Higher Engineering Mathematics

Professor P. N. Agrawal
Department of Mathematics
Indian Institute of Technology Roorkee
Lecture - 47
Big M Method – II (Special Cases)

Hello friends, welcome to my second lecture on Big M Method, here we shall discuss some
special cases. Say for example, let us consider the following problem we shall see that this
problem has no feasible solution.

So, let us consider the L.P.P. maximize Z = - x 1 - x2 such that 3 x 1 + 2 x2 ≥ 30, - 2 x 1 + 3 x2 ≤ -

30, x 1 + x2 ≤ 5 and x 1 and x2 are non-negative.

(Refer Slide Time: 00:58)

Now, what we will do in order to make the right hand side in the second constraint positive?
Here, it is - 30, so in order to make it positive we multiply this in equality by - 1 and so this
reduces to 2 x 1 - 3 x2 ≥ 30, ok. So, we in order to make this right hand side of the second
constraint + positive value we multiplied by - 1.
(Refer Slide Time: 01:36)

So, we have multiplying the second constraint by - 1 to make the right hand side positive and
adding slack and surplus variables, so here what we will do? We have here 3 x 1 + 2 x2 ≥ 30,
so we add, so we subtract surplus variable here and here this is 2 x 1 - 3 x2 ≥ 30, here also we
subtract surplus variable and here we have x 1 + x2 ≤ 5.

So, we add slack variable to convert it into equality and what we will have then? Maximum
of Z = - x 1 - x2 , ok - x 1, - x2 ≥ 0 into s1≥ 0 into s2≥ 0 into s3 where s1 is surplus variable (in the
first equation) in the constraint we need surplus variable. So, we take surplus variable as s1,
so we get 3 x 1 + 2 x2 - s1 = 30 and then in the second constraint also we have constraint ≥ type,
so we use surplus variable s2.
So, 2 x 1 - 3 x2 - s2 = 30 and third constraint is ≤ type x 1 + x2 ≤ 5 we have, so we use slack
variable s3. So, we have x 1 + x2 + s3 = 5, so we need three variables s1, s2, s3; two variables s1
, s2 are surplus variables and the third variable s3 is a slack variable, ok. Now, let us write the
corresponding B.F.S. here, so we have here 1, 2, 3, 4, 5 variables, five variables are there, ok
and the equations are three.

So, we can take any two variables as zeros, so taking x 1 = 0, x2 = 0, ok what we get? s1 = -
30, and here we get s2 = - 30, and here we get s3 = 5, so s1, s2 are coming out to be negative
but the variables must all be non-negative in L.P.P. so we use 2 artificial variables A 1 and A 2,
ok. So, we write the using two artificial variables in order to get the initial starting B.F.S. ok
we write the given equations in the following form.

(Refer Slide Time: 03:57)

So, in order to get the starting B.F.S. we have to get two artificial variables A 1 and A 2 in the
first two constraint because in the third constraint when you take x 1, x2 = 0, s3 is = 5. So, we
do not have to do this is anything here s3 is already in a non-negative but in the case of first
and second equations s1, s2 are coming out to be negative, so we use artificial variables.

So, let us say artificial variables are A 1 and A 2 then we write the L.P.P. in the following form,
ok maximum of Z = - x 1 - x2 + 0 s1 +0 s2+ 0 s3 and then we use the negative price vector - M for
the artificial variables. So, - M A 1 - M A 2 and here we get 3 x 1 + 2 x2 - s1 + 0 s2 + 0 s3 + A 1 + 0
A 2 = 30. The second equation is 3x (min) 2x - 3x2 + 0s1 - s2 + 0s3 + (0s1) 0 A 1 + A 2 = 30 and
then third equation is x 1 + x2 , 0 s1, 0 s2, then s3, 0 A 1 + 0 A 2 = 5, ok.

Now, we have 1, 2, 3, 4, 5, 6, 7, ok seven constraints are there and the equations are three, ok
so we take four constraints, four variables as zeros. So, taking x 1 = 0, x2 = 0, s1 = 0, s2 = 0,
taking x 1 = 0, x2 = 0, s1 = 0, s2 = 0 the first equation gives us A 1 = 30, ok so A 1 = 30. Second
equation gives us what? We gave we get from the second equation x 1, x2 0 s1, s2 0 and we get
A 2 = 30, ok.

So, we get A 2 = 30 and third equation gives x 1, x2 0, s1, s2 0, so we get s3 = 5, so we get s3 =

5. So, thus we see that x 1, x2 , s1, s2, s3, A 1, A 2 are all non-negative and therefore it can be
taken as a starting basic feasible solution.
(Refer Slide Time: 06:15)

Now, let is consider First Simplex Table. So, cj’s are coefficients of the variables in the
objective function, so coefficients of x 1 is - 1, coefficient of x2 is - 1, coefficient of s1 is 0,
coefficient of s2 is 0, coefficient of s3 is 0, coefficient of A 1 is - M, coefficient of A 2 is - M,
ok basic vectors are A 1, A 2, s3 and the (non-basic variables are) basic variables are A 1, A 2, A
3 these are A 1, A 2, A 3; A 1, A 2, A 3 these are basis variables and non-basic variables are x 1,
x2 s1, s2, s3, ok x 1, x2 , s1, s2, ok x 1, x2 , s1, s2 they are non-basic variables, these are non-basic
variables, ok.

Now, in the initial simplex table the x 1 column contains the coefficient of x 1 here, so
coefficient of x 1 are 3, 2, 1, ok we get 3, 2, 1 here, coefficient of x2 are 2, - 3, 1 so we get 2, -
3, 1 here, coefficient of s1 is - 1, 0, 0, ok we get - 1, 0, 0 coefficient of s2 is 0, - 1, 0 so we get
0, - 1, 0 coefficient of s3 is 0, 0, 1, ok we get 0, 0, 1 and then coefficient of A 1 is 1, 0, 0
coefficient of A 2 is 0, 1, 0, ok so we get this and b column is the right hand side of this
constraints, so 3, 30, 30, 5, ok that is b column, ok.

Now, the coefficient of A 1 in the objective function is - M, coefficient of A 2 in the objective

function is - M and coefficient of s3 is 0, so C B column contains the coefficients of the basic
variables in the objective function. Now, we then find Z j star, Z j star = sigma C B a ij, so we
multiply C B column to the column x 1 and get Z j (sta) Z 1 star, this is j = 1 this is first
column, so j = 1 so Z 1 star will be - 3M - 2M and 0 into 1 is 0, so we get total - 5M, ok and
C j is (sma) capital C j is small c j - Z j star, so from this C j value we subtract Z j star and we
what it get is? 5M - 1.

Similarly, we can find (in) Z j star for the second column, so we had - M into 2 - 2M, - M into
- 3 is 3M, 3M - 2M is M, ok 0 into 1 is 0, so Z j star is M. So, then C 2 capital C 2 is small c 2
- Z 2 star, so - 1 - M, so we get - M - 1. Similarly, in the column corresponding to s1 (we) it
comes out, we get Z j, Z 3 star = M and therefore C capital C 3 is 0 - M, so we get - M here.

In the column s2, Z j star which is Z 4, j = 4 Z 4 star is M and so C 4 capital C 4 is - M and in

the (c) s3 column Z j star j = 5 is 0 and so 0 - 0 is 0, C 5 is 0 and in the column for A 1 we get
Z j star = - M, so - M - - M becomes 0 and here for A 2 also Z j star is - M, so - M - - M = 0,
so we get this.

Now, so 5M - 1, ok this is a positive value, this is negative, this is negative, this is negative,
ok C j is greater than 0 for j = 1, ok so this is key column, this is key column let us find key
row.. So divide the elements in the b column, ok by the corresponding elements in the x 1
column, so (6) 30 by 3, 30 by 3 gives you 10, then 30 by 2 gives you 15, then 5 divided by 1
gives you 5, so we get 5, ok.

Now, we have to find the minimum positive ratio here, so clearly 5 is the minimum positive
ratio, so this is key row. Now, key row at the intersection of key row and the key column we
get the key element, so this is our key element, key element is 1. Now, with the so, that
means that s3 will be outgoing variable, ok and x 1 will be incoming variable, so in the next
simplex table basis vectors will be basis will be cover the basic variables will be A 1, A 2, s1, s3
will replaced by x 1, ok coefficient of A 1 is - M in the objective function, so - M here,
coefficient of A 2 is - M here, coefficient of x 1 is - 1, so we put - 1 here and then we have x 1,
x2 , s1, s2, s3 A 1, A 2, b, ok.

Now, the key element, the key element is 1 if it is not 1, we will divide it so that it becomes 1
but it is already 1, ok so what we will do? With the help of this key row we will be making
the other elements in the x 1 column 0 that is 3 and 2 we have to make 0. So, let us first write
this key row we have 1, 1, 0, 0, 1, 0, 0, ok and we have 5 here, ok. Now, we multiply by 3
this key row and subtract from this first row this row, so we get 0 there, ok and we multiply it
by 3.

So, 3 we subtract from 2 we get - 1 and we 3 we multiply to 0 and subtract from - 1, so it

remains - 1, 3 into 0 is 0, we 2 0 - 0 is 0, then 3 into 1 is 3, so we get - 3 there and then we get
3 into 0 is 0 so we get 1 here, 3 into 0 is 0 so we get 0 here, ok we are multiplying by 3 and
subtracting, so 5 into 3 is 15, 15 when we subtract from 30 we get 15, ok.

Now, second row we multiply key this row by 2 and subtract from this row, so we get 0 here,
ok we are multiplying this row by 2 and subtracting from here, so 2 - 2 is 0 then we (sub)
multiply by 2 and subtract from here, so - 3 - 2 is - 5, ok. Then, we are multiplying by 2 and
subtracting, so 2 into 0 is 0, 0 - 0 is 0 then we 2 into 0 is 0, so it remains - 1, 2 into 1 is 2.

So, we are subtracting so - 2 there, ok and then 2 into 0 is 0, so it is remains 0, 2 into 0 is 0 so

it remains 1 and then 2 into 5 is 10, 10 one when we subtract from 30 we get 20, ok so this is
the next simplex table we can see.
(Refer Slide Time: 13:59)

C B is - M, - M, - 1 in the C B column, in the basis we have A 1, A 2 A 1, A 2, x 1 we can see A 1,

A 2 x 1, x 1 column is 0, 0, 1; x 1 column is 0, 0, 1 x2 column is - 1, - 5, 1; - 1, - 5, 1, s3 column
is, sorry s1 column is - 1, 0, 0 we get - 1, 0, 0, s2 column is 0, - 1, 0 0, - 1, 0 s3 column - 3, - 2,
1 this is - 3, - 2, 1, ok A 1 column is 1, 0, 0 1, 0, 0 A 2 column 0, 1, 0 so we get 1, 0, 0, 1, 0, 1,
0 and b column is 15, 20, 5, ok so we get 15, 20, 5, alright.

Now, we can find Z j star, so - M into 0 is 0, - M into 0 is 0, - 1 into 1 is - 1, so we get - 1, - 1

when we subtract from - 1 we get 0 and here we get - M into - 1 so + M, - M into - 5M so +
5M, 5M and M becomes 6 M, 6M - 1 we get. So, when we subtract 6M - 1 from - 1 we get -
6M, ok and then similarly in the s1 column Z j star becomes M, s2 column Z j star becomes
M, s3 column Z j star becomes 5M - 1, in the A 1 column it is - M, A 2 column it is - M, ok and
the corresponding capital C j values are - M, - M, - 5M, 0, 0.

Now, you can see that all C j values are less than (zer0) or = 0, this is 0, this is - 6M, so ≤0
less than or, so all C j values are ≤0.

(Refer Slide Time: 16:03)

So, what we have? Here no C j is positive hence the optimality condition is satisfied and
therefore this solution is optimal, ok and so x 1 is = 5, we can see x 1 is = 5, ok x2 = 0, s1 = 0, s2
= 0, s3 = 0 and A 1 = 15, A 2 = 20, ok. Now, so artificial vectors A 1, A 2 appear in the basis at
positive level we can see A 1, A 2 occur at positive level, ok there b values are positive, so they
are occur at the positive level which indicates that the given problem has no feasible solution,
ok.

(Refer Slide Time: 16:48)

Now, let us go to problem two, consider the following L.P.P. here we will have unbounded
solution, so maximum of Z we have to find Z = 3 x 1 + 2 x2 + x 3 and we are given the
constraints - 3 x 1 + 2 x2 + 2 x 3 = 8, - 3 x 1 + 4 x2 + x 3 = 7 ; x1, x2 , x 3 are non-negative.

(Refer Slide Time: 17:09)

So, we will use to select two artificial variables A 1, A 2 to convert this form to the standard
form, so Z = 3 x 1 + 2 x2 + x 3 - M A 1 - M A 2, ok and we have - 3 x 1 + 2 x2 + 2 x3 and then we have
A 1 + 0 A 2 = 8, ok and - 3x 1 + 4x2 + x3 +¿ 0 A 1 + A 2 = 7, ok so x 1, x2 , x 3 greater than or (eq) 0
and A 1, A 2 ≥0.

Now, we have two equations and 5 constraints, 5 variables, ok x 1, x2 , x 3 , A 1, A 2, so we take

any 3 variables = 0, so let us take the variables x 1, x2 , x3 = 0 each, ok we get the value of A 1
as 8 and the value of A 2 as 7, so this is our starting bsf B.F.S., ok x 1 = 0, x2 = 0, x 3 = 0, A 1 =
8, A 2 = 7 this is our starting basic feasible solution.

(Refer Slide Time: 18:13)

Now, let us write the first simplex table, so C j the coefficients of x 1, x2 , x 3 A 1, A 2 they are
given here 3, 2, 1, - M, - M, ok x 1, x2 , x 3 are our non-basic variables A 1, A 2 are basic
variables, so A 1, A 2 occur here and their coefficient in objective function are - M each, so we
put them they are here and the coefficients in the x 1 column, x 1 column the coefficients are -
3, - 3 in the x2 column we have 2, 4.

So, - 3 - 3 2, 4 and then we get x 3 column coefficients are 2, 1, ok so 2, 1 here A 1 coefficient

is 1 here it is 0, so 1, 0 and then A 2 column has 0, 1, ok so 1, 0 0, 1 and b column is the
equation I mean the right hand side of the equations, equation at 8 and 7, so we get 8 and 7
here.

Now, we can find Z j star, we can find Z j star so Z j star is = for j = 1, so - M into - 3 3M this
is 3M, 3M + 3M is 6M, so 3 - 6M is here, ok. Similarly for the x2 column - 2M - 4M so - 6M
we get 6M + 2 that is c 2, ok capital C 2. Now - M into 2 - 2M, - M into 1 - M, so we get -
3M and then C 3 capital C 3 is 3M + 1 and similarly C 4 is 0, C 5 is 0.

Now, we can see C j is positive in the second column and C j is positive in the third column,
ok but 6M + 2 is greater than 3M + 1, ok so this is our key column, ok divide the elements of
the b column by the corresponding values in the key column, so (si) 8 by 2 is 4 and 7 by 4 is
here, ok for this 7 divided by 4 we get 7 by 4, so out of this 4 and 7 by 4, 7 by 4 by 7 by 4 is
clearly minimum, so this is our key row, ok at the intersection of key row and key column we
get 4, so key element is 4, ok.
Now, so this means what? A 2 will be outgoing variable, x2 will be incoming variable and, ok
so we have C B basis, in basis we get basis (vec) basis (var) basic variable size A 1, x2 will be
replace by A 2 will be replace by x2 , coefficient of A 1 is - M, coefficient of x2 is 2, ok and we
get then x 1, x2 , x3 ; A 2 column will be deleted. In the next simplex table x 1, x2 , x 3 , A 1 and then
we have b, so ok.

Now, what we will divide? What we will do? We will divide by key element all the elements
of the key row, ok so we get - 3 by 4, ok 1 we get 1 by 4, we get 0, we get 7 by 4. Now,
having made this key element 1 unity, ok we then subtract suitable multiples of this from the
other rows to make the other entries in the key column zeros, ok. So, we have to make this
two zero, ok that means we multiplied by 2 and subtract from the first row, so what we will
get?

When you multiplied by 2 it becomes - 3 by 2, so - 3 by 2 we have to subtract from - 3, so we

get - 3 + 3 by 2, ok and then we have multiplying it by 2, ok and subtracting from here, so we
get 0 here, we are multiplying it by 2 so 1 by 2, 1 by 2 we subtract from 2 so 2 - 1 by 2 and
then we are multiplying it by 2 and subtracting from the first row so we get 1 here and we are
multiplying it by 2 so 7 by 2 we have to subtract from 8, so 8 - 7 by 2.

So, what we will have? This is - 3 + 3 by 2 it is - 3 by 2, ok and here we get 2 - half that is 3
by 2 and here we get 16 - 7, so 9 by 2. So let us see in the new simplex table we should have
these values, ok.

(Refer Slide Time: 22:59)

So, C B column you can see, see contents C B column is - M, 2, ok - M, 2 basis variables are
A 1, x2 . So, you can see A 1, x2 x 1 column we have - 3 by 2, - 3 by 4, so - 3 by 2, - 3 by 4, x2
column 0, 1 x3 column 3 by 2, 1 by 4 3 by 2, 1 by 4 A 1 column 1, 0 and b column, b column
is 9 by 2, 7 by 4 so we get 9 by 2 and 7 by 4, ok. Now, we go to determining the values of Z j
star for each j, ok.

So, you can see here - 3 M into - 3 by 2 is 3 by 2 3M by 2 and this is two times - 3 by 2, so -
3 by 2, ok so total is 3M by 2 - 3 by 2 and when we subtract this value from 3 we get - 3M +
9 by 2, ok. And similarly in this is for j = 1, for j = 2 we get c 2 small c 2 is 2 and Z 2 star, Z
2 star is 2 because - M 1 2 0 is 0, 2 into 1 2 is 2 so we get 0 here, C 3 capital C 3 is 3M + 1 by
2 and capital C 4 is 0, now clearly 3M + 1 by 2 is positive, ok so this is our key column, ok.

Now, let us find key row? So divide the b (valu) b column values by the column values of x3 ,
so 9 by 2 divided by 3 by 2 so we get 3, ok 7 by 4 divided by 1 by 4 gives you 7, so minimum
positive ratio is 3, so this is our key row, ok. So this key element is 3 by 2. Now, we have to
so and after we have located the key element which variable will be going out or A 1 will be
outgoing variable, ok and x3 will be incoming variable.

So, in the new simplex table we will have C B column, the basis we will have (bas) vector the
variables x, A 1 will be replaced by x 3 , ok so x3 , x2 , coefficient of x 3 is 1, coefficient of x2 is 2,
ok and x 1, x2 , x 3 A 1 will also be going out, so one column will be deleted and so we have x 1,
x2 , x 3 and b, ok. Now we divide the key row by 3 by 2 by the key element, ok so dividing by
3 by 2 it becomes - 1, 0, 1 and 1 over 3 by 2 means though this A 1 column is not there 3 by 2
we 9 by 2 divided 3 by 2 gives you 3, ok and then after we have made this element key
element 1 we have to make the element 1 by 4, 0 ok.

So, we (div) multiplied by 1 by 4 and subtract from second row, so this becomes 0 and we are
multiplying by 1 by 4 subtracting from second row, so - 3 by 4 and + 1 by 4 so - 2 by 4 so we
get - half and then we multiplied by 1 by 4 subtract from here, so with this remains 1, ok we
multiplied by 1 by 4 so we get 3 by 4, 3 by 4 we subtract from 4 7 by 4 and we get 1, ok.

(Refer Slide Time: 26:58)

So, let us see in the new simplex table we have C B 1, 2 C B column 1, 2 basis column has x3
, x2 so x3 , x2 x 1 column has - 1, - half so - 1, - half x2 column 0, 1 0, 1 x3 column 1, 0 1, 0 x 4
column 3, 1 so 3, 1, ok so we get 3, 1 here. Now let us find Z j star, so 1 into - 1 - 1, 2 into -
half - 1, - 1 - 1 - 2, ok 3 - 2 is 5, ok and here we get 1 into 0 0, 2 into 1 2 so we get 2, 2 - 2 is 0
and then 1 into 1 is 1, 2 into 0 is 0, so 1 we subtract from 1, ok so we get 0 here, ok Z j star is
1, ok.

And then what we notice? C j is positive in for j = 1, so this is our key column, ok so x 1 is
incoming vector, x 1 is incoming vector, ok but we cannot select the outgoing vector because
x 1 column has all elements negative all the elements of the x 1 column are negative which is
the incoming vector thus we cannot select the outgoing vector because to select the outgoing
vector we find the minimum positive ratio and what we will have here? this is key column, so
we divide the elements of b column by the elements of key column, so 3 divided by - 1 this
will be - 3, 1 divided by - half will be - 2, ok so no positive ratio is there, theta is negative in
the first row as well as in the second row. So we cannot find the outgoing vector and
therefore the solution is unbounded.

(Refer Slide Time: 28:51)

Now, let us consider the case where the problem contents constant term in the objective
function. So in L.P.P. when the objective functions contains a constant term then the simplex
method is applied by leaving this constant term in the beginning and optimal solution is
obtained. In the end, the constant term which was left initially is added to the optimal value
of the objective function.

So, let us say we have to maximize 2 x 1 - x2 + x 3 + 50 then with subject to the constraints 2 x 1
+ 2 x2 - 6 x 3 ≤16, 12 x 1 - 3 x2 + 3 x 3 ≥6, - 2 x 1 - 3 x2 + x 3 ≤5, x 1, x2 , x3 non negative then what we
will do? We will consider the corresponding problem where we will drop this 50, we will
simply consider the objective function 2 x 1 - x2 + x3 with these constraints and find the
optimal solution and in the optimal value of Z we will add 50.

(Refer Slide Time: 29:54)

So, let us see we have leaving the constant term 50 from the objective function in the
beginning, we introducing slack, surplus and artificial variables, the given problem reduces to
maximum of Z dash, we will call the new Z as Z dash because we are dropping 50. So,
maximum of Z dash 2 x 1 - x2 + x 3 , ok we have 2 x 1 - x2 + x 3 and then you see first equation is
first constraint is 2 x 1 + 2 x2 - 6 x 3 ≤ 16, so we use a slack variable s1 here, ok so 2 x 1 + 2 x2 - 6
x 3 , ok + s1, s s11 is a slack variable = 16.
Second equation is ≥ type, so we use a surplus variable, so 12 x 1 - 3 x2 + 3 x3 , ok - s2 and then
we will have to add a artificial variable able artificial variable A 1 in order to get the starting
B.F.S. so artificial variable A 1 we add and we give a large penalty - M to this, so - M A 1, ok.
And then third equation is - 2 x 1 - 3 x2 + x 3 we use a slack variable here s3, ok so - 2 x 1 - 3 x2 +
x 3 + s3 = 4.

Now, let us take x 1, x2 , x3 s1, s2 where x 1, x2 , x3 s1, s2, s3 A 1 are all non-negative, we have
three equations here, ok and we have 6 7 variables, ok 7 variables so we take four variables as
zeros, ok. So x 1 = 0, x2 = 0, x 3 = 0 and s2 = 0, ok because if you instead of s2 you take A 1 = 0
we will get s2 = - 6, ok so x 1 = 0, x2 = 0, x 3 = 0, s2 = 0 we take, ok then what we will get? s1
will come out to be 16, A 1 will be 6 and then s3 will be = 4.

So, x 1, x2 , x 3 , s1, s2, s3 and A 1 (A) all are non-negative, ok A 1 are all non-negative so we get
the starting basic feasible solution, ok. The solution of the problem by simplex method is
given in the (())(32:18) in the following table.

(Refer Slide Time: 32:19)

So, c j we have c j 2, - 1, 1, ok 2, - 1, 1, 0, 0, 0, - M basis vectors are s1, A 1 basis basic
variables are s1, A 1, s3; s1, A 1, s3 they are basic variables, ok non-basic variables are x 1, x2 , x 3
, ok x 1, x2 , x 3 and s2, ok they are non-basic variables and the coefficients of s1, A 1, s3 in the
objective function are 0, - M and 0, ok in the x 1 column we get 2, 12, - 2 2, 12 - 2 x2 column
has coefficient of x2 in the constraints.

So, 2, - 3, 2, - 3, - 3 x 3 column - 6, 3, 1 - 6, 3, 1 s1 column 1 there is no s1 here so 0, 1, 0 there

is no s1 here so 1, 0, 0, ok so 1, 0, 0 s2 column 0, - 1, 0, ok so 0, - 1, 0 s3 column 0, 0, 1, ok
we get 0, 0, 1, ok A 1 column 0, 1 and then we get 0 so 0, 1, 0, ok as b column 16, 6, 4 so we
get 16, 6, 4.

Now, we can find Z j star for j = 1, so 0 into 2 is 0 - M into 12 is - 12M and 0 into - 2 is 0 so
we get - 12M here, we subtract this Z 1 star from C 1, so 2 + 12M we get. Similarly C 2
capital C 2 comes out to be - 1, - 3M and C 3 comes out to be 1 + 3M and then it is 0 this is 0
- M so - M here, 0 - 0 is 0 here and then we get - M - - M that is - M + M that is 0, ok.

Now, we can see the M is very large, so 2 + 12M is the maximum C j which is positive, of
course C j is positive in this column also and the so C j is positive for j = 3 and j = 1 but C 1
value is more than C 3 value, so we will add this is our key column, ok divide so this means
that x 1 will be incoming variable. Let us now find (out) outgoing variable, so divide the
elements of b column by the elements of key column.

So, 16 divided by 2 is 8, 6 divided by 12 is half, ok this half and then (mul) 4 divided by - 2 is
- 2, so we have to consider minimum positive ratio, minimum positive ratio is this half, ok so
this is our key element intersection of key row and key column gives us the key element, so
key element is 12, ok. Now so this means A 1 will be outgoing variable x 1 is incoming
variable and A 1 is outgoing variable and in the next simplex table A 1 will not be there, the A 1
column will not be there, ok.

What we do now? So new simplex table will be C B, ok 0, (minu) A 1 is going out, A 1 is

replaced by x 1, so ok. So basis so s1 A 1 is replaced by x 1, coefficient of x 1is 2 here, ok and
then we have s3, s3 coefficient is 0 here, ok and we add x 1, so x 1 is now divide the elements of
key row, ok by the key element, so this means when you divide the key row by 12 the key
element 12 then what is we will get? 1, - 3 by 12.

So, - 1 by 4 and then 3 by 12, so 1 by 4 and then we get 0 and then we get - 1 by 12, ok this is
- 1 by 12 and then this is 0, ok A 1 column will not be there in the new simplex table, so 6
divided by 12 so that is half, ok this is the x 1, x2 , x3 , ok x 1, x2 , x 3 and then we get s1, s2, ok
and then we get s3 and we get here b vector, ok b column.

So, now we will have we will make with the help of this key element change to 1 will make
this 0, ok first row in the this element 2, 0 in the column (condu corre) corresponding to x 1,
so we multiplied by 2 subtract from there so we get 0 here, we multiplied by 2 we get - half, -
half we subtract from there so we get 2 + half, so 5 by 2 and we get multiply by 2 we get 1 by
2, 1 by 2 we subtract from - 6 so we get - 13 by 2, so we get here (min) x in the x3 column we
get - 13 by 2, ok.

And then s1 column, ok we are multiplying by 2, ok after dividing by 12, so we have we add
2 come here, ok so we so this will remain 1, this will remain 1, we multiplied by 2 we get - 1
by 6 which we subtract from this 0 so we get 1 by 6 here, ok and then we get s3 in the s3
column we have 0 here b multiplied by 2 subtract from 0 we get 0 here, we multiplied by 2
we get 1 when we subtract from 16 we get 15, ok and then this - 2 also has to be made 0, so
we multiplied by 2, ok we multiplied by 2 and added to the third row, so this becomes 0, we
multiplied by 2 so we get - half, - half we add 2 - 3, ok so - 3 - half we get - 7 by 2, so we get
- 7 by 2.

And we will have to now go to (thi) x 3 , x 3 means we are multiplying by 2 and adding, ok
adding to this row, ok so here we have 1 here we have what? 1 by 4, so 1 by 4 we multiply by
2 1 by 2 we get 1 by 2 we add to 1 we get 3 by 2, ok 1 we multiply here by 2 this is 0 when
we multiply by 2 so we get 0, 0 we add to 0 and we get 0 here then we multiplied by 2 here
we get - 1 by 6, - 1 by 6 we add to 0 we get - 1 by 6 here and in the s3 column we have 0 we
multiplied by 2 and add 2 to 0 so we get 1 here and here we get 1 by 2 we multiply by 2 we
get 1, 1 we add to 4 we get 5.

(Refer Slide Time: 40:23)

So, let us see in the new simplex table C B column has 0, 2, 0 so we get 0, 2, 0 s1 column
basis (variab) variables are s1, x 1, s3 so s1, x 1, s3, ok then the x 1 column has 0, 1, 0 values so
we get 0, 1, 0 values x2 column 5 by 2, - 1 by 4, - 7 by 2, ok so we see 5 by 2, - 1 by 4, - 7 by
2, ok so we match this x2 column elements and then x3 column - 13 by 2, 1 by 4, - 13 by 2, 1
by 4 and then 3 by 2, ok this, this, this.
And then s1 column 1, 0, 0 we get 1, 0, 0 s2 column 1 by 6 - 1 by 12, ok and - 1 by 6, so they
also match and then 0, 0, 1 in the s3 column, ok b column has 15, 1 by 12 this should be 6
divided by 12 this should be 1 by 2, ok so here 1 by 2 it should be, it should be 1 by 2, ok.
Now, alright and then the, ok so and this one is 5, this one is 5, ok so they match now. Now,
we can see we find Z j star, Z j star is 2 - half, half, 0, - 1, 0, ok and C j values are then 2 - 2 is
0, so 0 - half half, 0, 1 by 6, 0 there are two positive values 1 by 6, 1 by 2 of the capital C j, so
1 by 2 is greater than in the two, so this is our key column, ok.

Now, let us so this means x 3 is incoming variable, so x3 is incoming variable let us find the
outgoing variable. So we divide the elements of the b column by the values in the column x3 ,
so 15 divided by - 13 by 2 so we get a negative value of theta and then 1 by 2 divided by 1 by
4, ok so 1 by 2 divided by 1 by 4 is = 4 by 2 we get 2, so we get 2 here, ok and what we get?
(fif) 5 divided by 3 by 2 so we get 10 by 3, so there are two positive values of theta 2 and 13,
10 by 3, so 2 is minimum so this is our key row and then add the intersection of key row and
the key column we get the element 1 by 4, so key element is 1 by 4, ok 1 by 4.

Now, we divide, so this means what will happen x 3 is incoming variable, x 1 is outgoing
variable, ok so now the C B column will be having 0 basis variables will be s1, x 1 will be
replaced by x 3 and we will have s3, x 3 coefficient is 1, s3 coefficient is 0, ok and we have x 1,
x2 , x3 s1,s2, s3 and b, ok. The elements of the key row, key row are divided by key element so
we divide by 1 by 4, so this when we 1 is divided by 1 by 4 we get 4, ok so we get 4 here 1 -
1 by 4 divided by 1 by 4 gives you - 1 this is 1 and this is 0 divided by 1 by 4 is 0, - 1 by 12
divided by 1 by 4.

So, 4 by - 1 by 12 divided by 1 by 4, so we get - 1 by 3, so we get here - 1 by 3 and we get 0

divided by 1 by 4 is 0 and we divide 1 by 2 divided by 1 by 4, so we get 2 here, ok. Now we
have made this element 1, ok with the help of this 1, we make the elements of the element -
13 by 2 0, ok so we multiplied by 13 by 2, ok and add to the elements of the first row, so 13
by 2 4 into 13 by 2 gives you 26, ok so 13 by 2 we have multiplying to the to 1 in order to
make this 0 13 by 2 multiplying 13 by 2 into 4 is 26, so 26 we add to this one 0 so we get 26.

Then, we have - 1, - 1 into 13 by 2 so - 13 by 2, - 13 by 2 we add to 5 by 2, so - 8 by 2 so we

get - 4 so we get - 4 here, x3 this is 0, ok then we multiplied by 13 by 2 and add to the first
row, right. So s1 here it is 0 when we multiplying 13 by 2 this will remain 1, ok - 1 by 3 we
multiply by 13 by 2, so - 1 by 3 into 13 by 2 so what we get is? - 13 by 6, so - 13 by 6 we add
to 1 by 6 so 1 by 6 - 13 by 6 and we get - 12 by 6 which is - 2, ok so we get - 2 here.

And then here what we will have? 0 will be multiply by 13 by 2 and add to 0 it will remain 0,
we multiply 2 by 13 by 2, so 2 into 13 by 2 means 13, so 13 we add to 15 so 15 and we get
28. Let us see whether these this is values match, ok.

(Refer Slide Time: 47:09)

So, 26, - 4, 0 26, - 4, 0 1, - 2, 0 1, - 2, 0, 28, so 1, - 2, 0, 28 and then second row is 4, - 1, 1 so

4, - 1, 1 0, - 1 by 3 0, - 1 by 3, 0, 2 0, - 1 by 3, 0, 2, ok so here we have 4, 4 divided by 13 by
2 so 4 divided by 3 2, ok this is to be made 0, 4 into 3 by 2 is 6, ok so six we subtract from 0
and we get - 6 here, ok and then what we have - 1 - 1 into 3 by 2, so we get - 3 by 2 - 3 by 2
we subtract from - 7 by 2, so - 7 by 2 + 3 by 2 and we get here - 4 divided by 2 we get - 2,
this is 0, ok.

And then we are multiplying by 3 by 2, so 0 into 3 by 2 is 0, ok we subtract that from 0, ok so

we get 0 here and here what we have? - 1 by 3 we multiply by 3 by 2 so we get - half, - half
we subtract from - 1 by 6 so - 1 by 6 + half and this is = what? 6 - 1 + 3, so we get 2 by 6 and
that gives you 1 by 3, so we get 1 by 3 here, ok then in the s3 column, s3 column the element
is 0 we multiplied by 3 by 2, so it becomes 0 subtracted from 1, so we get 1 here.

And then 3 by 2 into 1 by 2 3 by 2 oh no it is 2 now 3 by 2 into 2 is 3, so 3 we subtract from 5

and we get 2 here, so let us see this row 6 in the next table. So - 6, - 2, 0, 0, - 6, - 2, 0, 0 1 by
3, 1, 2 1 by 3, 1, 2, ok . So now we have found Z j star here and the corresponding C j values,
so C j values are now - 2, 0, ok 0, 0, 1 by 3, 0 so this is there is still one C j value which is
positive, ok so this is our key column.

So, we so that means s2, s2 will be incoming vector incoming variable. Now let us find
outgoing variable, ok so we divide the b values, ok by the key column values, so 28 divided
by - 2 that is - 14, so it is a negative value and then 2 divided by - 1 by 3, so what we get is -
6, so that is also negative then 2 divided by 1 by 3 so that is 6, ok so this is minimum positive
ratio is this, so this is our key row and this is key element, so key element is 1 by 3, ok this
means that s3 will be outgoing variable, ok.

So, we will have now C B column as basis, so basis will be having a (vect) variables s1, ok
and x 3 , s3 will be going out in place of x 3 will shall have s2, ok yeah s1 coefficient is 0, s x 3
coefficient is 1, s2 coefficient is 0, ok. We will have x 1, x2 , x3 s1, s2, s3 and b, ok. Now we
divide the elements of the key row by key element, so we divide by 1 by 3, ok so dividing by
3 means - 6 into 3 that is - 18, so it will become - 18, ok.

And then - 2 divided by 1 by 3 will get - 6 then 0 here, 0 here, ok and this will become 1 here,
ok 1 divided 1 by 3 gives you 3 and then 6 divided by 1 by 3 gives you 18, ok and then we
can make with the help of this one, we can with the help of this one we can make the other
entries in the s2 column zeros, so this still - 2 and - 1 by 3 they have to be made zeros, ok.

So, we multiply this new row, new key row by 2 and add 2 - 2 there, ok so that it become 0,
ok the new row we will have here and then we multiplied by 1 by 3 and add to this second
row, so that it is - 1 by 3 becomes 0. So after that we get the following table.
(Refer Slide Time: 52:47)

So, we see s1, x3 , s2 we have s1, x3 , s2 we have 0, 1, 0 they are the values in the C B column,
ok C B column and this is third row, third row we have written here - 18, - 6, 0, 0, 1 so - 18, -
6, 0, 0, 1 and then we have 3 , 6 3 and we have oh this two becomes 6, oh no not 18 it is 6, ok
so we get match they are matching. Now, we can make these elements - 2 and - 1 by 3 0 by
multiplying this row suitably an adding, ok and so we will get the following table, ok.

This table we have and in this table we have found the values of Z j star and the
corresponding values of C j, C j is 4, 2, 0, 0, 0, - 1, so there are two positive values but the
value 4 is the maximum, so this is our key column, ok. So that means x 1 is incoming variable,
let us find outgoing variable but we notice that all the values in the x 1 column are negative,
ok - 10, - 2, - 18, ok.
So, to get the outgoing vector what we have to do is? We have to divide the elements in the b
column by the corresponding elements in key column, so this is - 40 divided by - 10, so that it
will be - 4 here, 4 divided by - 2 will be - 2 here, 6 divided by - 18 will be - 1 by 3 here, so
there is no positive ratio, ok and therefore we cannot find the outgoing vector, ok.

So, x 1 is entering vector but all elements in the column 1, ok they are negative, so we cannot
select the outgoing vector, ok and hence the solution of the problem is unbounded. So we
have discussed various cases which occur in the linear programming problems and we have
seen how we can solve them by using the Big M Method, we have the we have discuss in a in
detail the all the methods. So, that is all in the lecture, thank you very much for your
attention.