You are on page 1of 3

Art of adjacency property

A feasible region is

convex not concave.

A LP region is always

convex.

in O

actual feasible region

war program region will always be a convex region


Slack tells how far we are from the boundery line. LHS-RHS format. Slack +ve is correct side, -ve is wrong side.

X X2 XIE Y

2 2 Xlt Xz Y

O
y X so

1 3
x2 30

I 5

7 3
only those solutions are feasible that do

not love a negative value

Simplex does not work for Urs variables

Augmentation Iteration 01

Initialization

slack variable
expresses

over audio why these names the diff bun nets


lets

we ament inequalities into

preparing a
problem for simple equations through
there stock

method variables

Decision variables t slack variables no no if variables

Max 2 3 1 5 2

x 4

2 2 E 12

3x t 2 2 618
XI X2 70

47 3 4
2 t Xy 12
2
3 1 2 2 Xy 18

zero's justifaction over audio

when variable value zero we will be on one of the boundary lines

Important here x so 9 2 0 worse for objective function

09 5 0 issue
Xy feasibility
However t when pint to zero will a CPF everytime but not
x2 give
optimal

At 41 09 x 0
2
9 43 9
xy 62 7 not feasible

simplex always needs a start from a point chichis opt and only origin
op guarantees that once into feasible region simplex does not
go out

non basic variables x x2


basic X 3 Xu 1 5

o o 9,12118
standard proper gansian form
Z 3 1 54 0 Row O

ft X3 4 Now I

2 12
Aguated form complete
2
t Xy Row 2

3 1 2 2 Xy 18 Row 3

simplex works
XI x2 o Int required as with non ve van only

saves for
2
only max problems

A basic variable will love co efficient I m one equation but 0 in at


Others Allows us to value from pits
study a

By definition of non basic value is zero

By definition of gale form co elf of basic var will always be zero

You might also like