DEPARTMENT OF MATHEMATICS

MATHEMATICS FOR ENGINEERS (21MT2102)

I/IV, B. Tech. – II Semester
A.Y.2021 - 2022
COURSE MATERIAL-CO3
Fourier Series
Session-16
Periodic Functions
If the function f(x) satisfies for all values of x the relation 𝑓 (𝑥 + 𝑇) = 𝑓(𝑥),
where T is a real number, the function is said to be periodic if T is the smallest positive
number for which such a relation is satisfied then T is called period of the function. The
graph of such a function is obtained by periodic repetition of its graph in any interval
of length T.
when the above relation is satisfied, we find that
𝑓(𝑥) = 𝑓(𝑥 + 𝑇) = 𝑓 (𝑥 + 2𝑇) = … … … = 𝑓 (𝑥 + 𝑛𝑇) = ⋯.
Examples:
The simplest periodic functions are Sin x and Cos x, having the period 2  . Their reciprocals
𝐶𝑜𝑠𝑒𝑐𝑥 and Sec x are also periodic with period 2  ; Tan x and Cot x are periodic with period
.
The functions 𝑆𝑖𝑛 𝑛𝑥 and 𝐶𝑜𝑠 𝑛𝑥 are periodic with period 2  /n.
2x 2x
The functions Sin and Cos are periodic with period T.
T T
In the subsequent articles we shall use the following results.
a + 2 a + 2
 1   0)
i) a Sin nx dx = − n Cos nx  a =0 (n

a + 2 a + 2
 1   0)
ii) a Cos nx dx = − n Sin nx  a =0 (n

a + 2 a + 2
iii)  Sin nx Cos nx dx =
a
1
2  Sin (m − n) x + Sin(m + n)xdx = 0
a

a + 2 a + 2
=0 , (m  n)
1
iv)  Sin mx Sin nx dx =
a
2  [Cos (m − n) x + Cos (m + n) x] dx
a

a + 2 a + 2
= 0, (m  n)
1
v)  Cos mx Cos nx dx =
a
2  [Cos (m − n) x + Cos (m + n) x] dx
a

1
 a + 2 a+2 x
vi) 2 1
a Sin nx dx = 2  (1 − Cos 2nx) dx =  , (n  0)
a

a + 2 a + 2
vii)  Cos nx dx =
2 1
 (1 + Cos 2nx) dx =  , (n  0)
a
2 a

Euler’s Formulae
It can be shown that under certain conditions a function f(x) can be expressed as an
infinite series of sines and cosines of x and its integral multiples. Thus if f(x) be a function of
x known in interval c<x<  +c we can write

a0*
f (x) = + a1 Cos x+ a2 Cos 2x + ….an Cos nx + b1 Sin x + b2 Sin 2x + ….+ bn
2
Sin nx+ ….…. ….(1)
a0
* to write instead of a0is a conventional device to be able to get more symmetric.
2

i.e., f(x) = a0 +  (a n Cos nx + bn Sin x)
2 n =1

c + 2
1
where a0 =
 
c
f ( x) dx …(2)

c + 2
1
an =
  c
f ( x) Cos nx dx ….(3)

c + 2
1
bn =
  c
f ( x) Sin nx dx …(4)

These values of a0, an, bnare known as Euler’s formulae and the series (1) as Fourier
series for f(x).
Example 1. Find the Fourier series to represent the function, f(x) given by
f(x)= -x, −   x  0,
f(x)= x, 0<x< 
a0 
Taking f(x)= +  a n Cosnx + bn Sinnx
2 n =1
We have
 0 
1 1 1
a0 =
  f ( x)dx =   (− x)dx +   x dx
− − 0

 
= + =
2 2

2
  0 
1 1 1
an =
 −
 f ( x)Cos nx dx =   (− x)Cox nx dx +   x Cos nx dx
− 0

0 
1  xSin x Cosnx 
+ 
1 xSin nx Cos nx 
= −  +  +
 n n  −   n
2
n 2  0

1 Cos n Cos n 1
= − + + −
n 2
n 2
n 2
 n2
2
= + (Cos n − 1).
 n2
 0 
1 1 1
bn =
 −
 f ( x) Sin nx dx =   (− x)Sinnx dx +   xSin nx dx
− 0


1  − Cos nx Sinnx 
0
+ 
1 xCos nx Sin nx 
=− + +
 n 
n  −  
2
n n 2  0

Cos n Cos n
= − =0
n n
Consequently, we get the series
 4 1 1 
F(x)= − Cos x + 2 Cos 3x + 2 Cos 5 x + ....
2  3 5 
Observation. Putting x=0 in (ii), we find another interesting series involving
 4 1 1 1 
0= − 1 + 2 + 2 + 2 + ....
2   3 5 7 

2 1 1 1
i..e, =1+ + + + .....
8 32 5 2 7 2

Example 2. Obtain the Fourier Series for ex in the interval -  <x< 

Taking f(x)= ex
a0 
= +  (a n cCos nx +b n Sin nx) ………….(1)
2 n =1
We have
 
1 1
a0 =
2 
−
f ( x)dx =
2 − 
e x dx

1 
= (e − e − )
2
 
1 1
an=  f ( x)Cos nx dx =  e
x
Cos nx dx
 −
 −

3
=
 (n
1
2
+ 1)
e Cosnx + ne
x −
Sin x 
−

=
 (n
1
2
+ 1)
e Cosnx − ne Cos nx
x x

 
1 1
bn =  f ( x) Sin nx dx =  e
x
Sin nx dx
 −
 −

=
 (n
1
2
+ 1)
e Sinnx − ne Cos nx 
x x 
−

=
−n
 (n + 1)
2

e  Cosnx − ne − Cos nx 
− n Cos nx 
= ( e − e − )
 (n + 1)
2

Subsequently, we get the series

e  − e −  1  1 1 1  1 2 3 
ex =  −  Cosx − Cos 2 x + Cos 3x.... +  Sin x − Sin 2 x + Sin3x.....
 2  2 5 10  2 5 10 
Class Room Delivery Problems
1. Find the Fourier series to represent the function, f(x)=x-x2 in the interval –π<x<π
2. Constuct the periodic function for the graph in the interval from 𝑥 = 0 to 𝑥 = 2𝜋

and then express it as Fourier series.

 x, for 0  x  
3. Let f(x) be a function of period 2π such that f ( x) =  Write Fourier series
 for   x  2
of f(x).
4. Evaluate the Fourier series of f(x) be a function of period 2l such that f(x) = x+2 in the range
− 1 < x < 1.
Practise Problems:
1. Let f(x) be a function of period 2π such that f(x) = x, 0 < x < π
= π, π < x < 2π .

4
 Write Fourier series of f(x).
2. Find the Fourier series of f(x) be a function of period 2π such that
f(x) = π – x, 0 < x < π
= 0, π < x < 2π
Session-17
Even and Odd Functions
(i) A function f(x) is said to be odd, if f(x-)= -f(x).
e.g., Sin x Tan x, x2 all odd functions. Graphically, an odd function is symmetrical about the
origin.
(ii) A function f(x) is said to be even, if f(-x) = f(x).
e.g., Cos x, Sin x2, x2 all even functions. Graphically, an even function is symmetrical about
the y-axis.
We shall be using the following property of definite integrals in the next paragraph.
c

 f ( x)dx = 0, when f ( x) is an odd

−c
function, .

c
= 2  f ( x)dx, when f ( x) is an even function, .
0

Expansion of Odd or Even Functions

We know that a periodic function f(x) defined in (− ,  ) can be represented by the Fourier
series

nx nx
a
a0
f ( x) = + n cos + bn sin
2 n =1 c c
c
1
where a0 =
c −cf ( x) dx

nx
c
1
an = 
c −c
f ( x) Cos
c
dx

nx
c
1
bn = 
c −c
f ( x) Sin
c
dx

Case I. When f(x) is an odd function.

c
1
a0 =
c −cf ( x) dx = 0

5
Since Cos nx is an even function, therefore f(x) Cos nx is an odd function.
c c

nx
c
1
 a n =  f ( x) Cos dx =0
c −c c

Again, Since Sin nx is an odd function, therefore f(x) .Sin nx is an even function.
c c

Sin nx dx
c
1
 bn =
c −c
 f ( x) c

nx
c
2
=
c  f ( x) Sin
0
c
dx

Hence, if a periodic function f(x) is odd, its Fourier expansion contains only Sin e terms and

Sin nx dx .
c
2
bn =
c  f ( x)
0
c

Case II. When f(x) is an even function.

c c
1 2
a0 = 
c −c
f ( x) dx =  f ( x) dx
c0

Since f(x) Cos nx is also an even function

c

nx
c
1
 an = 
c −c
f ( x) Cos
c
dx .

nx
c
2
=
c  f ( x) Cos
0
c
dx

nx
Again, Since f(x) Sin is an odd function
c

Sin nx dx =0
c
1
 bn =
c −c
 f ( x) c

Thus if a periodic function f(x) is even, its Fourier expansion contains only cosine terms,
and
c
2
a0 =
c  f ( x) dx
0

nx
c
2
an =
c  f ( x)
0
Cos
c
dx.

Example 1.
Find a Fourier series to represent x2 in the interval (-l, l)

6
 Since f(x) = x2, is an even function in (-l l)

nx
 f(x) = a 0 +
2
 an Cos
n =1 l
..(i)

l
2 2l 2
 x dx =
2
Then a0 =
l 0
3

Cos nx dx
l
2
x
2
an =
l 0
l
l
 2  Sin nx / l   Cos nx / l   Sin nx / l 
= 2  x  n / l  − 2 x  − n 2 2 / l 2  + 2 − n 3 3 / l 3 
l        0

4l 2
= Cos n  .
n 2 2
− 4l 2 4l 2 − 4l 2 − 4l 2
 a1 = , a2 = 2 2 , a 3 = 2 2 , a 4 = 2 2 , etc.
2 2 3 4
Substituting these values in (i), we get

l 2 4l 2  Cos x / l Cos 2x / l Cos 3x / l 

2
x = −  − + 
3  2  12 22 32 
the required series.
Example 2:
Express x as a Fourier series from − to 
The function is odd in the given interval.
Therefore, a0 = an = 0

also bn = 2  f (x) Sin nx dx
 0


2
=
 x
0
Sin nx dx


2  x − Cos nx  −  − Sin nx 
=     
      0
2
n n

2(−1) n
bn = -
n
Hence for −  x  
x=2 (Sin x -
1
2 Sin 2 x + 13 Sin 3x − 14 Sin 4 x + ...........) .

7
Half Range Series
In some problems it is desired to expand a function in a Fourier series of a function f(x)
for the interval (0, c) which is half the period of the Fourier series. In addition, we may be
forced by the conditions of the problem to expand a given function in a series of sines alone or
a series of cosines only. As it is immaterial whatever the function may be outside the range
0<x<c, we extent the function to cover the range –c<x<c so that the new function may be odd
or even. The Fourier expansion of such a function of half the period, therefore, consists of Sine
or cosine terms only.
Sine series: If it be required to expand f(x) as series of since in 0<x<c, then we extend the
function reflecting it in the origin, so that
F(-x)=-F(x)
Then extended function is odd in (-c,c) and its expansion will give the Fourier series :

nx
f ( x) =  bn Sin
n =1 c
where
nx
c
2
bn =
c0 f ( x)Sin
c
dx.

Cosine series. If it be required to express f(x) as a cosine series in 0<x<c, we extend the
function reflecting it in the y-axis, so that
f(-x) = f(x)
Then the extended function is even in (-c, c) and its expansion will give fourier cosine series:

nx
f(x) = a 0 +
2
a
n =1
n Cos
c

where
c
2
a0 =
c  f ( x) dx
0

nx
c
2
and an =
c  f ( x)
0
Cos
c
dx.

Problem 1.
Obtain the Fourier half-range Cosine series for the function f(x)=x,0<x<  .

1
Here an =
  f ( x) dx
0

8
 
1 1

= xdx = 
2
0


2
=
  x Cos nx dx
0


2  Sin nx  Cos nx 
= x − − 
 n  n 2  0
−4
= 0(n − even, (n − odd )
n 2
Hence the Fourier Cosine series is

−
4
 Cosx + Cos23x + Cos25 x + .... 
2  3 5
Problem 2.

Expand f(x)= 1 − x, if 0  x  1
4 2

= x − 3 , if 1  x  1
4 2
in the Fourier series of Sine terms
Let f(x) represent an odd function in (-1,1) so that

f ( x) =  bn Sinnx
n =1

1
where bn = 2  f ( x) Sin nx dx
0

 12 
 1 
1
3
= 2  ( − x)Sinnx dx +  ( x − ) Sinnx dx 
4 4
0 1

 2 
1
1
 1 Cos nx Sin nx  2   3  Cosn x Sinnx 
= 2 ( − − x ) − 2 2  + 2 −  x −  + 2 2 
 4 n n  0   4  n n  1
2

1 4Sin(n / 2)
bn = (1 − Cosn  )
2n n 2 2
Thus
1 4   1 4   1 4 
f ( x) =  − 2  Sinx +  + 2 2  Sin3x +  − 2 2  Sin5x + .........
    3 3    5 5  

9
Problem: 3
 
3 0x
3
  2
Expand f(x)=  0 x as Fourier cosine and sine series
 3 3
− 2
x
 3 3
(i) For Cosine Series

2
a0=
  f ( x)dx
0

 3 2


2  3
 
=   dx +  0 dx +  − dx 
 03 3
  2

 3 3 

2   2  2 2 2 
= − + =0
  9 3 9 

2
an =
  f ( x)Cos nx dx
0

 3 2


2  3
 
=   Cos nx dx +  0.Cos nx dx +  − Cos nx dx 
 03 3
  2

 3 3 

  
2   Sinnx  3  Sinnx  
= .   − 
 3  n  0  n  2 
 3 
2 1 n 1 2 n 
=  Sin + Sin
3 n 3 n 3 
2 n  n 
= Sin 1 + 2 Cos 3 
3n 3  
Thus
2  
a1 = Sin 1 + 2 Cos 
3 3 3

2 3 2
= . .(1 + 1) =
3 2 3

10
 2 2  2 
a2 = Sin 1 + 2Cos 3  = 0
3.2 3  
2  3 
a3 = Sin 1 + 2Cos  = 0
3.3  3
2 4  4 
a4 = Sin 1 + 2Cos =0
3.4 3  3 
2 5  5 
a5 = Sin 1 + 2Cos
3.5 3  3 

2 3.2 2
=− . =− .etc.
3.5 2 5 3
2  1 1 
 f ( x) = Cosx − Cos 5 x + Cos 7 x.......
3 5 7 
(ii) For Sine Series

2
bn =
  f ( x) Sin nx dx
0

 2


2 3  3
 
=   Sin nx dx +  0.Sinnx dx +  − Sin nx dx 
 03 3
  2

 3 3 

  
2   Cosnx  3  Cos nx  
= . −  −  − 
 3  n 0  n  2 
 3 

2 1 n 1 1 1 2n 
= −  Cos − − Cos nx + Cos
3 n 3 n n n 3 
2 1 n 2n 
= −  Cos − 1 − Cos n + Cos
3 n 3 3 
2   2 
Thus b1 = − Cos − 1 − Cos  + Cos  = 0
3.1  3 3
2  2 4 
b2 = − Cos − 1 − Cos 2 + Cos =1
3.2  3 3 

2  3 6 
b3 = − Cos − 1 − Cos 3 + Cos =0
3.3  3 3 

11
 2  4 8  1
b4 = − Cos − 1 − Cos 4 + Cos  = etc.
3.4  3 3 2

Hence f ( x) = Sin2 x + 1 Sin4 x + 1 Sin10 x + .........

2 10
Class Room Delivery Problems
1. Expand the Fourier cosine for f(x) = x, on the interval [0, π].
2. Obtain the Fourier sine series for f(x) = 4-x2 on [0, 2].
3. Obtain the Fourier cosine series for f(x) = 4-x on [0, 1].
4. Expand the Fourier cosine for f(x) = x, on the interval [0, 2].
Practise Problems
1) Write the Fourier series of f(x) = x2 , over the interval −π < x < π and has period 2π,
Sketch a graph of f(x) in the interval −π< x <π
2) Compute the Fourier series for the function f(x) =|x| in– π x  π.
 
3 0x
3
  2
3) Expand f(x)=  0 x as Fourier cosine and sine series
 3 3
− 2
x
 3 3
4) Expand the Fourier sine for f(x) = x2, on the interval [0, π ].
5) Expand the Fourier Series to represent the function f(t) = t 2 , -3 <t< 3.
6) Obtain the Fourier Series to represent the function f(t)=1, - π ≤ t ≤ π.
Session 18
Partial differential equations
We will now consider differential equations that model change where there is more than one
independent variable. For example, the temperature in an object change with time and with
the position within the object. The rates of change lead to partial derivatives, and the
equations relating them are called partial differential equations. The applications of the
subject are many, and the types of equations that arise have a great deal of variety. We will
limit our study to the equations that arise most frequently in applications. These model heat
flow and simple waves. solving them in some cases using the method of separation of
variables.
Partial differential equation is an equation relating to the partial derivatives of some unknown
functions.
Goals: 1. Formulate PDE, 2. Solve PDE

12
A partial differential equation (PDE) is an equation involving one or more partial derivatives
of an unknown function that depends on two or more variables, often time t and one or several
variables in space
z  2 z
Ex: = 2
t x
The order of the highest derivative in PDE is called as the order of the PDE.
z  2 z
Ex: = 2 second order PDE
t x
Standard notation: Let x and y represent the independent variables and z the dependent
variable so that 𝑧 = 𝑓 (𝑥, 𝑦) then
z z
= z x = p, = z y = q, and
x y
2z 2z 2z
= z xx = r , = z xy = s, = z yy = t.
x 2 xy y 2
These are the first , second order parial derivative s of z = f(x, y) respective ly.
Just like as in the case of ordinary differential equations, second order PDE’s will be the most
important ones in applications.
A Partial differential equation can be obtained by elimination of arbitrary constants or by
elimination of arbitrary functions, involving two or more variables.
Here we are discussing about how to obtain partial differential equations by eliminating
arbitrary constants and functions from the solution.
Elimination of arbitrary constants:
Consider z to be a function of two independent variables x and y defined by
𝑓(𝑥, 𝑦, 𝑧, 𝑎, 𝑏) = 0 ---- (1) in which a and b are arbitrary constants.
To Eliminate two constants we need at least three equations .Therefore partially differentiating
above equation w.r.t x and y we get two or more equations. From these three equations we can
eliminate the two constants ‘a’ and ‘b’
Differentiating equation (1) partially with respect to x and y, we obtain
𝜕𝑓 𝜕𝑓 𝜕𝑧 𝜕𝑓 𝜕𝑓
+ = +𝑝 =0 − −(2)
𝜕𝑥 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑧
𝜕𝑓 𝜕𝑓 𝜕𝑧 𝜕𝑓 𝜕𝑓
+ = +𝑞 =0 − −(3)
𝜕𝑦 𝜕𝑧 𝜕𝑦 𝜕𝑦 𝜕𝑧
By means of above three equations, two arbitrary constants can be eliminated.

13
So that we obtained a result i.e., a required Partial differential equation of order one in the form
𝐹(𝑥, 𝑦, 𝑧, 𝑝, 𝑞) = 0
Elimination of arbitrary functions:
Let 𝑢 = 𝑢(𝑥, 𝑦, 𝑧) 𝑎𝑛𝑑 𝑣 = 𝑣(𝑥, 𝑦, 𝑧)𝑏𝑒 𝑖𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 functions of the variables x, y ,z and
let ∅(𝑢, 𝑣) = 0 − (1) be an arbitrary relation between them . We shall obtain the Partial
Differential Equation by eliminating the functions 𝑢 and 𝑣. Regarding z as the dependent
variable and differentiating (1) partially with respect to x and y, then we get
  u u z    v v z 
 + +  + =0
u  x z x  v  x z x 
  u u    v v 
 + p +  + p  =0
u  x z  v  x z 
and
  u u    v v 
 + q  +  + q  = 0
u  y z  v  y z 
e lim inating
 
, from (2)and (3) wehave
u v

u u v v
+p +p
| x z x z |
| u u v v |=0
+q +q
y z y z

 u u   v v   u u   v v 
i.e.  + p   + q  -  + q   + p  =0
 x z   y z   y z   x z 

this is the required differential equation.

𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
Writing 𝑃 = 𝜕𝑦 𝜕𝑧 − 𝜕𝑧 𝜕𝑦
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
𝑄= −
𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑧
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
𝑅 = 𝜕𝑥 𝜕𝑦 − 𝜕𝑦 𝜕𝑥

Then the above equation takes of the form 𝑃𝑝 + 𝑄𝑞 = 𝑅 a partial differential in p and q free
of the arbitrary function ∅(𝑢, 𝑣).Thus from this equation which involves arbitrary function ∅
we obtain a partial differential equation which is linear.
Example 1. Formulate the PDEs from the relation 2z=x2/a2+y2/b2 by eliminating the arbitrary
constants a, b.
Solution Given that 2z = x2/a2 + y2/b2 (1)

14
Differentiate equation (1) partially w.r.to x and y to obtain
1 1 z 1 1 z
= ; =
a 2 x x b 2 y y
Substituting these values in equation (1), we get the required PDE 2z=px+qy.
Example 2. Formulate the partial differential equation of all spheres whose centre lie on Z-
axis and given by equation x2 + y2 + (z-a)2 = b2, a and b being constants.
Solution. Equation of the sphere given by x2 + y2 + (z-a)2 = b2 (1)
Where a and b are two arbitrary constants.
Differentiating (1) partially w. r. t ‘x’ and ‘y’
x + (z-a) p=0 (2)
y + (z-a) q=0 (3)
Eliminating ‘a’ from equations (2) & (3), we get
yp - xq=0
This is the required partial differential equation.
Example 3. Formulate the partial differential equation from z=f (x2+y2).
Solution. Given z=f(x2+y2) (1)
Differentiating (1) partially w. r. t ‘x’ and ‘y’
z
p= = f ( x 2 + y 2 )2 x (2)
x
z
q= = f ( x 2 + y 2 )2 y (3)
y
p x
Dividing (2) by (3), we get =  py − qx = 0 which is required partial differential
q y
equation.
Class Room Delivery Problems
1. Obtain a Partial Differential Equation from the equation 𝑧 = 𝑎𝑥 + 𝑏𝑦
2. Obtain a Partial Differential Equation from the equation 𝑧 = 𝑥 𝑦 + (𝑥 2 + 𝑦 2 )
3. Obtain a Partial Differential Equation from the equation 𝑧 = 𝑓(𝑥 − 𝑎𝑡) + 𝐹(𝑥 + 𝑎𝑡)
4. Obtain the Partial Differential Equation of all spheres having radius ‘r’ and centered on
𝑥𝑦-plane.
Practice Problems
1. Eliminate the arbitrary constants from z=f(y/x) to form a PDE
2. Eliminate the arbitrary constants from z=f(x2-y2) to form a PDE

15
 3. Formulate the partial differential equation of all spheres whose Centre lie on Z-axis and
given by Equation, 𝑥 2 + 𝑦 2 + (𝑧 − 𝑎)2 = 𝑏 2 a and b being constants.
4. Formulate the partial differential equation of the following by eliminating the arbitrary
Functions from z = 𝑓(𝑥 2 + 𝑦 2 + 𝑧 2 )
𝑥2 𝑦2
5. Formulate the Partial Differential Equation from the relation 2𝑧 = 𝑎2 + 𝑏2 by

eliminating the arbitrary constants a, b.

Session 19
Solving the linear partial differential equations through Lagrange’s method
Linear partial Differential Equations of First Order
Lagrange’s Linear Equation:
A linear partial differential equation of the first order, commonly known as Lagrange’s linear
equation is of the form
Pp + Qq = R
Where P,Q and R are functions of x,y,z
u u
And p = and .
x y
Method of solving
To solve the equation Pp+Qq = R
(i). form the auxiliary equations
dx/P =dy/Q= dz/R
(ii). Solve the auxiliary equations obtaining two independent solutions u =a and v =b
(iii). Then the solution is (u,v)=0 or u = F(v) or v=F(u).
Solution of the subsidiary equations
The subsidiary equations are
dx/P =dy/Q= dz/R
I - Method of Grouping:
The equations are dx/P =dy/Q= dz/R by taking first two members dx/P =dy/Q and then
Integrating we get an equation, say u=a which gives one equation of the solution similarly,
taking any other two members, and then integrating, we get another equation, say v= b which
gives another equation of the solution.
II Method of Multiplies
If l,m,n are functions of x,y,z are constants, then

16
 ldx + mdy + ndz
dx/P =dy/Q= dz/R =
lP + mQ + nR
Now if the multiplies l,m,n are chosen in such a way that lP+mQ+nR = 0 then ldx+mdy+ndz
=0. Integrating we get u =a. This gives one equation of the solution.
Similarly, taking multiplies l1, m1, n1 in such a way that l1P+m1Q+n1R = 0 then l1dx+m1dy+n1dz
=0. Integrating v =b. this gives another equation of solution.
The two solution u=a, v=b so obtained form the complete solution.
Note: We may get one solution u=a from the method of grouping and another solution v=b
from the method of multiplies.
Example 1
1). Solve x(y-z)p+y(z-x)q=z(x-y).
Sol:- The Auxiliary equations are
dx dy dz
= =
x( y − z ) y ( z − x) z ( x − y )
Taking 1,1,1 as multiplies
dx + dy + dz dx + dy + dz
Each fraction = =
x( y − z ) + y ( z − x) + z ( x − y ) 0
This gives dx+dy+dz = 0
 x+y+z =a (u=a)
Again taking 1/x, 1/y, 1/z as multiplies
1 1 1
dx + dy + dz
x y z
Each fraction =
x( y − z ) + y ( z − x) + z ( x − y )
1 1 1
dx + dy + dz
x y z
=
0
1 1 1
This gives dx + dy + dz =0
x y z
log x+log y+log z = log b
 xyz =b (v=b)
The general integral is
(x+y+z, xyz) =0.

17
Example 2
Solve y2zp+x2zq = xy2
Sol: Auxiliary equations are
dx dy dz
2
= 2 = 2
y z x z xy
dx dy
From 2
= 2
y z x z

 x2dx = y2dy
 x3/3 = y3/3+a
 x3-y3 = a1 where a1= 3a (u=a)
dx dz
Again 2
= 2  xdx = zdz
y z xy

x2/2 = z2/2+b
 x2-z2 = b1 where b1 = 2b (v=b1)
The general integral is
(x3-y3, x2-z2) =0.
Example 3
Solve (x2-y2-z2)p+2xyq = 2xz
Sol: Auxiliary equations are
dx dy dz
= =
x −y −z
2 2 2
2 xy 2 xz

dy dz
from =
2 xy 2 xz
dy dz
=
y z

 log y = log z+log a

 y/z =a (u=a)

Again using x,y,z as the multieplies

dx dy dz xdx + ydy + zdz
= = =
x −y −z
2 2 2
2 xy 2 xz x( x 2 + y 2 + z 2 )
xdx + ydy + zdz
Now = dz/2xz
x( x 2 + y 2 + z 2 )

18
 2( xdx + ydy + zdz )
 = dz/z
x2 + y2 + z2
Integrating
Log(x2+y2+z2) = log z+ log b
x2 + y2 + z2
=b (v =b)
z
x2 + y2 + z2
The general integral is (y/z, )=0
z
Class Room Delivery Problems
1. Solve 𝑦 2 𝑧𝑝 + 𝑥 2 𝑧𝑞 = 𝑦 2
2. Obtain the solution of 𝑝√𝑥 + 𝑞 √𝑦 = √𝑧 using Lagrange’s method
3. Solve (𝑚𝑧 − 𝑛𝑦)𝑝 + (𝑛𝑥 − 𝑙𝑧)𝑞 = 𝑙𝑦 − 𝑚𝑥
4. Obtain the solution of partial differential equation 𝑝𝑦𝑧 + 𝑞𝑧𝑥 = 𝑥𝑦

Practice Problems
1. Obtain the solution of partial differential equation 𝑝𝑦𝑧 + 𝑞𝑧𝑥 = 𝑥𝑦
2. Obtain the solution of partial differential equation (𝑦 2 + 𝑧 2 ) 𝑝 – 𝑥𝑦𝑞 + 𝑥𝑧 = 0
3. Obtain the solution of partial differential equation 𝑝 tan 𝑥 + 𝑞 tan 𝑦 = tan 𝑧. Using
Lagrange’s method
4. Obtain the solution of partial differential equation of (𝑦 − 𝑧) 𝑝 + (𝑧 − 𝑥) 𝑞 = 𝑥 −
𝑦
Session 20
RANDOM VARIABLES
A random variable is a function that associates a real number with each element in the sample
space.
Ex: The testing of three of electronic components for defectives is a random experiment. The
associated sample space is
S = NNN , NND , NDN , DNN , NDD, DND, DDN , DDD where N denotes non defective
and D denotes “defective”. If X is the number defectives. Then for each point in the sample
space, X associates real numbers 0, 1, 2, or 3.
Types of Random Variables: Random variables are of two types: Discrete random variable
and continuous random variable.

19
Discrete Random Variable: A random variable is called a discrete random variable if its set
of possible outcomes is countable. If X is the number of defective components out of 3
components. Then X is a discrete random variable.
Discrete probability distribution function: A discrete random variable assumes each of its
values with a certain probability. The table showing the values of the random variable and
associated probabilities is called the discrete probability distribution.
Ex: In the case of tossing a coin three times, the variable X, representing the number of heads
has the following probability distribution.
x 0 1 2 3
f(x)=P(X=x) 1/8 3/8 3/8 1/8

Probability function: The set of ordered pairs (x, f(x)) is called the probability function or
probability mass function or probability distribution of the discrete random variable X, if for
each possible outcome x,
1. f ( x)  0.

2.  f ( x) = 1
x

3. P( X = x) = f ( x).
Cumulative distribution: The cumulative distribution function F(x) of a discrete random
variable X with probability distribution f(x) is
F ( x) = P( X  x) =  f (t ) for −   x .
tx

Properties:
(i) F (−) = 0
(ii) (ii) F () = 1
(iii) (iii) P(a  X  b) = F (b) − F (a)
(iv) F(x) is a non-decreasing function.

Example:
Given that f(x)=k/2x is a probability distribution for a random variable that can take on the
values x=0, 1, 2, 3 and 4. Then, (a) Find K, (b) Find the Cumulative probability distribution
F(x)
Solution: (a) since f(x) is the probability mass function

20
 4

∑ f(x) = 1
x=0

 f(0)+f(1)+f(2)+f(3)+f(4)=1
 k/1+k/2+k/4+k/8+j/16=1
 K[31/16]=1
 K=16/31.
(b) F(0)=f(0)=k=16/31
F(1)=f(0)*f910=K+(K/2)=(3/2)K=24/31
F(2)=f(0)+f(1)+f(2)=K+(K/2)+(K/4)=(7/4)*K=28/31
F(3)=f(0)+f(1)+f(2)+f(3)=K+(K/2)+(K/4)+(K/8)=30/31.
F(4)=f(0)+f(1)+f(2)+f(3)+f(4)=K+(K/2)+(K/4)+(K/8)+(K/16)=1
c)f(2)=F(2)-F(1)=(28/31)-(24/31)=4/31.
Continuous Random Variable: When a random variable can take on values on a continuous
scale, it is called a continuous random variable.
Ex: Le X be the random variable defined by the waiting time, in hours, between successive
speeders spotted by a radar unit. The random variable X take on values x for which x  0.
Continuous Probability Distribution: A continuous random variable has a probability of
zero of assuming exactly any of its values. Consequently, its probability distribution cannot be
given in tabular form.
Density Function: In dealing with continuous variables, f(x) is usually called the probability
density function or simply the density function of X.
The function f(x) is a probability density function for the continuous random variable X,
defined over the set of real numbers R, if
1) f ( x)  0, for all x  R

2)  f ( x)dx = 1
−

b
3) P(a  x  b) =  f ( x)dx.
a

Cumulative Distribution Function (CDF): The cumulative distribution function F(x) of a

continuous random variable X with density function f(x) is
x
F ( x) = P( X  x) =  f (t )dt ,
−
for −   x  

21
 dF ( x)
Note: P(a  X  b) = F (b) − F (a) and f ( x) = , if the derivate exits.
dx
Example:
An important factor in solid missile fuel is the particle size distribution. Significant problems
occur if the particle sizes are too large. From the production data in the part, it has been
determined that the particle size (in micrometers) distribution is characterized by
f(x) =3x-4,
= 0, elsewhere
a) Verify that this is a valid density function
b) Evaluate the Cumulative distribution function F(x)
c) What is the probability that a random particle from the manufactured fuel exceeds 4
micrometers?
d) What is the probability that a random particles’ size is between 2 and 4 micrometers?
Solution:
∞
a) f(x) is a valid density function if ∫−∞ 𝑓(𝑥)𝑑𝑥 = 0, x≤X
∞ ∞ ∞
−4
𝑥 −3
∫ 𝑓(𝑥)𝑑𝑥 = ∫ 3𝑥 = 3[ ] =1
−3 1
0 1

Therefore, f(x) is a valid density function.

b) Cumulative distribution function F(x) is given by
x 1
f(x)=P(X≤ 𝑥)=∫1 3x −4 = 1 − x3
0, x < 1
∴ F(x) = { 1
1 − x3 , x ≥ 1

c) Probability that a random particle’s size exceeds 4 micrometers

=P(X>4) = 1−F(4) = 1−(1−(1/43)=0.0156.
d) Probability that a random particle’s size is between 2 and 4 micrometers
=P(2≤X≤4)=F(4)-F(2)=(1-1/43)-(1-1/23)=1/8-1/64 =0.109.
Class room delivery problems
1. A lot containing 7 components is sampled by a quality inspector; the lot contains 4 good
components and 3 defective components. A sample of 3 is taken by the inspector. Find the
expected value and variance of the number of good components in this sample.

22
2. A shipment of 8 similar microcomputers to a retail outlet contains 3 defectives. If a school
makes a random purchase of 2 of these computers, find the probability distribution for the
number of defectives. Also determine the mean and variance of X.

3. Consider the density function 𝑓(𝑥) = {𝑘√𝑥, 0 < 𝑥 < 1

0, 𝑒𝑙𝑠𝑒𝑤ℎ𝑒𝑟𝑒
i. Evaluate k.
ii. Evaluate P(0.3<X<0.6) using the density function
iii. Determine mean and variance of X
 x2
 −1  x  2
4. Let X be a random variable with density function f ( x) =  3
 0 elsewhere

Find the expected value and variance of g ( X ) = 4 X + 3.

Practice Problems
1. The shelf life (in hour) of a certain perishable packaged food is a random variable whose
20,000
3 , 𝑓𝑜𝑟
𝑥>0
probability density function is given by 𝑓(𝑥) = {(𝑥+100) . Find the
0, 𝑒𝑙𝑠𝑒 𝑤ℎ𝑒𝑟𝑒
probabilities that one of these packages will have a shelf life of (a) at least 200 hours, (b)
at most 100 hours, (c) anywhere from 80 to 120 hours.
2. The probability density function of the random variable X is given by
𝑐
, 𝑓𝑜𝑟 0 < 𝑥 < 4
√𝑥
𝑓(𝑥) = { 0, 𝑒𝑙𝑠𝑒 𝑤ℎ𝑒𝑟𝑒 .

Find (i) the value of c (ii) P(X<1) and (iii) 𝑃(𝑋 ≥ 1).
Session 21
Binomial and Poisson distributions
BINOMIAL DISTRIBUTION
A random variable X is said to follow a binomial distribution if it assumes only non negative
integral values and its probability mass function is given by
b(x; n, p) =𝑛𝑐𝑥 𝑝 𝑥 𝑞 𝑛−𝑥 , for x = 0, 1, , n, p + q = 1
Where
n is the number of independent trails
p is the probability of success of each trail
q is the probability of failure of each trail
x is the number of success

23
Note
1) X ~ b(n, p) to denote that X follows binomial distribution with parameters n and p
2) Each trial has two possible outcomes, one is called “success” and the other is “failure”.
3) The outcomes are mutually exclusive and collectively exhaustive.
4) The probabilities of success p and of failure 1 − p remain the same for all trials.
5) The outcomes of trials are independent of each other.
6) The mean of the binomial distribution is ‘np’
7) The variance of the binomial distribution is ‘npq’
8) The standard deviation of binomial distribution is √𝑛𝑝𝑞
9) In binomial distribution, the mean is always greater than the variance
POISSON DISTRIBUTION
When n is large and p is very small, binomial probabilities are often approximated by means
of the Poisson distribution with the parameter λ equal to the product np i.e., Poisson
distribution is used in case of rare events.
Experiments yielding numerical values of a random variable X, the number of outcomes
occurring during a given time interval or in a specified region (Here n is not known and
information regarding the number of occurrences of event is known), are called Poisson
experiments.
The number X of outcomes occurring during a Poisson experiment is called a Poisson random
variable and its probability distribution is called the Poisson distribution.
Definition: The probability distribution of the Poisson random variable X, representing the
number of outcomes occurring in a given time interval or specified region denoted by t, is
e−λt (λt)x
P(x; λt) = , x = 0,1,2, … ..
x!
where λ is the average number of outcomes per unit time, distance, area or volumand e=2.71828
Note:Both the mean and variance of the Poisson distribution P(x; λt) are λt.
1) If the time t is one unit then
e−λ (λ)x
P(x; λ) = , x = 0,1,2, … ..
x!
2) We write, X ~ P(x, ) to denote that X follows Poisson distribution with parameter .
The following are the some of the examples of random variables following Poisson
distribution:
➢ The number of customers arrived during a time period of length t.

24
 ➢ The number of telephone calls per hour received by an office.
➢ The number of typing errors per page.
➢ The number of accidents occurred at a junction per day.
Class room delivery problems
1. It has been claimed that in 60% of all solar heat installations the utility bill is reduced by at
least one third. Accordingly, what are the probabilities that the utility bill will be reduced
by at least one third in (i) four or five installations (ii) at least four of five installations? (iii)
P(x>1).
2. A car hire firm has two cars which it hires out day by day. The number of demands for a car
on each day is distributed as a Poisson distribution with mean 1.5. Calculate the proportion
of days. (i) on which there is no demand (ii) on which demand is refused. (iii)Also find
P(x>2).
3. 4. A hospital switch board receives an average of 4 emergency calls in a 10 min.
interval. What is the probability that (i) there are at the most 2 emergency calls (ii) there
are exactly 3 emergency calls(iii) there are at least 2 emergency calls in a 10 min. interval?
Practice Problems
1. During one stage in the manufacture of integrated circuit chips, a coating must be applied.
if 70% of chips receive a thick enough coating, find the probabilities that, among 15 chips:
i. At least 12 will have thick enough coating
ii. At most 6 will have thick enough coating
iii. exactly 10 will have thick enough coating
2. In a given city, 6% of all drivers get at least one parking ticket per year. Use the Poisson
approximation to the binomial distribution to determine the probabilities that among 80
drivers (i) 4 will get at least one parking ticket in any given year (ii) at least 3 will get at
least one parking ticket in any given year (iii) anywhere from 3 to 6, inclusive, will get at
least one parking ticket in any given year.
3. It is conjectured that an impurity exists in 30% of all drinking wells in a certain rural
community. In order to gain some insight on this problem, it is determined that some tests
should be made. It is too expensive to test all of the wells in the area, so 10 were randomly
selected for testing,
i. Using the binomial distribution, what is the probability that exactly 3 wells
have the impurity assuming that the conjecture is correct?
ii. What is the probability that more than 3 wells are impure?
iii. What is the probability that not more than 2 wells?
25
4. If 3 % of the electric bulbs manufactured by a company are defective, find the probability
that in a sample of 100 bulbs
i. None is defective
ii. one is defective
ii. 4 are defective
iii. Find mean and variance of the distribution
5. Ten percent of the tools produced in a certain manufacturing process turn out to be
defective. Find the probability that in a sample of 10 tools chosen at random exactly two
will be defective by using the Poisson approximation to the binomial distribution?
Session 22
Normal Distribution
The Normal Distribution (N.D.) was first discovered by De-Moivre as the limiting form of the
binomial model in 1733, later independently worked Laplace and Gauss.
The Normal distribution is the most important distribution in statistics. It is a probability
distribution of a continuous random variable and is often used to model the distribution of
discrete random variable as well as the distribution of other continuous random variables. The
basic from of normal distribution is that of a bell, it has single mode and is symmetric about its
central values. The flexibility of using normal distribution is due to the fact that the curve may
be centered over any number on the real line and it may be flat or peaked to correspond to the
amount of dispersion in the values of random variable.
Definition
A random variable X is said to follow a Normal Distribution with parameter mean (µ) and
variance (𝜎 2 ) if its density function is given by the probability law
1 1 𝑥−µ 2
− ( )
𝑓(𝑥) = 𝑒 2 𝜎 , −∞ < 𝑥 < +∞; −∞ < 𝜇 < +∞; 𝜎 > 0
𝜎√2𝜋
Symbolically we can represent the normal variate as X ~ N(, 2)
The Normal Probability Curve
The graph of the normal distribution depends on two factors – the mean and the standard
deviation. The mean of the distribution determines the location of the center of the graph, and
the standard deviation determines the height and width of the graph. When the standard
deviation is large, the curve is short and wide; when the standard deviation is small, the curve
is tall and narrow. All normal distributions look like a symmetric, bell-shaped curve, as shown
below.

26
The curve on the left is shorter and wider than the curve on the right, because the curve on the
left has a bigger standard deviation.
Standard Normal Distribution
𝑋−𝜇
If X is a normal variate with mean (µ) and standard deviation () then 𝑍 = is a standard
σ

normal variate with mean (0) and standard deviation (1).

The probability density function of the standard normal variate Z is given by the probability
law
1 𝑧2
−
𝑓(𝑧) = 𝑒 2 , −∞ < 𝑧 < +∞
√2𝜋
Symbolically we can represent the standard normal variate as Z ~ N(0, 1)
A graph representing the density function of the Normal probability distribution is also known
as a Normal Curve or a Bell Curve (see Figure below). To draw such a curve, one needs to
specify two parameters, the mean and the standard deviation. The graph below has a mean of
zero and a standard deviation of 1, i.e., (µ =0,  =1). A Normal distribution with a mean of
zero and a standard deviation of 1 is also known as the Standard Normal Distribution.

Characteristics of Normal distribution and normal curve:

The normal probability curve with mean µ and standard deviation  is given by the equation
1 1 𝑥−µ 2
𝑒− ( )
𝑓(𝑥) = 2 𝜎 , −∞ < 𝑥 < +∞; −∞ < 𝜇 < +∞; 𝜎 > 0
𝜎√2𝜋
and has the following properties
i) The mean, median and mode of the normal distribution coincide, i.e., mean = median =
mode = µ. (The height of normal curve is at its maximum at the mean. Hence the mean and

27
mode of normal distribution coincides. Also the number of observations below the mean in a
normal distribution is equal to the number of observations above the mean. Hence mean and
median of N.D. coincides.)
ii) The curve is bell shaped
iii) The normal curve is symmetrical about the line x= µ
iv) As x increases numerically, f(x) decreases rapidly, the maximum probability occurring at
1
the point x = µ, and given by 𝑝[(𝑥)]𝑚𝑎𝑥 = 𝜎√2𝜋

v) Linear combination of independent normal variates is also a normal variate

+∞
vii) The total area under the normal curve (∫−∞ 𝑓(𝑥)𝑑𝑥 = 1) is distributed as follows
➢ (𝜇 − 𝜎) < 𝑥 < (𝜇 + 𝜎)covers 68.26% of the area
➢ (𝜇 − 2𝜎) < 𝑥 < (𝜇 + 2𝜎)covers 95.44% of the area
➢ (𝜇 − 3𝜎) < 𝑥 < (𝜇 + 3𝜎)covers 99.74% of the area, and it can be represented as
follows

Note
1) The standard normal distribution, N (0, 1), is very important because probabilities of
any normal distribution can be calculated from the probabilities of the standard normal
distribution.
X −
2) If X is a normal random variable with mean  and standard deviation , then Z =

x1 −  x2 − 
is a standard normal random variable and hence P( x1  X  x2 ) = P( Z )
 
3) Suppose Z ~ N(0, 1) is standard normal variate then by using the standard normal
distribution area tables, we can calculate the various probabilities as explained below:
i) P(Z<a)P(Z≤a) is Directly from the table we can get the value and the
corresponding probability as shown below

28
 ii) P(Z>b) = this probability can be represented by using the following graph of
standard normal distribution and it cannot be obtained directly from the tables

P(Z>b)=1- P(Z≤b), where P(Z≤b) available directly from table.

iii) P(a≤Z≤b) = this probability can be represented by using the following graph of
normal distribution

P(a≤Z≤b) = P(Z≤b) – P(Z≤a), where P(Z≤b) and P(Z≤a) available directly

from the tables.
4) The Normal Approximation to the Binomial Distribution
Given X is a random variable which follows the binomial distribution with parameters
n and p, then the limiting form of the distribution function of standard normal variate
𝑋−𝑛𝑝
is 𝑍 = , 𝑤ℎ𝑒𝑟𝑒 𝑞 = 1 − 𝑝 provided, if n is large and p is not close to 0 or 1. If
√𝑛𝑝𝑞

both np and nq are greater than 5, the approximation will be good.

Class room delivery Problems
1. If a random variable having standard normal distribution, find the probabilities that it will
take on a value:

29
 i. less than 1.50;
ii. greater than 2.16;
iii. less than -1.20;
iv. greater than -1.75;
v. between 0.87 and 1.28;
vi. between -0.34 and 0.62
2. Given a standard normal distribution find the value of k such that
i. 𝑃(𝑧 < 𝑘) = 0.0427;
ii. 𝑃(𝑧 > 𝑘) = 0.2946;
iii. 𝑃(−0.93 < 𝑍 < 𝑘) = 0.7235.
3. Suppose that the actual amount of instant coffee that a filling machine puts int6-ounce.
Jans is a random variable having a normal distribution with SD=0.05 ounce. If any 3% of
Jans are to contain less than 6 ounces of coffee, what must be the mean fill of these Jans?
4. Suppose that during periods of transcendental meditation the reduction of a person’s
oxygen consumption with mean=37.6cc/min and SD =4.6CC/min. Find the probabilities
that during a period of transcendental meditation a person’s oxygen consumption will be
reduced by
i. at least 44.5 CC/min
ii. at most 35.0CC/min
iii. anywhere from 30.0 to 40.0CC/min
5. In a photosnapic process, the developing time of prints may be looked upon as a random
variable having the normal distribution with mean 15.40 seconds and standard deviation
0.48 sec. Find the probabilities that the time it takes to develop one of the prints will be
i. at least 16.00 sec
ii. at most 14.20 sec
iii. anywhere from 15.00 to 15.80 sec
Practice Problems
1. Given a Standard Normal distribution, find the area under the curve which lies
i. To the left of z=1.43;
ii. To the right of z= -0.89
iii. Between z=-2.16 and z=-0.65
iv. To the left of z=-1.39
v. To the right of z=1.96

30
 vi. Between z=-0.48 and z= -1.74
2. Given the normally distributed variable X with mean 18 and standard deviation 2.5,
find
i. 𝑃(𝑋 < 15);
ii. the value of k such that 𝑃(𝑋 < 𝑘) = 0.2236;
iii. the value of k such that 𝑃(𝑋 > 𝑘) = 0.1814;
iv. 𝑃(17 < 𝑋 < 21).
3. Given random variable X having a normal distribution with μ=50 and σ =10, find the
probability that X assumes a value
i. between 45 and 62
ii. less than 20
iii. more than 40
4. A random variable has a normal distribution with SD=10. If the probability that the
random variable will take on a value less than 82.5% is 0.8212. What is the probability
that it will take on a value greater than 58.3?
5. If 23% of all patients with high blood pressure have bad side effects from a certain kind of
medicine use the normal distribution to determine the probability that among 120 patients
with high blood pressure treated with this medicine more than 32 will have bad side
effects?
Sessions 23
Markov Process
Stochastic process: A family of random variables which are functions of say, time are known
as stochastic process (or random process).
Eg 1. Consider the experiment of throwing an unbiased die. Suppose that X n is the outcome

of the nth throw, n  1. Then X n , n  1 is a family of random variables such that for distinct

values of n (=1,2,3…), one gets distinct random variable X n ; X n , n  1 constitutes a

stochastic process.
Eg 2. Suppose that X n is the number of sixes in the first n throws. For a distinct value ofn=

1,2,3……., we get a distinct binomial variable X n ; X n , n  1 which gives a family of random

variables is a stochastic process.

Eg 3. Suppose that X n is the maximum number shown in the first ‘n’ throws. Here X n , n  1

constitutes a stochastic process.

31
Eg 4. Consider the number of telephone calls received at a switch board. Suppose that X(t) the
random variable which represents the number of incoming calls in an interval (o, t) of duration
t units. The number of calls in one unit of time is X(1). The family { X (t ), t  T } constitutes a
stochastic process (T = [0, )) .
State space:The set of possible values of a single random variable X n of a stochastic process

X n , n  1 is known as its state space. The state space is discrete if it contains a finite or

denumerable infinite of points; otherwise, it is continuous.

Eg 1. If X n is the total number of sixes appearing in the first n throws of a die, the set of

possible values of X n is the finite set of non negative integers 0,1,…n Hence the state space

of X n is discrete. We can write X n = y1 + y 2 + ... + y n , where yi is a discrete random variable

denoting the outcome of ith throw and yi=1 or 0 according as ith throw shows six or not.
Eg 2. Consider X n = z1 + z 2 + ..... + z n , where zi is a continuous random variable assuming

values in [0, ). Here, the set of values of X n is the interval [0, ), and so the state space of

X n is continuous.

Discrete time (parameter) stochastic process:In the above two examples we assume that the
parameter n of Xn is restricted to be non-negative integer, n=0,1,2,…. We consider the state of
system at discrete time points n=0,1,2,…, only. Here the word time is used in a wide sense. In
the first example “time n” implies throw number n.
Continuous time (parameter) stochastic process: we can visualize a family of random
variables { Xt , t  T } (or { X (t ), t  T } ) such that the state of the system is characterized at
every instant over a finite or infinite interval. The system is then defined for a continuous
range of time and we say that we have a family of random variables in continuous time.
A stochastic process in continuous time may have either a discrete or continuous state
space.
Eg 1: Suppose that X(t) is the number of telephone calls at a switch board in an interval (o,
t). Here the state space of X(t) is discrete though X(t) is defined for a continuous range of time.
This is a continuous time stochastic process with discrete state space.
Eg 2. Suppose that X(t) represents the maximum temperature at a particular place in (0, t), then
set of possible values of X(t) is continuous. This is a continuous time stochastic process with
continuous state space.
Thus the stochastic processes can be classified into the following four types of processes:
(i) Discrete time; discrete state space

32
 (ii) Discrete time; continuous state space
(iii) Continuous time; discrete state space
(iv) Continuous time; continuous state space
All the four types may be represented by { X (t ), t  T } . In case of discrete time, the
parameter generally used is n, i.e,.the family is represented by { X (n), n = 0,1,2....} . In case
of continuous time both the symbols { X t , t  T } and { X (t ), t  T } ( where T is finite or

infinite interval) are used. The parameter t is usually interpreted as time, through it may
represent such characters as distance, length, thickness and soon.
Markov process: If { X (t ), t  T } is a stochastic process such that, for, t1  t 2  ...  t n  t

Pr a  X (t )  b / X (t1 ) = x1 , X (t 2 ) = x 2 ,... X (t n ) = x n  = Pra  X (t )  b / X (t n ) = x n  the

process { X (t ), t  T } is a Markov process.

Markov chain: A discrete parameter Markov process is know as a Markov chain.
Definition:The stochastic process X n , n = 0,1,2,3,... is called a Markov chain if, for

j , k , j1 ,  j n −1  N

Pr X n = k / X n −1 = j , X n − 2 = j1 ,  , X 0 = j n −1  = PrX n = k / X n −1 = j,

whenever the first term is defined.

The outcomes are called the states of the Markov chain; If X n has the outcome j

(i.e. X n = j ) , the process is said to be at state j at the nth trail. To a pair of states (j, k) at

the two successive trails ( nth and (n+1)th trails) there is an associated conditional probability
Pjk .

Transition probability: Pjk is called the transition probability and represents the

probability of transition from state j at the nthtrial to the state k at the (n+1)th trail.
Homogeneous Markov chain: If the transition probability Pjk is independent of n, the

Markov chain is said to be homogeneous. If it is dependent on n, the chain is said to be

non-homogeneous.
One step transition probability: The transition probability Pjk refer to the states (j, k) at

two successive trails (say nth and (n+1)th trails); the transition is one step transition
probability.

33
 If we are concerned with the pair of states (j, k) at two non-successive trails, say, j at
the nth trail and k at the (n+m)th trail, the corresponding probability is then called m-step

= PrX n + m = k / X n = j.
( m)
transition probability and is denoted by Pjk

Transition probability Matrix or Matrix of transition probabilities: The transition

probability Pjk satisfy Pjk >0, P jk =1 for all j. These probabilities may be written in the

matrix form
 p11 p12 p13 
p p 22 p 23 
 21
P=   
 
   
 

This is called the transition probability Matrix of the Markov chain. P is a stochastic matrix.
Thus a transition matrix is a square matrix with non-negative elements and unit-row sums.
Example 1: Consider a communication system which transmits the two digits 0 and 1
through several stages. Let X n , n  1 be the digit leaving the nth stage of system and X0 be

the digit entering the first stage. At each stage there is a constant probability q that the digit
which enters will be transmitted unchanged, and probability p otherwise, p+q=1.
Here X n, n  0 is a homogeneous two-state Markov chain with single step transition

matrix
0 1

0 q p
P= 
1p q 
Example 2. A particle performs a random walk with absorbing barriers say, as 0 and 4.
Whenever it is at any position r (0 < r < 4), it moves to r+1 with probability p or to (r-1)
with probability q, p+q=1. But as soon as it reaches 0 or 4 it remains there itself. Let Xn be
the position of the particle after n moves. The different states of Xn are the different
positions of the particle {Xn} is a Markov chain whose unit-step probabilities are given by
PrX n = r + 1 / X n −1 = r = p
PrX n = r − 1 / X n −1 = r = q, 0  r  4
and
PrX n = 0 / X n −1 = 0 = 1
PrX n = 4 / X n −1 = 4 = 1

34
The transition matrix is given by
States of X n

0 1 2 3 4

0 1 0 0 0 0
1 q 0 p 0 0 
States of X n −1 2  0 q 0 p 0
 
3 0 0 q 0 p
4  0 0 0 0 1 
Eg 3: General case of random walk: In the above example, as soon as the particle reaches
0 it remains there with probability ‘a’ and is reflected to state 1 with probability 1-a(0 < a
<1); if it reaches state 4 it remains there with probability b and reflected to state 3 with
probability 1-b (o < b<1) then {Xn} is a Markov chain with state space {0,1,2,3,4}.
The transition matrix is given by
0 1 2 3 4

0 a 1 − a 0 0 0
1 q 0 p 0 0 
2 0 q 0 p 0
 
3 0 0 q 0 p
4 0 0 0 1− b b 

If a=1, then ‘0’ is an absorbing barrier and if a=0, then ‘0’ is a reflecting barrier similar is
a case with state 4.
Finite Markov chain: A Markov chain X n , n  0 with k states, where k is finite, is said

to be a finite Markov chain. The transition matrix P in this case is a square matrix with k
rows and k columns.
The number of states could however infinite. Where the possible values of X n form a

denumerable set, then the Markov chain is said to be denumerable infinite or denumerable
and the chain is said to have a countable state space.
Probability distribution:
The probability distribution of X 0 , X 1 , X 2 ,  , X n can be computed in terms of the

transition probability p jk and the initial distribution of X 0 .

35
 PrX 0 = a, X 1 = b,  , X n − 2 = i, X n −1 = j , X n = k 
= PrX n = k / X n −1 = j , X n − 2 = i,  , X 1 = b, X 0 = a
= PrX n = k / X n −1 = j prX n −1 = j , X n − 2 = i,  , X 0 = a
= PrX n = k / X n −1 = jpr X n −1 = j / X n − 2 = i PrX 1 = b / X 0 = aPrX 0 = a
= p jk p ij  p ab PrX 0 = a.

Eg: Let X n , n  0 be a Markov chain with three states 0,1,2 with transition matrix

0 1 2
0 3 1 
4 0
4
1  
1 1 1
2 4 2 4
 3 1 
0
 4 4 
and the initial distribution

PrX 0 = i = , i = 0,1, 2.
1
3

We have PrX 1 = 1 / X 0 = 2 =
3
4

PrX 2 = 2 / X 1 = 1 =
1
4
PrX 2 = 2, X 1 = 1 / X 0 = 2

= PrX 2 = 2 / X 1 = 1PrX 1 = 1 / X 0 = 2 = . =
1 3 3
4 4 16

PrX 2 = 2, X 1 = 1, X 0 = 2 = PrX 2 = 2, X 1 = 1 / X 0 = 2PrX 0 = 2 =

3 1 1
. = .
16 3 16
PrX 3 = 1, X 2 = 2, X 1 = 1, X 0 = 2

PrX 3 = 1 / X 2 = 2Pr X 2 = 2, X 1 = 1, X 0 = 2 = . =
3 1 3
4 16 64
Note:The transition probabilities with the initial distribution completely specifies a
Markov chain.
Higher order transition probabilities:

𝒑𝒋𝒌 (𝟐) = 𝑷𝒓{𝑿𝒏+𝟐 = 𝒌/𝑿𝒏 = 𝒋} = ∑ 𝑷𝒓{𝑿𝒏+𝟐 = 𝒌, 𝑿𝒏+𝟏 = 𝒓/𝑿𝒏 = 𝒋}

𝒓

= ∑ 𝒑𝒋𝒓 𝒑𝒓𝒌
𝒓

𝑝𝑗𝑘 (𝑚+𝑛) = ∑ 𝑝𝑟𝑘 (𝑛) 𝑝𝑗𝑟 (𝑚) = ∑ 𝑝𝐽𝑟 (𝑛) 𝑝𝑟𝑘 (𝑚)
𝑟

36
Let p = ( p jk ) denote the transition matrix of the unit-step transition and P ( m ) = Pjk ( m ) denote

the transition matrix of the m-step transitions. For m=2, we have the matrix 𝑝(2) = 𝑝. 𝑝 = 𝑝2
similarly 𝑝(𝑚+𝑛) = 𝑝(𝑚) 𝑝 = 𝑝. 𝑝(𝑚)
Classification of chains: The Markov chains are of two types (i) ergodic (ii) regular
An ergodic Markov chain has the property that it is possible to pass from one state to another
in a finite number of steps, regardless of present state.
A special type of ergodic Markov chain is the regular Markov chain.
A regular Markov chain is defined as a chain having a transition matrix P such that for some
power of P it has only non-zero positive probability values.
Note: Thus all regular chains must be ergodic chains.
Example:Consider a communication system which transmits the digits 0 and 1 through several
stages. At each stage the probability that the same digit will be received by the next stage, as
transmitted, is 0.75. What is the probability that a 0 that is entered at the first stage is received
as a 0 by the 5th stage?
Solution: Now we want to find the P004
0.75 0.25
The state transition matrix P is given by
P = 0.25 0.75
 
0.625 0.375
Hence P = 0.375 0.625
2

 

0.53125 0.46875
And
P 4
= P 2 2
P = 0.46875 0.53125
 
Therefore, the probability that a zero will be transmitted through four stages as a zero is
P004=0.53125.
Classification of states: Transient and persistent(Recurrent) let f jk (n ) be the probability that

the system starts with state j and reaches the state k for the first time after n transitions. Let
(n )
p jk be the probability that it reaches state k after n transitions(not necessarily first time).

Let F jk denote the probability that starting with state j ever reach state k. Clearly

F jk =  f jk
(n)

n =1

When F jk = 1, it is certain that the system starting with state j will reach state k.

The mean (first passes) time from state j to state k is given by  jk =  nf jk
(n)
.
n =1

37
 
 jj =  nf jj ( n ) is known as the mean recurrence time for the state j.
n =1

Persistent (recurrent) state: A state j is said to be persistent if F jj = 1 and transient if F jj  1

. A persistent state is said to be null persistent if  jj = , i.e, if the mean recurrence time is

infinite and is said to be non-null(positive) persistent if  jj  .

Thus the states of Markov chain can be classified as transient and persistent, and persistent
states can be subdivided as non-null and null persistent.
A non-null and periodic state of Markov chain is said to be ergodic.
Periodic and Aperiodic: A recurrent state is said to be aperiodic if its period di=1, and
periodic if di>1.
Communicating Classes: States i and j communicate if each is accessible from the other.
Transient state:Once the process is in state i, there is a positive probability that it will never
return to state i,
Absorbing state:A state i is said to be an absorbing state if the (one step) transition probability
Pii = 1.

Note:
1. ifj is transient P(Xn = j | X0 = i) = Pij(n)→ 0
1 n (k )
2. If chain is irreducible:as n →∞ Pij →  j
n k =1
3. If chain is irreducible and aperiodic: Pij(n)→𝜋j as n→∞

38
Example: Consider a Markov chain on S = {1, 2, . . . , 7} with transition probabilities

0.5 0.4 0.1 0 0 0 0

0 0.8 0 0 0 0.1 0.1
0 0 0 0 0 0 1
𝑃 = 0.5 0 0 0 0.5 0 0
0.3 0 0 0 0.7 0 0
0 0.3 0.7 0 0 0 0
[ 0 0 0 0 0 0.8 0.2 ]
From the transition graph, one can see that C = {2, 3, 6, 7}and S are the only two closed sets.
Also, C is a communication class and T= {1, 4, 5} are communication classes. From results
below, it follows that C is a class of positive recurrent states and T is a class of transient states.
Furthermore, the states in these classes are aperiodic.

We now establish the major result that all of the states in an irreducible set are of the same
type, and they have the same period.
Example:Consider a communication system which transmits the digits 0 and 1 through several
stages. At each stage the probability that the same digit will be received by the next stage, as
transmitted, is 0.75. What is the probability that a 0 that is entered at the first stage is received
as a 0 by the 5th stage?
Solution: Now we want to find the P004
0.75 0.25
The state transition matrix P is given by
P = 0.25 0.75
 
0.625 0.375
Hence P = 0.375 0.625
2

 

0.53125 0.46875
And
P 4
= P 2 2
P = 0.46875 0.53125
 
Therefore the probability that a zero will be transmitted through four stages as a zero is
P004=0.53125.

39
It is clear that this Markov chain is irreducuble and aperidoic.
We have the equations

The unique solution of these equations is 𝜋0= 0.5, 𝜋1 = 0.5. This means that if
data are passed through a large number of stages, the output is independent of the original input
and each digit received is equally likely to be a 0 or a 1. This also means that
0.5 0.5
lim 𝑃𝑛 = [ ]
𝑛→∞ 0.5 0.5
Class room delivery Problems
1. A city is served by two cable TV companies, Best TV and Cable cast. Due to aggressive
sales tactics, each year 40% of Best TV customers switch to Cable cast and the other 60%
of Best TV customers stay with Best TV. On the other hand, 30% of the Cable cast switch
to Best TV.
i. Express the information as a transition probability matrix.
ii. Suppose that this year ¼ of customers switch to Best TV and ¾ customers
subscribe to Cable cast. After one year, two years and in the long run what
percentage of customers subscribe to each company?
2. Consider a bike share problem with only 3 stations: A, B, C. Suppose that all bikes must
be returned to the station at the end of the day, so that all the bikes are at some station.
Each day, the distribution of bikes at each station changes, as the bikes get returned to
different stations from where they are borrowed. Of the bikes borrowed from station A,
30% are returned to station A, 50% end up at station B, and 20% end up at station C. Of
the bikes borrowed from station B, 10% end up at station A, 60% of have been returned to
of the bikes borrowed from station C, 10% end up at station A, 10% end up at station B,
and 80% are returned to station C.
i. Express this information as a transition probability matrix and determine the
probabilities of bike being at a particular station after two days.
ii. suppose when we start observing the bike share program, 30% of the bikes are at
station A, 45% of the bikes are at station B and 25% are at station C, determine
the distribution of bikes at the end of the next day and after two days.

40
 3 1
0
4 4
1 1 1
3. The transition probability matrix of a Markov chain with states 0,1,2 is P = 4 2 4
3 1
0
( 4 4)
1
and the initial state distribution of the chain is P(𝑋0=i) i) = 3 ,i = 0,1,2. Find (i) P(𝑋2=2)

and (ii) P(𝑋2=2, 𝑋1=1, 𝑋0=2).

4. In an industry the production of a certain product is regulated according to its existing stock
at the end of the day. Namely, if there exist unsatisfied orders or the stock is zero, then the
production of the next day covers the unsatisfied orders plus two more metric units (m.u.).
On the contrary, if there exists a non-zero stock, there is no production P (𝑥0 =i) for the next
day. We further know that the consumers’ demand for the product is either 1 m.u. per day
with probability 60%, or 2 m. u. per day with probability 40%. Calculate the probability to
have unsatisfied orders in the long run.
Practice Problems
1. A company circulates for first time in the market a new product, say K. The market’s
research shows that the consumers buy on average one such product per week, either K, or
a competitive one. It is also expected that 70% of those who buy K they will prefer it again
next week, while 20% of those who buy another competitive product, they will turn to K
next week. Find the market’s share for K two weeks after its first circulation, provided that
the market’s conditions remain unchanged.
2. Two boys B1 , B2 and two girls G1 , G2 are throwing a ball from one to another. Each boy
throws the ball to the other boy with probability ½ and to each girl with probability ¼. On
the other hand, each girl throws the ball to Each boy with probability ½ and never to the
other girl. In the long run, how often does each receive the ball?
3. A market research team has conducted a survey of customer buying habits with respect to
three brands of talcum power in an area. It estimates at present, 20% of the customers buy
brand A, 50% of the customers buy brand B and 30% of the customers buy brand C. In
addition, the team has analysed its survey and has determined the following brand
switching matrix.
Brand Next Bought A B C

Brand Just Bought

41
 A 0.6 0.3 0.1
B 0.4 0.5 0.1
C 0.2 0.1 0.7

Determine the expected distribution of consumers two time periods later.

4. Three children (denoted by 1,2,3) arranged in a circle play a game of throwing a ball to one
another. At each stage the child having the ball is equally likely to throw it into any one of
the other two children. Suppose that X 0 denote the child who had the ball initially and

X n (n  1) denotes the child who had the ball after n throws Find the transition probability

matrix P. Calculate PrX 2 = 1 / X 0 = 1. , PrX 2 = 2 / X 0 = 3 , PrX 2 = 3 / X 0 = 2 and also

the probability that the child who had originally the ball will have it after 2 throws.

******

42