Higher Engineering Mathematics

Professor P.N. Agrawal

Department of Mathematics
Indian Institute of Technology Roorkee
Lecture - 41
Graphical Methods - I
Hello friends, welcome to my lecture on Graphical Methods. So, let us see what is a linear
programming?

(Refer Slide Time: 00:41)

Linear programming is an important optimization that is maximization or minimization

technique used in decision making in business and everyday life for obtaining the maximum or
minimum values as required of a linear expression subject to satisfying certain number of given
linear restrictions. Now, the linear programming problem in general calls for optimizing that is
maximizing or minimizing a linear function of variables called the objective function subject to a
set of linear equations and or linear in equations called the constraints or restrictions, a linear
programming problem in short is written as LPP.
(Refer Slide Time: 01:17)

Now, mathematical District description of a general linear programming problem, a general

linear programming problem can be stated as so this is linear function. Okay, which is to be
maximize or minimize subject to the constraints. So these x1 , x2 , … , xn are subjected to these in
equations. Okay, a11 x1 +a12 x 2 +…+a1n x n it could be less than or equal to, it could be equal to, it
could be ¿ ≥b1 , a21 x 1 +a22 x2 +…+ a2 n xn ≤=≥ b2 , and am 1 x1 +am 2 x2 +…+ amn xn ≤=≥ bm . Now, if
there is equality sign equality here, then they are called equations, if there is inequality here then
they are known as in equations. So the linear expressions Z=x1 c1 + x 2 c 2 + …+ xn c n
has to be optimized that is maximized or minimized subject to these equations are in equations,
linear equations are linear in equations.
(Refer Slide Time: 02:29)

Now, and the non-negative restrictions. There is a restriction on the values of x1 , x2 , … , xn that
they are all non-negative. Now, aii here, these coefficients a11 ,a12 , … ,a1n ,

a21 , a22 , … , a2n , am 1 , am 2 , … . amn and these coefficients a11 , a12 , … ,amn and b1 , b2 , …, bm,
c 1 , c2 , … , c m, okay, they are constants. Okay. aij ,i=1¿ m∧ j=1¿ n Okay, they are constants
along b j=ab j , j=1¿ m okay and c i , i=1 , 2,.. , n so aij , b j and c i are all constants, xi , i=1¿ n are
all non-negative.

So these are non-negative restrictions and these are constants x1 , x2 ,… , xn are called variables.
Now, the function Z is called the objective function. We have to maximize or minimize this
function. So this is known as objective function the conditions given in 2, okay the conditions
given in 2 these conditions, these conditions are called as constraints.

Okay, linear constraints and the conditions is given in three. These are called as non-negative
restrictions on the linear programming problem.
(Refer Slide Time: 04:15)

Now, this linear programming problem may be stated in Matrix form. We can write it in Matrix
form Z=CX , where X =x1 , x 2 , … , x n C=c1 , c2 , … , c n okay, so Z=CX okay and

A=(aij)m× n okay X as we have seen X =x1 , x 2 ,… , x n so Ax will give you Ax ≤=≥ b b is equal to
the column vector b1 , b2 there are m equations b1 , b2 , … , bm . Okay. So the b is a column vector
column matrix, X is a column matrix, C is a row matrix. Okay, and A is m by n matrix. Ok so
A=(aij)m× n is called the coefficient matrix of the LPP.

(Refer Slide Time: 05:38)

C is row matrix as we said we call it a price vector, b is a column matrix, which is called
requirement vector and x column matrix of variables 0 is the null matrix of type n by 1, so 0 is
equal to n by 1 means 0 is of the type n rows. Okay, so 0 0 0 so this is n by 1 matrix.

(Refer Slide Time: 06:03)

Now, let us see what do we mean by solution of a linear programming problem a set of values of
the variables x1 , x2 , .. , xn satisfying the constraints that is the equations linear equations are linear
inequations given by 2. They are these are the constants is called a solution of the linear program
linear programming problem. Okay, so a set of values of the variables x1 , x2 ,.. , xn satisfying the
constraints and non-negative restrictions of a linear programming problem is called a feasible
solution of the linear programming problem.

So (we were) the set of values of the variables x1 , x2 , .. , xnwhich satisfy all the constraints and
also the non-negative restriction. That is x1 , x2 ,.. , xn are all non-negative that will be called a
feasible solution of the linear programming problem. Now, a feasible solution of a linear
programming problem is called optimal are optimum if it also optimizes that is maximizes or
minimizes as the case may be the objective function of the problem.

So, a feasible solution of the linear programming problem will then be called optimal okay, if it
is if it optimizes that is maximizes or minimizes the objective function of the problem.
(Refer Slide Time: 07:25)

Now, if the value of the objective function if the value of the objective function can be increased
or decreased indefinitely. Okay, such a solution will be called unbounded solution. So, the graph
or the graph or the solution set of a system of simultaneous linear inequations is region okay
containing the points xi we satisfy all the in equations of the given system simultaneously. Okay,
so we shall draw we shall use the graphical method. So we have to understand it very carefully
the graph or the solution set of a system of simultaneous linear inequations.

Okay, that is the values of the set of values of X =x1 , x 2 ,.., x nwhich satisfy the inequations will
determine this solution set and that will be the graph. Okay, so which satisfy all the inequations
of the given system simultaneously.

(Refer Slide Time: 08:29)

Now, let us see in general we use the now which are the methods which are to be which are the
methods that we use to solve a linear programming problem? First one is graphical method, if the
objective function Z is a function of two variables. Let us consider two variables case

Z=c1 x1 , c 2 x 2 +…+cn x n So when Z is a function of two variables only then the problem can be
solved by graphical method. Okay, a problem of three variables can also be solved by this
method but it is very complicated. Okay. So when we are when we have a problem of two
variables that is Z=c1 x1 , c2 x 2 +…+cn x n . We solve we find the optimum value of Z by using we
can find the optimum value object by using the graphical method.

Now, the LPP having more than two variables, if there is a case where Z is a function of more
than two variables. Okay, then we do not use graphical method as I as we have seen even for
three variables it is quite complicated. So we the problem even the problem three variables
becomes complicated so in such cases the analytical method, which is also called trial and error
method is useful.

(Refer Slide Time: 09:49)

Now Simplex method, this method is the most powerful method to solve a linear programming
problem. Any problem can be solved by this method and it is this method is an algebraic
procedure which progressively approaches the optimal solution as we shall see.
(Refer Slide Time: 10:07)

Now, graphical method to solve a linear programming problem. There are two techniques to
solve a LPP by graphical method. Okay one is the Corner Point method. The other one is Iso-
Profit or Iso-cost method. So what is the corner Point method? In this method we first obtain the
region before stopped in the region in the xy −¿ plane containing all points that simultaneously
satisfy all constraints including non-negative restrictions. Okay. So we have the constraints given
by the system of equations 2 okay and the non-negative restriction. That is x1 , x2 , … , xn ≥=0 , are
so we determine that region in the xy plane which is satisfied by all the constraints given by
system 2 end the non-negative restrictions on the n variables x1 , x2 ,… , xn This polygonal region
it will be will see that it is a polygonal region, this polygonal region is called the convex polygon
of the set of all feasible solutions of the linear programming problem.

Okay, so that that regular that polygonal region, which is a convex set is the cons is the set of all
feasible Solutions of the linear programming problem and this feasible, we call it feasible region
or we call it permissible region. So, the feasible region are permissible region is obtained by
finding the set which is the set of values of X and Y or x 1 and x2 which satisfy all the constraints
given in the, given by the system to end the non-negative restrictions.
(Refer Slide Time: 11:57)

Now, we determine the vertices of this poly convex polygon. So come the convex polygon that
that is the feasible region that we have got we determine the (vertex) vertices of that feasible
region. These vertices are called Extreme points of the set of all feasible Solutions of the linear
programming problem. Okay, after obtaining the extreme points, once we have found the
extreme points or we have located in the corner points of the convex polygon, we will find the
value of the objective function at all these extreme points.

Okay, the point where the objective function attains its optimum value that is maximum value or
minimum value of Z gives the optimal value of the given programming problem. Okay, so we
will determine the value of Z at all, the corner points then whichever value is minimum will give
us the optimum that is minimum value of the given programming problem or the corner point
which gives us the maximum value of Z will give us the optimum that is maximum value of the
given programming problem.

So this depends on the case as the so the given function will obtained its optimum value
maximum or minimum as the case may be okay by determining the value of Z at all the corner
points and then seeing which one is less or which one is maximum.
(Refer Slide Time: 13:22)

Now, if the two vertices of the convex polygon, suppose there may be a case where two vertices
of the convex polygon give the same optimal value of the objective function, then all points on
lying on the line segment joining these two vertices give the optimal value of the objective
function.

So, if there are two vertices of the convex polygon at which the value of Z comes out to be same
then all points lying on the line segment joining the two vertices they will all be giving the
optimal value of the objective function. So, in that case we say that LPP has infinite number of
Optimal Solutions.

(Refer Slide Time: 14:03)

What is Iso-Profit, Iso-Cost method? To determine the vertex of the polygon region which gives
the optimal value of the objective function Z, what we can do is? We draw a straight line in the
feasible region corresponding to the equation obtained by giving some convenient value to K say
Z=c1 x1 + c1 x 2. Ok, then we can consider Z=K that is c 1 x1 +c1 x 2=K , K is a constant we draw
this line we draw this a straight line in the feasible region. This straight line is drawn in the
feasible region. Okay, which is obtained by giving some Convent value K to the objective
function. This line is called Iso-Profit or Iso-Cost Line. (Iso-Profit or Iso-Cost line) Okay, since
every Point within the permissible region on this line will yield the same value of K.

Now, every point on this line within the permissible region will have same value K because K is
a constant. So whatever points x1 , x2 lie on this line, okay, c 1 x1 +c1 x 2=K means they will all
have the same give the same value of Z that is K. Now, we can also take k equal to 0 now it is it
is a makes it makes the determination of the optimum value of Z easier if we choose K=0, so if
we choose K=0 then we will have c 1 x1 +c1 x 2=0 that means it will give a line passing through
the origin and parallel to the Iso-Profit line, okay.

So take k equal to 0 draw the line passing through the origin. Okay, the line c 1 x1 +c1 x 2=0 that
will be Iso-profit line or Iso-cost line parallel to that will be a line this will be a line parallel to
the Iso-cost or Iso-profit line. Because Iso-cost or Iso-profit line is the one which is drawn in the
feasible region having the value K it is c 1 x1 +c1 x 2=K . So if you consider the line

c 1 x1 +c1 x 2=0 then this line has the same slope as the line this one Iso-profit line so they are
parallel. So this line is a line which is passing through the origin and is parallel to the Iso-cost or
Iso-profit line.
(Refer Slide Time: 16:51)

Now, we draw the lines parallel to this line. Okay over the permissible region. So, what we will
do is, so now we will draw lines parallel to this line. Okay, so this is suppose this is line. Suppose
this line is c 1 x1 +c1 x 2=0.Okay and this is your feasible region say let us draw rough figure.
Suppose this kind of figure we have okay, then we will draw lines parallel to this line. Okay, and
parallel to this line.

So, what will happen? Now, we draw lines parallel to this line over the permissible region if the
objective function Z is to be maximized then the parallel lines are drawn in the increasing the
direction away from the origin till it passes through a vertex of the polygon and its distance from
the origin is maximum. So, we want to find the maximum value, if we want to find the maximum
value of Z, then we will go on drawing these lines till it passes through a vertex, this one okay till
it passes through a vertex. This vertex is the farthest from the origin, so distance of this line this
line from the origin must be greatest the maximum okay till it passes through a vertex of the
polygon and it's distance of the line from the origin must be maximum. If the vertical parallel
line is drawn further away from the origin then this line will not intersect the feasible region.

So, if you draw the line further away like this, okay, then it will not intersect the feasible region
if you draw a line further away from the origin. So, now the coordinates of this vertex will give
the maximum value of the objective function. So this this vertex okay the coordinates of this
vertex this this line which is parallel to the Iso-profit or which is parallel to the line
c 1 x1 +c1 x 2=0 and has at the maximum distance from the origin. Okay, this vertex will then give
us the point at which we will have the maximum value of Z.

So we will put the coordinates of this point in the given expression of Z , Z=c1 x 1 +c1 x2 We will
have the maximum value of Z.

(Refer Slide Time: 19:38)

Now, if the objective function is to be minimized if we want to minimize then the parallel lines
are drawn towards the origin till it passes through the vertex whose distance from the origin is
minimum. So we will, in this case what we will do? Let us say this is your c 1 x1 +c1 x 2=0 and the
region is this. Okay, then we go on drawing lines parallel to this (Iso) this line c 1 x1 +c1 x 2 . Okay,
so the (cor) then the then we line whose distance from the origin is minimum this line, this line is
passing through the vertex. Okay through this vertex and its distance from the origin is
minimum. Okay so this vertex will then give us the minimum value of Z. We put the we put the
coordinates of this vertex in the given expression of Z . Z=c1 x 1 +c1 x2to get the minimum value
of the objective function Z. This called as Iso-cost or Iso-profit method.

(Refer Slide Time: 21:11)

Now that, in case, this line passing through, you know in the case of maximum or minimum if
this line okay, which has mixed maximum or minimum distance from the origin passes through
two or more vertices of convex polygon. Okay, then each point on the line segment between
these two vertices in the permissible region give the optimal value of the objective function.
Okay and in this case, we share that the problem has infinite solutions. So, here what may
happened that this line say for example in the case of minimum. It passes through two vertices.
Okay, then every point on this line will give the minimum value of Z, and we will have infinite
number of solutions.

Similarly, if this line parallel to the line c 1 x1 +c1 x 2=0 passes through the point this one. Okay,
where the distance of this line will be maximum from the origin then if again two vertices on
onto for two vertices, we have this line passes through two vertices this one this line like this this
passes through two vertices, then we will again have infinite number of solutions because every
point on that line will give us the maximum value of Z.
(Refer Slide Time: 22:30)

Now, consider this case we have Z=20 x1 +10 x 2 and we have these constraints x1 +2 x2 ≤ 40. So
we are given in equation 3 equations are there. Okay and B are given Z=20 x1 +10 x 2 , x1 , x 2 ≥0
let us see how we can apply the graphical method to determine solution to this problem. So what
we will do we will draw the lines straight lines given by x1 +2 x2 =40 , 3 x1 + x2 =30
4 x 1 +3 x 2=60 and from there and the non-negative restrictions X1 X2 greater than or equal to 0
we will find the permissible region to determine the minimum value of Z.

(Refer Slide Time: 23:20)

Now, let us see so this is the figure okay we consider three lines x1 +2 x2 =40. So this is

x1 +2 x2 =40 this line. Okay, then there is the then consider the line 3 x1 + x2 =30. So this is
3 x1 + x2 =30. Okay, then consider 4 x 1 +3 x 2=60 , 4 x 1 this is 4 x 1 +3 x 2=60 and when you draw
these three lines then the region which satisfies all the three inequalities, okay, this, this and this
and also the non-negative restrictions x1 , x2 ≥0 is this region this region.

Okay, this one shown by dots. Okay this region is satisfied when by those inequalities, although
inequality satisfy this region. Okay.

So a every point x1 , x2 in this dotted region satisfies those inequalities and the non-negative
restrictions. Now, what we will do? Let us let us locate the vertices of this convex polygon. This
is convex polygon. Okay, so vertices of this convex polygons are this 40, 0 the intersection of

x1 +2 x2 =40 and this line okay that is x 1 axis.

Okay on the x 1 axis x2 =0 so x1 =40 . So this 40, 0 point and here this line is 4 x 1 +3 x 2=0 so
x2 =0. So 4 x 1=60 so x1 =15, So this 15, 0 point and then when you consider the intersection
point of 3 x1 + x2 =30 that is this line and this line x1 +2 x2 =40 you can solve you will get the
intersection Point as 4,18. Okay, and then the intersection point of 4 x 1 +3 x 2=60 that is this line,
this line, and this line 3 x1 + x2 =30, this line. Okay that is this one. This has coordinates 6,12 so
6,12 is the intersection of lies on 3 x1 + x2 =30 and 4 x 1 +3 x 2=6 0 . Okay. So intersection of these
two is the point R. That is 6, 12. Okay.
So we have found the vertices of the convex polygon, convex polygon is you are S P Q R S, (S P
Q S)(SPQRS) Okay, this convex polygonor feasible region. So the vertices are vertices are P
okay Q okay R and we have S. Now, at all these four vertices of the convex polygon okay, we
will determine the value of Z OK and see which value of Z is minimum.

(Refer Slide Time: 26:55)

So, let us see P=40 , 0 Q=4 , 18 R=6 , 12 S=15 , 0 the value of Z comes out to be

Z=20 x1 +10 x 2, Z=800 for P=260 for Q and 240 for R and 300 for 15,0. Okay so whichever is
the least okay 260 is the least value. Okay, so Z is minimum at the point 240 is the least value
240 is the minimum value and it occurs at the point 6,12. Okay so the Z is minimum at the point
R okay and the optimal solution of the given LPP is Z occurs at and is obtained at x1 =6 and
x2 =12and the minimum value of Z=240. So that is how we apply this corner point method. This
is Corner Point method, solution of this LPP by Corner Point method.
(Refer Slide Time: 28:03)

Now, let us say we obtained the solution by the Iso-profit method then what we will do we draw
the line through the origin corresponding to Z=0so you see this line. Okay this line is

20 x 1 +10 x2=0 which means that 2 x 1 + x2 =0. Okay. So this line is 2 x 1 + x2 =0 and then we draw
lines parallel to this 2 x 1 + x2 =0 , this line, this line, this line and then this line. Okay. This line is
at the has the minimum distance from the origin and passes through (16) 6, 12 Point Okay, so the
point 6, 12. Okay, will give us the minimum value here.

Okay. So the dotted line through the origin is shown in the figure, drawing parallel lines from
away from the origin. We see that the nearest line, okay because it is a minimization problem. So
the nearest line in the permissible region passes through the vertex 6, 12 therefore the optimal
solution x1 =6 , x2 =12and minimum value of Z=240.
(Refer Slide Time: 29:14)

Now, let us take one more example suppose we consider the we want to find the maximum value
of Z=2 x1 +3 x2 subject to these constraints. Okay, and the non-negative restrictions

x1 , x2 ≥0 , then as in the case of the previous example here also we consider the straight lines
2 x 1 + x2 =5, x1 −x2 =1 and x2 =2 and then obtain the Convex polygon. We locate the vertices of
the convex polygon okay and then find the value of Z at all these vertices and the value of Z for
what whichever vertex the value of Z is maximum that vertex gives us the optimal value that is
maximum value of Z.

(Refer Slide Time: 30:05)

So, let us see we have the lines 2 x 1 + x2 =5 this is this line is 2 x 1 + x2 =5

this one. Okay, and then we have x1 −x2 =2, x1 −x2 =1.So this is x1 −x2 =2this line. Okay, and
then we have x2 =2. So we have x2 =2. This line is x2 =2this one. Okay. Now, let us see the
region which is this which satisfied the point x1 , x2 which satisfy all the three inequations so that
region is this one shown as the by dots here. Okay this this this region, okay, because

x2 ≤ 2. Ok 2 x 1 + x2 ≤5 , okay, and then x1 −x2 ≤ 1. Okay. So this is the region.

Alright, now, so we will have let us see which are the vertices here this point. Okay this point is
0,2 because here this is one of the vertex of the polygon region. There are five vertices 1 2 3 4 5
OK. This is 0, 0 there is a there is 1, 0 because x 2 is 0 here on x 1 axis. Okay, x 1 axis x2 =0so
this can be obtained. This point can be obtained by putting x2 =0 here. So this is 1 0 and then this
point, this point is the intersection of x1 −x2 =1this line and this line 2 x 1 + x2 =5. So their
intersection point is 2,1 that is q and then x2 =2and 2 x 1 + x2 =5 that gives the intersection point R
3 by 22.

Okay, and then we have this as 0 2 because x2 =2 , x1 =0so now, so P=(0 ,1)okay 2Q=(2, 1)

3
then you have R= and we have S=(0 , 2) and O=( 0, 0 )Z is equal to, so (O P Q) O P Q. Okay
2
R S O that is the feasible region. Okay, the values of Z at these points are okay so values of Z
are. Let me see Z=2 x1 +3 x2 . Okay, Z=2 x1 +3 x2 . So x1 =1, x2=0 so this will be for this Z=2.

3
Okay for this Z=2.2+3.7 . Okay, and for this Z= .2 so that is 3 and then 3 into 2 so this is 9
2
okay for 3 by 2 2 and 0 2 Z=0 here and 2 here or 6 here and Z=0 here. Okay so this value is the
maximum of all these five values of Z.

Okay so and so optimum that is maximum value of Z at occurs at 3 by 2 2 this point. Okay. So
the maximum value of Z (is 3) is 9 and it occurs at the point 3 by 2 2 by if you use the corner

3
Point method, okay, so by Corner Point method x1 = , x 2=2 , Z=9.
2

Now, let us see Iso-cost or Iso-profit method so draw the line 2 x 1 +3 x 2=0 . Okay. So
2 x 1 +3 x 2=0 that is this line this line is 2 x 1 +3 x 2=0 and 2 x 1 +3 x 2=0 means 3 x2 =2 x1

x1 x2 −3
so so is minus 3 over 2. So slope of this line is this line.
x2 x1 2
Okay, now we go on drawing lines parallel to this line which passes through the origin with

−3
slope minus . So this line, then this line, this line, and we go on drawing lines. Okay till we
2
reach a line which passes through this vertex. Okay, which passes through this vertex this line
has maximum distance from the origin. Okay and is passing through this vertex. So this vertex is

3 −3
the one which will give us the maximum value of Z and that at this vertex as by into and
2 2
we have seen gives us 9.So maximum value of Z=9

(Refer Slide Time: 35:28)

x1 −3
So here we have just said the same here we have written = . Okay so yeah, so we have
x2 2

−2 −2
this slope is −2 y 3 not minus 1 3 by 2, slope is minus , slope is by Y point X so slope is
3 3

−2 −2
Okay, slope is . So slope will be minus okay, the line corresponding to Z=0. This is the
3 3
line corresponding to Z=0 Okay it is shown by dotted line through O which is parallel to the
Iso-profit line now drawing parallel lines away from the origin since it is a maximization
problem shown by dotted lines. We see that the farthest line from the origin passes through the

3
vertex R 2. Okay, if you draw lines parallel lines parallel to the line 2 x 1 +3 x 2=0 even further,
2
they will not (pass the) cross the feasible region.
Therefore, they are not considered so we do not have to draw them. So the farthest line from the

−3 3
origin passes through the vertex has the optimal solution is x1 = , x 2=2 and maximum value
2 2
of Z=9. So this is how we solve a linear programming problem when Z is a function of two
variables by using graphical method in and we can solve the problem by using Corner Point
method and by using Iso-cost or Iso-profit method. We have shown that the two in the case of
both the examples how to apply the corner Point method and the Iso-cost or Iso-profit method.
So thank you very much for your attention.