0,1), (1,3), (2,7) and (3,13). Find the
polynomial.

Solution: The difference table is as follows:

𝑥 𝑓(𝑥) Δ 𝑓(𝑥) Δ² 𝑓(𝑥) Δ³ 𝑓(𝑥)

0 1

1 3 2

4 0

2 7 2

3 13

𝑢(𝑢−1)
We know that f(ɑ+uh) = f(ɑ) + uΔf(ɑ) + Δ²f(ɑ),
2!

expanding upto second order differences only.

Put, ɑ = 0, h = 1, u = 𝑥,

so that,
𝑥(𝑥−1)
f( 𝑥) = f(0) + 𝑥Δf(0) + Δ² f(0)
2!

𝑥(𝑥−1)
= 1+ 𝑥×2+ ×2
2!

= 1+ 2 𝑥 + 𝑥²− 𝑥

= 𝑥²+ 𝑥+1.
 Name : Ekra Binte Elais

Roll : 170520

Exercise 5: In an examination the number of candidates who secured marks between

certain limits were as follows:

Marks Number of candidates

0-19 41
20-39 62
40-59 65
60-79 50
80-99 17

Estimate the number of candidates of getting marks less than 70.

Solution:

First we prepare the cumulative frequency table, as given below:

Marks above 0 but less then Number of students

19 41
39 62
59 65
79 50
99 17
Now we prepare the difference table for this date.

Marks N0. of Δf(x) Δ2 f(x) Δ3 f(x)

(x) candidates
F(x)
19 41
39 62 21
59 65 3 -18 0
79 50 -15 -18
99 17 -33 -18
We shall find:

F (70) = Number of candidates with marks less than 70, Taking a+uh=70, we get

19+u×2o=70

U=2.55
Using Newton’s formula for forward interpolation,

We get,

F (70) =f (19) +2.55Δf (19) +2.55(2.55-1)0.5Δ2 f(19)+0

=41+2.55×21+2.55×1.55×0.5× (-18)

=41+53.55-35.5725

=58.9775 (on simplification)

Hence the number of candidates with marks less than 70 is 58.9775,

MD. Parvez Mia

Roll : 170522

Exercise 8:
Estimate u2 from the following table:

X 1 2 3 4 5

U2 7 ? 13 21 37

we have 4 values,

so, ∇4 f(x)=0
 (E-1)4f(x)=0
 (E4-4E3+6E2-4E+1)f(x)=0
 f(x+4)-4f(x+3)+6f(x+2)-4f(x+1)+f(x)=0
 here the interval difference is 1………. (1)
 putting, x=1 in (1)
 f(5)-4f(4)+6f(3)-4f(2)+1=0
 37-4×21+6×13-f(2)+7=0
 f(2)=9.5
 MD. Mahamood Reza

Roll : 170523

Ex:11. The following are the numbers of deaths successive ten year age groups .Find the
number of deaths at 45-50 and 50-55.
Age group:25-35 35-45 45-55 55-60
Deaths :13229 18139 24225 31496

Sol: First we shall form the cumulative frequency table.

Age above 25 and below N0. of deaths
x f(x)
35 13229
45 31368
55 55593
65 87089
The difference table is as under:

X 𝑓(𝑥) ∇𝑓(𝑥) ∇2 𝑓(𝑥) ∇3 𝑓(𝑥)

35 13229

18139
45 31368
6086
24225 1185
55 55593 7271

31496
65 87089

First we shall find 𝑓(50) i.e the no. of deaths above the age 25 and below 50.
𝑥−𝑎 50−35
Here 𝑢= = = 1.5
ℎ 10

By Newton’s formula, we get

(1.5)(1.5 − 1) 2
𝑓(50) = 𝑓(35) + (1.5) ∇𝑓(35) + ∇ 𝑓(35)
2!
(1.5)(1.5 − 1)(1.5 − 2) 3
+ ∇ 𝑓(35)
3!
(1.5)(.5) (1.5)(.5)(−.5)
=13229+(1.5)(1813)+ × 6086 + × (1185)
2 6

=13229+27208.5+2282.25-74.0625
=42646

Hence the reqired number of deaths between 45 and 50

=42616-31368-11278

Therefore the number of deaths between 50 and 55

=24225-11278=12947

Md. Tanvir Hossain

Roll:170524
Exercise 16: Find y when x=8 for

x 0 5 10 15 20 25

y 7 11 14 18 24 32

❖ Solution:
The difference table is as given below:

(x) F(x) Δf(x) Δ2 f(x) Δ3 f(x) Δ4 f(x) Δ5 f(x)

0 7 4

5 11 -1

10 14 3 2

15 18 1 -1

4 1 0
20 24
2 -1
25 32
6 0

8
Here,

f(8) = f (a + uh)

8 = 0+u×5

U=1.6

By Newton’s interpolation formula, we have

F(8) = f (o) + 1.6 Δf (0) + 1.6(1.6-1) 0.5Δ2 f(0) + 1.6(1.6-1) (1.6-2) 0.1667Δ3f(x)
+ 1.6(1.6-1) (1.6-2) (1.6-3) 0.041667 Δ4f(x) + 0

= 7 + (1.6×4) + {0.48×(-1) } – (0.064×2) + {0.0224× (-1) }

= 7+ 6.4 - 0.48 - 0.128 - 0.0224

= 12.7696 ≈ 13

= 13 (approx)

Md. Naimur Rahman Durjoy

Roll:170525

Divided Differences: The differences used in the Newton formula are known as divided
differences which are differences obtained in the usual manner and then divided by certain
differences of the values of the argument.

Let f(x0),f(x1),f(x2),…,f(xn) be the entries corresponding to the arguments x0,x1,x2,… ,xn where the
intervals x1-x0,x2-x1,…,xn-xn-1 may not be equal, i.e. not necessarily equally spaced. Then the first
f(x1 )−f(x0 ) f(x0 )−f(x1 )
divided difference of f(x) for the arguments x0,x1 is defined as or and is denoted
x1 −x0 x0 −x1

by f(x0,x1) or by x⍋ f(x0 ) .
1

Similarly the other first divided differences of f(x) for the arguments x1,x2; x2, x3 ;…, xn-
1,xn are
f(x2 )−f(x1 )
f(x1,x2)= = x⍋ f(x1 )
x2 −x1 2

f(x3 )−f(x2 )
f(x2,x3)= x3 −x2
= x⍋ f(x2 )
3
……………………………………
f(xn )−f(xn−1 )
f(xn-1,xn)= = 𝑥⍋ f(xn−1 )
xn −xn−1 𝑛

The second divided difference of f(x) for the three arguments x0,x1 and x2 is defined as
f(x1 ,x2 )−f(x0 ,x1 ) f(x0 ,x1 )−f(x1 ,x2 ) 2
f(x0,x1,x2)= = =x⍋,x f(x0 )
x2 −x0 x0 −x2 1 2

The nth divided difference is given by

f(x1 ,x2 ,…,xn )−f(x0 ,x1 ,…,xn−1 ) f(x0 ,x1 ,…,xn−1 )−f(x1 ,x2 ,…,xn )
f(x0,x1,x2,...,xn)= =
xn −x0 x0 −xn

⍋n
=x f(x0 )
1 2 ,…,xn
,x

Golam Sarwar

Roll:170526

The nth divided differences of a polynomial of the nth degree are constant:

Proof: First consider the function 𝐹(𝑥) = 𝑥 𝑛 The first divided differences of this function are
given by
𝑓(𝑥𝑟+1 )−𝑓(𝑥𝑟 ) 𝑥𝑟+1 𝑛 −𝑥𝑟 𝑛
𝑓(𝑥𝑟 , 𝑥𝑟+1 ) = =
𝑥𝑟+1 −𝑥𝑟 𝑥𝑟+1 −𝑥𝑟

= 𝑥𝑟+1 𝑛−1 + 𝑥𝑟 𝑥𝑟+1 𝑛−2 + ⋯ + 𝑥𝑟 𝑛−2 𝑥𝑟+1 + 𝑥𝑟 𝑛−1

Which is a homogeneous expression of degree (𝑛 − 1) in 𝑥, 𝑥𝑟+1

The second divided differences are given by

𝑓(𝑥𝑟 , 𝑥𝑟+1 ) − 𝑓(𝑥𝑟+1 , 𝑥𝑟+2 )

𝑓(𝑥𝑟 , 𝑥𝑟+1, 𝑥𝑟+2 ) =
𝑥𝑟+1 − 𝑥𝑟+2

𝑓(𝑥𝑟+1 , 𝑥𝑟+2 ) − 𝑓(𝑥𝑟 , 𝑥𝑟+1 )

=
𝑥𝑟+2 − 𝑥𝑟
 1
= [(𝑥 𝑛−1 + 𝑥𝑟+1 𝑥𝑟+2
𝑛−2 𝑛−2
+ ⋯ + 𝑥𝑟+1 𝑛−1
𝑥𝑟+2 + 𝑥𝑟+1 )
(𝑥𝑟+2 − 𝑥𝑟 ) 𝑟+2
𝑛−1 𝑛−2
− (𝑥𝑟+1 + 𝑥𝑟 𝑥𝑟+1 + ⋯ + 𝑥𝑟𝑛−2 𝑥𝑟+1 + 𝑥𝑟𝑛−1 ))]
𝑛−1 𝑛−2
𝑥𝑟+2 − 𝑥𝑟𝑛−1 𝑥𝑟+2 − 𝑥𝑟𝑛−2 𝑛−2
(𝑥𝑟+2 − 𝑥𝑟 )
= + 𝑥𝑟+1 + ⋯ + 𝑥𝑟+1
𝑥𝑟+2 − 𝑥𝑟 𝑥𝑟+2 − 𝑥𝑟 𝑥𝑟+2 − 𝑥𝑟
𝑛−2 𝑛−3 𝑛−2
= (𝑥𝑟+2 + ⋯ + 𝑥𝑟𝑛−2 ) + 𝑥𝑟+1 (𝑥𝑟+1 + ⋯ + 𝑥𝑟𝑛−3 ) + ⋯ + 𝑥𝑟+1 ,

Which is a homogeneous expression of degree 𝑛 − 2 𝑖𝑛 𝑥𝑟 , 𝑥𝑟+1 𝑎𝑛𝑑 𝑥𝑟+2

By induction it can be shown that 𝑓(𝑥𝑟 , 𝑥𝑟+1 , ⋯ , 𝑥𝑟+𝑚 ) is a homogeneous expression of degree
𝑛 − 𝑚.In particular , the nth divided difference of 𝑓(𝑥) = 𝑥 𝑛 is an expression of degree zero
,i.e., is a constant and is therefore independent of the values 𝑥𝑟 , 𝑥𝑟+1 𝑎𝑛𝑑 𝑥𝑟+2 , ⋯ , +𝑥𝑟+𝑛.

Since the divided differences of 𝑥 𝑛 are constant , therefore The divided differences of 𝑥 𝑛 of
order greater than n will all be zero.

If 𝑓(𝑥) = 𝑎𝑥 𝑛 , where a is constant , then the nth divided differences of 𝑓(𝑥) =

𝑎. (𝑡ℎ𝑒 𝑛𝑡ℎ 𝑑𝑖𝑣𝑖𝑑𝑒𝑑 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑜𝑓 𝑥 𝑛 ), which is a constant .

Hence If 𝑓(𝑥) = 𝑎0 𝑥 𝑛 + 𝑎1 𝑥 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑥1 + 𝑎𝑛 be a polynomial of degree n , then the nth

divided difference of all the terms except 𝑎0 𝑥 𝑛 will be zero and so the nth divided difference of
the whole polynomial will be constant.

Newton's Formula for unequal intervals:

Let 𝑓(𝑥0 ), 𝑓(𝑥1 ), ⋯ , 𝑓(𝑥𝑛 ) be the values of 𝑓(𝑥) corresponding to the arguments 𝑥0 , 𝑥1,⋯ , 𝑥𝑛, not
necessarily equally spaced .

From the definition of divided differences,

𝑓(𝑥) − 𝑓(𝑥0 )
𝑓(𝑥, 𝑥0 ) =
𝑥 − 𝑥0

𝑜𝑟 𝑓(𝑥) = 𝑓(𝑥0 ) + (𝑥 − 𝑥0 )𝑓(𝑥, 𝑥0 )

𝑓(𝑥, 𝑥0 ) − 𝑓(𝑥0 , 𝑥1 )
𝐴𝑙𝑠𝑜 𝑓(𝑥, 𝑥0 , 𝑥1 ) =
𝑥 − 𝑥1

𝑜𝑟 𝑓(𝑥, 𝑥0 ) = 𝑓(𝑥0 , 𝑥1 ) + (𝑥 − 𝑥1 )𝑓(𝑥, 𝑥0 , 𝑥1 ).

Similarly
𝑓(𝑥, 𝑥0 , 𝑥1 ) = 𝑓(𝑥0 , 𝑥1 , 𝑥1 ) + (𝑥 − 𝑥2 )𝑓(𝑥, 𝑥0 , 𝑥1 , 𝑥2 ).

⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯
 ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯ ⋯

𝑓(𝑥, 𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛−1 ) = 𝑓(𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 ) + (𝑥 − 𝑥𝑛 )𝑓(𝑥, 𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 ) ⋯ (4)

Multiplying the equation (2) by (𝑥 − 𝑥0 ), (3) 𝑏𝑦 (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) and so on and finally the
equation (4) by (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) ⋯ (𝑥 − 𝑥𝑛−1 ) and adding to the equation (1) , we have

𝑓(𝑥) = 𝑓(𝑥0 ) + (𝑥 − 𝑥0 )𝑓(𝑥0 , 𝑥1 ) + (𝑥 − 𝑥0 )(𝑥 − 𝑥1 )𝑓(𝑥0 , 𝑥1 , 𝑥2 ) + ⋯

+ (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) ⋯ (𝑥 − 𝑥𝑛−1 )𝑓(𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 ) + 𝑅𝑎

where the reminder 𝑅𝑎 is given by

𝑅𝑎 = (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) ⋯ (𝑥 − 𝑥𝑛 )𝑓(𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 )

Assuming that 𝑓(𝑥) is a polynomial of degree n ,

𝑓(𝑥, 𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 ) vanishes so that

𝑓(𝑥) = 𝑓(𝑥0 ) + (𝑥 − 𝑥0 )𝑓(𝑥0 , 𝑥1 ) + (𝑥 − 𝑥0 )(𝑥 − 𝑥1 )𝑓(𝑥0 , 𝑥1 , 𝑥2 ) + ⋯

+ (𝑥 − 𝑥0 )(𝑥 − 𝑥1 ) ⋯ (𝑥 − 𝑥𝑛−1 )𝑓(𝑥0 , 𝑥1 , ⋯ , 𝑥𝑛 ). ⋯ (5)

This formula is called Newton's divided difference interpolation formula.

Farhad Kabir

Roll:170527

Ex 1: Construct a divided difference table for the following:

X 1 2 4 7 12

F(x) 22 30 82 106 216

Solution: the divided difference table for the given data is as follows:
 x F(x) ∆f(x) ∆2f(x) ∆3f(x) ∆4f(x)

1 22
30−22
=8
2−1
26−8
2 30 =6
4−1
(−3.6)−6
82−30
= 26 = −1.6
7−1
4−2
0.535−(−1.6)
4 82 8−26
= = 0.194
12−1
7−2
−3.6 1.75−(−3.6)
106−82
=8 = 0.535
12−2
7−4
7 106
22−8
=
216−106 12−4
= 22 1.75
12−7
12 216

Ex 2: find the third divided difference with arguments 2,4,8,10 of the function f(x) =x2 -2x.

Solution: the divided difference table is:

x F(x) ∆f(x) ∆2f(x) ∆3f(x)

2 4

56−4
=26
4−2
131−26
= 15
9−2
4 56
23−15
711−56 =1
= 131 10−2
9−4

269−131
= 23
10−4
9 711

980−711
= 269
10−9
 Md. Fahim Ahmed

Roll:170528

Example 03:-By means of Newton’s divided difference formula, find the value of 𝒇(𝟖) and
𝒇(𝟏𝟓) from the following table:

x: 4 5 7 10 11 13

𝒇(𝒙): 48 100 294 900 1210 2028

Sol. The divided difference table is:

x 𝒇(𝒙) Δ 𝒇(𝒙) 𝜟𝟐 𝒇(𝒙) 𝜟𝟑 𝒇(𝒙) 𝜟𝟒 𝒇(𝒙)

4 48

100 − 48
= 52
5−4
5 100

97 − 52
294 − 100 = 15
= 97 7−4
7−5 21 − 15
=1
10 − 4
7 294 0
202 − 97
900 − 294 = 21
= 202 10 − 5
10 − 7 27 − 21
=1
11 − 5
10 900 310 − 202 0
1210 − 900 = 27
= 310 11 − 7
11 − 10 33 − 27
=1
13 − 7
409 − 310
11 1210 = 33
13 − 10
2028 − 1210
= 310
12 − 11
13 2028

Using Newton’s interpolation formula for unequal intervals, we get,

𝑓(8)=48+ (8-4) ×52+ (8-4) (8-5) ×15+ (8-4) (8-5) (8-7) ×1

 =448

And,

𝑓(8)=48+ (15-4) ×52+ (15-4) (15-5) ×15+ (15-4) (15-5) (15-7) ×1

=3150

Exercise 07 : Find the polynomial of lowest possible degree which assumes the values 3,
12, 15,-21 when ‘x’ has the values 3, 2, 1,-1 respectively.

Sol. For the given data the divided deference table is as given as blew:

x 𝑓(𝑥) Δ 𝑓(𝑥) 𝛥2 𝑓(𝑥) 𝛥3 𝑓(𝑥)

-1 -21

15 + 21
= 18
1+1

1 15 −3 − 18
= −7
2+1
−3 + 7
12 − 15 =1
3+1
= −3
2−1
−9 + 3
2 12 = −3
3−1
3 − 12
= −9
3−2

3 3

By Newton’s divided difference formula, we get,

𝑓(𝑥) = 𝑓(𝑥0 )+(x-𝑥0 ) Δ 𝑓(𝑥0 )+ (x-𝑥0 ) (x-𝑥1 )𝛥2 𝑓(𝑥0 )+ (x-𝑥0 ) (x-𝑥1 ) (x-𝑥2 )𝛥3 𝑓(𝑥0 )

Or,
 𝑓(𝑥)=-21+[x-(-1)] (18) +[x-(-1)] (x-1) (-7) +[x-(-1)] (x-1) (x-2) (1)

=-21+(x+1)18+(x+1) (x+1) (x-1) (-7) +(x+1) (x-1) (x-2)

=𝑥 3 -9𝑥 2 +17x+6

Md. Minhaz Abedin

Roll: 170529

Numerical Differentiation:

The process of finding the derivative or derivatives of a function at some value of the
independent variable, when we know the value of the function corresponding to the given values
of the independent variable, is called Numerical Differentiation.

Example - 1
Find the first, second and third derivation of the function tabulated below, at the Point x =1.5
X: 1.5 2.0 2.5 3.0 3.5 4.0
Y=f (x): 3.375 7.000 13.625 24.000 38.875 59.000

Since the derivatives are required at x=1.5, which is near the beginning of the table, therefore we
shall use Newton’s forward formula. The difference table is given blow:

x y ∆y ∆2y ∆3y ∆4y

1.5 3.375
3.625
2.0 7.000 3.000
6.625 0.750
2.5 13.625 3.750 0
10.375 0.750
3.0 24.000 4.500 0
14.875 0.750
3.5 38.875 5.250
20.125
4.0 59.000
Newton’s forward formula is

f (a + xh) = f (a) + xC1∆f (a) + xC2∆2f (a) + xC3∆3f (a), up to third differences
𝑥(𝑥−1) 2 𝑥(𝑥−1)(𝑥−2) 3
= f (a) + x∆ f (a) + ∆ f (a) + ∆ f (a).
2! 3!

Differentiating w.r.t. x thrice and then putting x=0 in the equation obtained, we get
1 1
h f’ (a) = ∆ f (a) - 2 ∆2 f (a) + 3 ∆3 f (a),

h2f’’ (a) = ∆2 f (a) + ∆3 f (a),

h3f”’ (a) = ∆3 f (a).

Putting a= 1.5, h=0.5 and the value of various differences in these equation, we get
1 1 1
f’ (1.5) = 0.5[3.625 - 2 (3.000) + 3 (0.750)] = 4.750

1
f’’ (1.5) = 0.25[ 3.000 + (0.750)] = 9.000

1
f’’’ (1.5) = 0.125[ 0.750] = 6.000

Note. The function tabulated in the given table of this problem is

Y = x3-2x+3

Hence y1 = 3x2-2, Y2= 6x, y3 = 6.

Putting x=1.5, we get

(y1)1.5 = 3(1.5)2-2 = 4.750,

(y2)1.5 = 6(1.5) = 9.000,

(y3)1.5 = 6.000.

Thus we get the same value as obtained by numerical differentiation.

 Kismot Reza

Roll:170531

Ex.2: Find the first and second derivatives of the function tabulated below, at the point of
x=3.0

𝑿 ∶ 𝟑. 𝟎 𝟑. 𝟐 𝟑. 𝟒 𝟑. 𝟔 𝟑. 𝟖 𝟒. 𝟎

Y: -14.000 -10.032 -5.296 0.256 6.672 14.000

Solution: since the derivatives are required at x=3.0, which is near the beginning of the table,
therefore we shall use Newton’s forward formula. The difference table is given below:

x y ∆𝐲 ∆𝟐 𝐲 ∆𝟑 𝐲 ∆𝟒 𝐲

3.0 -14.000

3.968

3.2 -10.032 .768

4.736 .048

3.4 -5.269 .816 0

5.552 .048

3.6 0.256 .864 0

6.416 .048

3.8 6.672 .912

7.328

4.0 14.000
Newton’s forward formula is

𝑥(𝑥−1) 𝑥(𝑥−1)(𝑥−2)
𝑓(𝑎 + 𝑥ℎ) = 𝑓(𝑎) + 𝑥∆𝑓(𝑎) + ∆2 𝑓(𝑎) + ∆3 𝑓(𝑎), Taking up to 3rd difference
2 6
only.

Differentiating w.r.t. x twice and then putting x=0 in the equations obtained, we get
1 1
ℎ𝑓 ′ (𝑎) = ∆𝑓(𝑎) − 2 ∆2 𝑓(𝑎) + 3 ∆3 𝑓(𝑎),

ℎ²𝑓"(𝑎) = ∆²𝑓(𝑎) − ∆3 𝑓(𝑎)

Putting a=3, h=.2 and the values of various differences in these equations, we get

1 1 1
𝑓 ′ (3) = [3.968 − (. 768) + (. 048)] = 18
.2 2 3
1
𝑓"(3) = .04 [. 768 − .048] = 18.

Raihan Babu

Roll:170532

Ex 3: find the first and second derivatives of the function rebuilded below at
the point x=1.1

x: 1 1.2 1.4 1.6 1.8 2.0

f(x) : 0 .1280 .5440 1.2960 2.4320 4.00

Solution: since the derivatives are required at x= 1.1, which is near the beginning
of the table, we shall use the Newton’s forward formula. The difference table is as
below
 x y ∆𝑦 ∆2y ∆3y ∆4y

1.0 0

.1280

1.2 .1280 .2880

.4160 .0480

1.4 .5440 .3360 0

.7520 .0480

1.6 1.2960 .3840 0

1.1360 .0480

1,8 2.4320 .4320

1.15680

2.0 4.00

Dipto Debnath

Roll:170533

Example 04: Find the difference of 𝒇(𝒙) 𝒂𝒕 𝒙 = 𝟒 from the following table:

𝒙 .1 .2 .3 .4
𝒇(𝒙) 1.10517 1.22140 1.34986 1.49182

Solution: Since the derivative is required at x=4, which is near the end of the table, therefore

we shall use newtons Backward formula. The difference table is given below:
 𝒙 𝒚 𝛁𝒚 𝛁𝟐𝒚 𝛁𝟑𝒚
.1 1.10517
.11623
.2 1.22140 .01223
.12846 .00127
.3 1.34986 .01350
.14196
.4 1.49182

Newtons Backward formula is

𝑥(𝑥 + 1) 2
𝑓(𝑎 + 𝑛ℎ + 𝑥ℎ) = 𝑓(𝑎 + 𝑛ℎ + 𝑥∇𝑓(𝑎 + 𝑛ℎ) + ∇ 𝑓(𝑎 + 𝑛ℎ)
2!
𝑥(𝑥+1) (𝑥−2)
+ ∇2 𝑓(𝑎 + 𝑛ℎ),
3!

taking upto third difference only

differentiating w.r.t. x ,we get

2𝑥 + 1 2
ℎ𝑓′(𝑎 + 𝑛ℎ + 𝑥ℎ) = ∇𝑓(𝑎 + 𝑛ℎ) + ∇ 𝑓(𝑎 + 𝑛ℎ)
2
3𝑥 2 +6𝑥+2
+ ∇2 𝑓(𝑎 + 𝑛ℎ),
6

Putting 𝑎 + 𝑛ℎ = −4, ℎ = −1 , 𝑥 = 0, we get

1 1
(−1)𝑓 ′ (−4) = ∇𝑓(−4) + ∇2 𝑓(−4) + ∇2 𝑓(−4)
2 3
1 1
= .14196 + (. 01350) + (. 00127) = 0.14913
2 3
∴ 𝑓 ′ (−4) = 1.4913
 Name: Syed Sadat Raj

Roll:170534
Introduction:

The process of computing the value of a definite integral from a set of numerical values
of the integrand is called Numerical Integration. When applied to the integration of a function of
a single variable, the process is known as quadrature.

The problem of numerical integration, like that of numerical differentiation, is solved by

representing the integrand by an interpolation formula and then integrating this formula between
𝑏
desired limits. Thus, to find the value of the definite integral ∫𝑎 𝑦 𝑑𝑥, we replace the function y
by an interpolation formula, usually one involving differences, and then integrate this formula
between the limits a and b. In this way we can derive quadrature formulas for the approximate
integration of any function for which numerical values are known.

The Trapezoidal Rule:

We know,
𝑛2 𝑛3 𝑛2 ∆2 𝑦0 𝑛4 ∆3 𝑦0
I=h [n𝑦0 + 2 ∆𝑦0 + ( 3 − ) + ( 4 − 𝑛3 +𝑛2 ) +….upto (n+1) terms]
2 2! 3!

… (1)

Putting n=1 in the formula (1) and neglecting second and higher order differences, we get:
𝑥0 +ℎ 1
∫𝑥0 𝑦 𝑑𝑥=h [𝑦0 +2 ∆𝑦0]

𝑦1 −𝑦0 𝑦0 +𝑦1
= h [𝑦0 + ]=h [ ]
2 2

𝑥 +2ℎ 𝑦1 +𝑦2
Similarly, ∫𝑥 0+ℎ 𝑦 𝑑𝑥=h [ ]
0 2

……………………
𝑥0 +𝑛ℎ 𝑦𝑛−1 +𝑦𝑛
̅̅̅̅̅̅ℎ 𝑦 𝑑𝑥 = h [
∫𝑥0+𝑛−1 2
]

Adding these n integrals, we get:

𝑥0 +𝑛ℎ 1
∫𝑥0 𝑦 𝑑𝑥=h [2(𝑦0 + 𝑦𝑛 )+( 𝑦1 + 𝑦2 +…+𝑦𝑛−1 )]
=distance between two consecutive ordinates×{mean of the first and the last coordinates+ sum
of all the intermediate coordinates}.

This rule is known as the Trapezoidal rule.

Note. Here we have assumed that y is a function of x of first degree, i.e., the equation of the
curve is of the form:

y=a+bx.

Simpson’s One-third Rule:

We know,
𝑛2 𝑛3 𝑛2 ∆2 𝑦0 𝑛4 ∆3 𝑦0
I=h [n𝑦0 + ∆𝑦0 + ( − ) +( − 𝑛3 +𝑛2 ) +….upto (n+1) terms]
2 3 2 2! 4 3!

… (1)

Putting n=2 in the formula (1) and neglecting third and higher order differences, we get:
8
𝑥0 +2ℎ ( −2)
∫𝑥0 𝑦 𝑑𝑥=h [2𝑦0 +2∆𝑦0 + 3
∆2 𝑦0]
2

1
=h [2𝑦0 +2(𝑦1 − 𝑦0 )+3(𝑦2 − 2𝑦1 + 𝑦0 )]

ℎ
=3(𝑦0 + 4𝑦1 + 𝑦2 )

Similarly,
0𝑥 +4ℎ ℎ
∫𝑥0+2ℎ 𝑦 𝑑𝑥=3(𝑦2 + 4𝑦3 + 𝑦4 )

…………………………
𝑥0 +𝑛ℎ ℎ
∫𝑥0+(𝑛−2)ℎ 𝑦 𝑑𝑥=3(𝑦𝑛−2 + 4𝑦𝑛−1 + 𝑦𝑛 ), when n is even.

Adding all these integrals, we get:

𝑥0 +𝑛ℎ ℎ
∫𝑥0 𝑦 𝑑𝑥=3 [(𝑦0 + 𝑦𝑛 )+4(𝑦1 + 𝑦3 + ⋯ + 𝑦𝑛−1)+2(𝑦2 + 𝑦4 + ⋯ + 𝑦𝑛−2)

This formula is known as Simpson’s one-third rule.

Note. Here we have neglected all differences above the second, so y must be a polynomial of
second degree only, that is:

y=a𝑥 2 +bx+c.
 M Mostakim Rahman

ID: 170535

#5. Simpson’s Three-eighth’s Rule.

𝑛4
𝑥 +𝑛ℎ 𝑛2 𝑛3 𝑛2 ∆2 𝑦0 ( −𝑛3 +𝑛2 )∆3 𝑦0
4
Putting n=3 in the formula∫𝑥 0 𝑦𝑑𝑥 = ℎ[𝑛𝑦0 + ∆𝑦0 + ( 3 − ) + +
0 2 2 2! 3!
⋯ 𝑢𝑝𝑡𝑜 (𝑛 + 1) 𝑡𝑒𝑟𝑚𝑠] and neglecting all differences above the third, we get

81 3
𝑥0 +3ℎ
9 27 9 ∆2 𝑦0 ( 4 − 27 + 4) ∆ 𝑦0
∫ 𝑦𝑑𝑥 = ℎ[3𝑦0 + ∆𝑦0 + ( − ) + ]
𝑥0 2 3 2 2! 3!

𝑥0 +3ℎ
9 9 3
∫ 𝑦𝑑𝑥 = ℎ [3𝑦0 + (𝑦1 − 𝑦0 ) + (𝑦2 − 2𝑦1 + 𝑦0 ) + (𝑦3 − 3𝑦2 + 3𝑦1 − 𝑦0 )]
𝑥0 2 4 8

𝑥0 +3ℎ
3ℎ
∫ 𝑦𝑑𝑥 = [𝑦 + 3𝑦1 + 3𝑦2 + 𝑦3 ]
𝑥0 8 0

𝑥 +6ℎ 3ℎ
Similarly, ∫𝑥 0+3ℎ 𝑦𝑑𝑥 = [𝑦3 + 3𝑦4 + 3𝑦5 + 𝑦6 ]
0 8

… … … … … … … … … …
𝑥0 +𝑛ℎ
3ℎ
∫ 𝑦𝑑𝑥 = [𝑦 + 3𝑦𝑛−2 + 3𝑦𝑛−1 + 𝑦𝑛 ]
𝑥0 +(𝑛−3)ℎ 8 𝑛−3

Adding all these integrals where n is a multiple of 3, we have

𝑥0 +𝑛ℎ
3ℎ
∫ 𝑦𝑑𝑥 = [(𝑦0 + 𝑦𝑛 ) + 3(𝑦1 + 𝑦2 + 𝑦4 + 𝑦5 + ⋯ + 𝑦𝑛−1 ) + 2(𝑦3 + 𝑦5 + ⋯ + 𝑦𝑛−3 )]
𝑥0 8

This formula is known as Simpson’s three-eighth’s rule.

Note. Here we have neglected all differences above the third so y is a polynomial of the third
degree, i.e., 𝑦 = 𝑎𝑥 3 + 𝑏𝑥 2 + 𝑐𝑥 + 𝑑
 Akash datto

Roll:170536

S6. Weddle's Rule. Putting n=6 in the formula (1) of$2 and neglecting all differences of seventh
and higher order, we get

𝑥0 +6ℎ 123 33
∫𝑥0 𝑦𝑑𝑥 = ℎ[ 6𝑦0 +18∆𝑦0 + 27∆2 𝑦0 + 24∆3 𝑦0 + ∆4 𝑦0 + 10 ∆5 𝑦0 +
10
41
∆6 𝑦0 ]
140

Here the coefficient of ∆6 𝑦0 differs from 3/10 by the small fraction 41/40. Hence if we replace
this coefficient by 3/10, we commit an error of only( h/140)∆6 𝑦0 . If the value of h is such that
the sixth difference are small, the correct committed will be negligible. We therefore change the
last term to (3/10) ∆6 𝑦0 and replace all difference by their values in term of the given y`s.
the result become

𝑥0 +6ℎ 3ℎ
∫𝑥0 𝑦𝑑𝑥 = [ 𝑦0 +5𝑦1 + 𝑦2 + 6𝑦3 + 𝑦4 + 5𝑦5 + 𝑦6 ]
10

Similarly,
0𝑥 +12ℎ 3ℎ
∫𝑥0+6ℎ 𝑦𝑑𝑥 = 10
[ 𝑦6 +5𝑦7 + 𝑦8 + 6𝑦9 + 𝑦10 + 5𝑦11 + 𝑦12 ]

…. .... …. .... …. .... …. ....

0𝑥 +𝑛ℎ 3ℎ
∫𝑥0+(𝑛−6)ℎ 𝑦𝑑𝑥 = 10
[ 𝑦𝑛−6 +5𝑦𝑛−5 + 𝑦𝑛−4 + 6𝑦𝑛−3 + 𝑦𝑛−2 + 5𝑦𝑛−1 + 𝑦𝑛 ]
if n is a multiple of 6.

Adding all these integrals, we have if n is a multiple of 6,

𝑥0 +𝑛ℎ 3ℎ
∫𝑥0 𝑦𝑑𝑥 = [ 𝑦0 +5𝑦1 + 𝑦2 + 6𝑦3 + 𝑦4 + 5𝑦5 + 2𝑦6 + 5𝑦7 + 𝑦8 + ⋯ ]
10

This formula is known as Weddle's Rule. It is more accurate, in general, than Simpson's Rule,
but it requires at least seven consecutive values of the function.

Note. Here we have assumed that the function y is of the form

y=a𝑥 6 + 𝑏𝑥 6 + 𝑐𝑥 6 + 𝑑𝑥 6 + 𝑒𝑥 6 + 𝑓𝑥 + 𝑔
 Tammana Islam

Roll:170537

𝟓.𝟐
Example 01:.𝐂𝐚𝐥𝐜𝐮𝐥𝐚𝐭𝐞 𝐭𝐡𝐞 𝐯𝐚𝐥𝐮𝐞 𝐨𝐟 𝐭𝐡𝐞 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐥 ∫𝟒 𝐥𝐨𝐠𝐱𝐝𝐱 𝐛𝐲

𝟏
(𝐚)𝐓𝐫𝐚𝐩𝐞𝐳𝐨𝐢𝐝𝐚𝐥 𝐫𝐮𝐥𝐞 (𝐛)𝐒𝐢𝐦𝐩𝐨𝐧′ 𝐬 𝐫𝐮𝐥𝐞
𝟑
𝟑
(𝐜)𝐒𝐢𝐦𝐩𝐨𝐧′ 𝐬 𝐫𝐮𝐥𝐞 (𝐝)𝐖𝐞𝐝𝐝𝐥𝐞′ 𝐬 𝐫𝐮𝐥𝐞 .
𝟖
After finding the true value of the integral, compare the errors in the four cases.

sol: Taking h = .2 divide the whole range of integration (4, 5.2) into

six equal parts. The values of logx for each point of sub − division

are given below:

x y = log e x x y = log e x
x0 = 4.0 yo = 1.3862944 x0 + 4h = 4.8 y4 = 1.5686
xo + h = 4.2 y1 = 1.4350845 x0 + 5h = 5.0 y5 = 1.6094
x0 + 2h = 4.4 y2 = 1.4816045 x0 + 6h = 5.2 y6 = 1.6486
xo + 3h = 4.6 y3 = 1.5260563
(a)By Trapezoidal rule, we have
5.2
h
∫ log e xdx = [y + y6 + 2{y1 + y2 + y3 + y4 + y5 }]
4 2 0

.2
= [3.034953 + 2 × 7.6207991]
2
= .1(18.276551)

= 1.8276551

1
(b)By Simpon′ s rule, we have
3
5.2
h
∫ log e xdx = [y0 + y6 + 4(y1 + y3 + y5 ) + 2(y2 + y4 )]
4 3
 .2
= [3.034953 + 4(4.5705787) + 2(3.0502204)]
3
.2
= [3.034953 + 18.282315 + 6.1004408]
3
.2
= × 27.417709
3
= 1.8278472

3
(c)By Simpon′ s rule, we have
8
5.2
3h
∫ log e xdx = [y + y6 + 3(y1 + y2 + y4 + y6 ) + 2y5 ]
4 8 0

3(. 2)
= [3.034953 + 3(6.0947428) + 2(1.5260563)]
8
.6
= [3.0345953 + 18.284228 + 3.0521126]
8
.6
= × 24.371294
8
= 1.827847

(d)By Weddle′ s rule, we have

5.2
3h
∫ log e xdx = [y + y6 + 5(y1 + y5 + y2 + y4 + 6y3 ]
4 10 4

3(. 2)
= [3.034953 + 5(3.044522) + 3.0502204 + 6(1.5260563)]
10
.6
= [3.034953 + 15.22612 + 3.0502204 + 9.1563378]
10
.6
= × 30.464123
10
= 1.8278474
5.2
Actual value of ∫ log e xdx = [x(logx − 1)]5.2
4
4

= [5.2(log5.2 − 1) − 4(log4 − 1)]

 = [3.3730249 − 1.5451774]

= 1.8278475

Hence the errors due to different formulae are

(a). 0001924 (b). 0000003

(c). 0000005 (d). 0000001

We observe that Weddle′ s rule is more accurate.

1
dx 1 3
2. Find ∫ 2
by using Simpon′ s and rule
0 1+x 3 8

Hence obtain the approximate value of π in each case.

Sol: Divide the range of integration (0,1)into 6 equal parts of width

1−0 1 1
= so that h = . The values of f(x)at each point of sub − division
6 6 6
are given below:

x 1
y=
1 + x2
x0 = 0 1
= 1.0000000
1
1 36
x0 + h = = 0.9729729
6 37
2 36
x0 + 2h = = 0.9000000
6 40
3 36
x0 + 3h = = 0.8000000
6 45
4 36
x0 + 4h = = 0.6923076
6 52
5 36
x0 + 5h = = 0.5901639
6 61
x0 + 6h = 1 1
= 0.5000000
2
1
By Simpon′ s rule, we get
3
1 x0+6h
dx dx h
∫ = ∫ = [y + y5 + 4(y1 + y3 + y5 ) + 2(y2 + y6 )]
0 1+x
2
x0 1 + x2 3 0
 1
= [1.5000000 + 4(2.3631359) + 2(1.5923077)]
18
1
= (14.137163)
18
= 0.7853979 … (1)

3
By Simpon′ s rule, we get
8
1
dx 3h
∫ = [y + y6 + 3(y1 + y2 + y4 + y5 ) + 2y3 ]
0 1 + x2 8 0

1
= [1.5000000 + 3(3.1554446) + 2(. 8000000)]
16
1
= (12.566334)
16
= 0.7853958 … (2)
1
dx
But ∫ = [tan−1 x]10
0 1 + x2

= tan−1 1 − tan−1 0
π
= … (3)
4
Now from (1)and (3), we get
π
= 0.7853979 or π = 3.1415916
4
From (2)and (3), we get
π
= 0.7853958 or π = 3.1415835
4
 Tamanna Sarif

Roll:170538

𝟏 𝒅𝒙 𝟏 𝟏
Example 02:- Find ∫𝟎 by using Simpson’s 𝟑 𝒂𝒏𝒅 rule. Hence obtain the approximate
𝟏+𝒙𝟒 𝟖
value of 𝝅 in each case.
1−0 1 1
Sol. Divide the range of integration (0, 1) into 6 equal parts each of width = 6 .so that h=6
6
.The values of 𝑓(𝑥) at each point of sub-division are given blew:

1
x y=1+𝑥 2

𝑥0 = 0 1
= 1.0000000
1
1
𝑥0 + ℎ = 36
6 = 0.9729729
37
2
𝑥0 + 2ℎ = 36
6 = 0.9000000
40
3
𝑥0 + 3ℎ = 36
6 = 0.8000000
45
4
𝑥0 + 4ℎ = 36
6 = 0.6923076
52
5
𝑥0 + 5ℎ = 1
6 = 0.5000000
2
𝑥0 + 6ℎ = 1

1
By Simpson’s ‘3’ rule, we get,

1 𝑑𝑥 𝑥 +6ℎ 𝑑𝑥 ℎ
∫0 =∫𝑥 0 = 3 ⌊𝑦0 + 𝑦6 + 4(𝑦1 + 𝑦3 + 𝑦0 ) + 2(𝑦2 + 𝑦4 )⌋
1+𝑥 4 0 1+𝑥 2

1
= 18 [1.5000000 + 4(2.3631369) + 2(1.5923077)]
 1
= 18 (14.137163)

= .7853979 ............ (1)

3
By Simpson’s ‘8’ rule, we get,

1
𝑑𝑥 3ℎ
∫ = [𝑦 + 𝑦6 + 3(𝑦1 + 𝑦2 + 𝑦4 + 𝑦5 ) + 2𝑦3 ]
0 1 + 𝑥2 8 0

1
= [1.5000000 + 3. (3.1554446) + 2(.8000000)]
16
1
= (12.56334)
16
= 0.7853958 ............ (2)

1 𝑑𝑥 1 𝜋
But ∫0 = [tan−1 𝑥] = tan−1 1 − tan−1 0 = 4 ................... (3)
1+𝑥 2 0
Now from ‘1’ and ‘3’, we get
𝜋
= 0.7853979
4
Or,𝜋 = 3.1415835

From ‘2’ and ‘3’, we get

𝜋
= 0.7853979
4
Or,𝜋 = 3.1415835

Salman Mollha

Roll:170539
𝟏𝟎 𝒅𝒙
Example- 7: Calculate (up to 4 places of decimal )∫𝟐 by dividing the range into eight
𝟏+𝒙
equal parts .

10−2
Solution : Divide the range of integration (2,10) into eight equal parts each of width =1
8
,Hence h=1. The values of y for each point of subdivision are given below:
 x 1
𝑦=
1+𝑥

X0=2 1/3

X0+h=3 1/4

X0+2h=4 1/5

X0+3h=5 1/6

X0+4h=6 1/7

X0+5h=7 1/8

X0+6h=8 1/9

X0+7h=9 1/10

X0+8h=10 1/11

By Simpson’s ‘1/3’ rule , we get

10 𝑑𝑥 ℎ
∫2 = 3 [𝑦0+y5+4(y1+y3+y5+y7) + 2(y2+y6+y8) ]
1+𝑥

1 1 1 1 1 1 1 1 1 1
= [ + 11 + 4( 4 + 6 + 8 + 10)+2(5 + 7 + 9) ]
3 3

1
= 3 [.4242424+4(.6416666)+2(.4539682) ]

1
=3 × 3.8988453

=1.2996151

=1.2996
 Dipta Das

Roll:170540

𝟔 𝒅𝒙
Example 08: Evaluate ∫𝟎 by using (i) Simpson’s ‘1/3’ rule (ii) Simpson’s ‘3/8’ rule (iii)
𝟏+𝒙𝟐
Weddle’s rule.
6−0
Solution: Divide the range of integration (0, 6) into six equal parts each width = 1. Hence
6
h=1. The values of y for each point of sub-division are given below:
1
X Y=1+𝑥 2

Xo= 0 1/1= 1.00000000

Xo+h= 1 1/2= 0.50000000

Xo+2h= 2 1/5= 0.20000000

Xo+3h= 3 1/10= 0.1000000

Xo+4h= 4 1/17= 0.058823529

Xo+5h= 5 1/26= 0.038461538

Xo+6h= 6 1/37= 0.027027027

By Simpson’s ‘1/3’ rule, we get
6 𝑑𝑥 ℎ
∫0 = 3[y0 +y4 +4(y1 +y3 +y5) +2(y2 +y4)]
1+𝑥 2

1
= 3[1.0270270+2.5538462+0.517647]

=1.3661734

By Simpson’s ‘3/8’ rule, we get

6 𝑑𝑥 3ℎ
∫0 = [y0 +y6 +4(y1 +y2 +y4+y5) +2y3]
1+𝑥 2 8

3
= 8 [1.0270270+2.391855+0.200000]

3
=8 × 3.618882=1.3570808

By Weddle’s rule, we get

6 𝑑𝑥 3ℎ
∫0 = 10[y0 +y6 +5 (y1+y5) +y2 +y4 +6y3)]
1+𝑥 2

3
= 10[1.0270270+2.6923075+0.8588235]

=1.3734474b

Sabbir Ahamed

Roll:170541

𝟒
Ex. 9. Evaluate ∫𝟎 𝒆𝒙 𝒅𝒙, by Simpson’s rule, given that 𝒆𝟏 = 𝟐. 𝟕𝟐, 𝒆𝟐 = 𝟕. 𝟑𝟗, 𝒆𝟑 = 𝟐𝟎. 𝟎𝟗,
𝒆𝟒 = 𝟓𝟒. 𝟔𝟎, and compare it with the actual value.

Sol. Divide the whole range (0,4) into 4 equal parts taking h=1.

By Simpson’s 1/3 rule, we get

4 ℎ
∫0 𝑒 𝑥 𝑑𝑥 = 3[y0 + y4 + 4(y1 + y3) + 2y2]

1
= 3 [1 + 54.60 + 4(2.72 + 20.09) + 2 × 7.39]
 1
= 3 × 161.62

= 53.873333

= 53.87

4
The actual value of ∫0 𝑒 𝑥 𝑑𝑥 = [𝑒 𝑥 ] 40 = 𝑒 4 − 𝑒 0

= 54.60 − 1

= 53.60

Akimul Islam
Roll:160503

𝟕 𝒅𝒙
Ex.10: Use Simpsons rule to prove that 𝐥𝐨𝐠 𝒆 𝟕 𝒊𝒔 𝒂𝒑𝒑𝒓𝒐𝒙𝒊𝒎𝒂𝒕𝒆𝒍𝒚 𝟏. 𝟗𝟓𝟖𝟕 𝒖𝒔𝒊𝒏𝒈 ∫𝟏 𝒙

Sol. Divide the range (1,7) into six equal parts each of width 1.hence ℎ = 1.The values of y for
each point of sub division are given below:

X: 𝑥0 = 1 𝑥1 = 2 𝑥2 = 3 𝑥3 = 4 𝑥4 = 5 𝑥5 = 6 𝑥6 = 7
1 1 1 1 1 1
Y: 𝑦0 = 1 𝑦1 = 2 𝑦2 = 3 𝑦3 = 4 𝑦4 = 5 𝑦5 = 6 𝑦6 = 7

1
By Simpsons 3 rule we have

7
𝑑𝑥 ℎ
∫ = [𝑦0 + 𝑦6 + 4(𝑦1 + 𝑦3 + 𝑦6 ) + 2(𝑦2 + 𝑦4 )]
1 𝑥 3

1
= [1.1428571 + 3.6666667 + 1.0666667]
3
1
= × 5.8761905 = 1.9587
3
7 𝑑𝑥
The exact value of ∫1 = [log 𝑥]17 = log 𝑒 7
𝑥

Thus log 𝑒 7 = 1.9587

 Jihad hasan
Roll:160545

•Newton-Raphson Mathod:

Let 𝑥0 denoted an approximate value of the desired root of the equation f(x)=0 and let h be the
correction which must be applied to 𝑥0 to get the exact value of the root. Then 𝑥0 +h is a root of
the equation f(x)=0, so that f(𝑥0 +h)=0………………..(1)

Expanding f(𝑥0 +h) by Taylor theorem, we get

ℎ2
f(𝑥0 +h)=f(𝑥0 )+hf'(𝑥0 )+ f′′(𝑥0 )+……=0
2!

Now if h is sufficiently small. we may neglect the terms containing second & higher power of h
& get simple relation f(𝑥0 )+hf'(𝑥0 )=0
𝑓(𝑥 )
This gives h=𝑓′(𝑥0 ) , provided f'(𝑥0 )≠0.The improved value of the root is
0

𝑓(𝑥 )
𝑥1 =𝑥0 +h=𝑥0 - 𝑓′(𝑥0 )
0

Successive approximation are given by 𝑥2 ,𝑥3 ,……,𝑥𝑛+1

𝑓(𝑥 )
Where 𝑥𝑛+1=𝑥0 - 𝑓′(𝑥0 ) ………….(3)
0

Formula (3) is known as Newton Raphson Formula.

………….The End ………….