Scribd
Upload
655 views

What-If Sensitivity Analysis For Linear Programming

The document discusses what-if analysis or sensitivity analysis for linear programming models. It introduces a example problem for a company called RMC that produces two chemical products from three raw materials. The document then outlines how to analyze: changes in objective function coefficients, changes in right-hand side values of constraints, simultaneous changes, reduced costs, and pricing out new variables. It provides details on calculating and interpreting shadow prices, ranges of feasibility and optimality, and using the 100% rule for simultaneous changes.

Uploaded by

luckiestmermaid
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
655 views

What-If Sensitivity Analysis For Linear Programming

The document discusses what-if analysis or sensitivity analysis for linear programming models. It introduces a example problem for a company called RMC that produces two chemical products from three raw materials. The document then outlines how to analyze: changes in objective function coefficients, changes in right-hand side values of constraints, simultaneous changes, reduced costs, and pricing out new variables. It provides details on calculating and interpreting shadow prices, ranges of feasibility and optimality, and using the 100% rule for simultaneous changes.

Uploaded by

luckiestmermaid
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 33

What-if Analysis (sensitivity analysis) for Linear Programming

ADM2302 /Rim Jaber

Introduction
Assumption: The parameters of the model were known with certainty In Reality: the model parameters are simply estimates that are subject to change Example (RMC Problem)

ADM2302 /Rim Jaber

RMC, Inc. is a firm that produces chemical based products. In a particular process three raw materials are used to produce two products. The Material requirements per ton are: Product Fuel additive Solvent base Material 1 2/5 1/2 Material 2 0 1/5 Material 3 3/5 3/10

For the current production period RMC has available the following quantities of each raw material. Because of spoilage, any materials not used for current production must be discarded. Number of Tons Material Available for Production Material 1 20 Material 2 5 Material 3 21 If the contribution to the profit is $40 for each ton of fuel additive and $30 for each ton of solvent base, How many tons of each product should be produced in order to maximize the total contribution profit?
ADM2302 /Rim Jaber 3

Introduction (contd)
RMC Problem Formulation (refer also RMC_problem.xsl)

Max Z = $ 40 x1 + 30 x2
Subject to

2/5 x1 + 1/2 x2 20 1/5 x2 5 3/5 x1 + 3/10 x2 21 x1 0, x2 0


x1: number of tons of fuel additive that RMC produces 4 x2: number of tons of solvent base that RMC produces

Answer Report
Target Cell (Max) Cell Name $D$5 Profit Totals Original Value $0.00 Final Value $1,600.00

Adjustable Cells Cell Name Original Value $B$3 Solution Fuel Additive (x1) 0.00 $C$3 Solution Solvent Base (x2) 0.00

Final Value 25.00 20.00

Constraints Cell Name $D$8 Material 1 LHS $D$9 Material 2 LHS $D$10 Material 3 LHS

Cell Value Formula Status Slack 20.0 $D$8<=$F$8 Binding 0 4.0 $D$9<=$F$9 Not Binding 1 21.0 $D$10<=$F$10 Binding 0

ADM2302 /Rim Jaber

Sensitivity Report
Adjustable Cells Final Reduced Objective Allowable Allowable Cell Name Value Cost Coefficient Increase Decrease $B$3 Solution Fuel Additive (x1) 25.00 0.00 40 20 16 $C$3 Solution Solvent Base (x2) 20.00 0.00 30 20 10 Constraints Cell Name $D$8 Material 1 LHS $D$9 Material 2 LHS $D$10 Material 3 LHS Final Shadow Constraint Allowable Allowable Value Price R.H. Side Increase Decrease 20.0 33.3 20 1.5 6 4.0 0.0 5 1E+30 1 21.0 44.4 21 9 2.25

ADM2302 /Rim Jaber

Introduction (contd)
What_if Analysis = Sensitivity Analysis = Postoptimality Analysis = Parametric Programming = Optimality Analysis The analysis of the effect of parameter changes (coefficient of the LP problem) on the optimal solution Centered around What-if question?

ADM2302 /Rim Jaber

Outline
Changes in the Objective Functions Coefficients Changes in the Right-Hand Sides (RHS) Values (changes in constraint quantity values) Shadow Prices Range of Feasibility Simultaneous Changes in the Objective Function Coefficients Simultaneous Changes in the RHS Reduced cost Pricing out New variables
ADM2302 /Rim Jaber 8

Changes in the Objective Functions Coefficients


Definition:
c1: the objective function coefficient of x1 (c1= ) c2: the objective function coefficient of x2 (c2 = )

Purpose Sensitivity analysis is performed to determine the range for cj over which the current solution remain optimal
This can be directly obtained from Excels sensitivity report

ADM2302 /Rim Jaber

Changes in the Objective Functions Coefficients (contd)


Graphically
Changes in the objective function coefficients values do not affect the size of the feasible region

The range of optimality


The range of values over which a variables coefficient can change without causing a change in the optimal solution (x1= 25, x2= 20) is: 24 c1 60 (refer to the Excel Sensitivity analysis report) 20 c2 50 (refer to the Excel Sensitivity analysis report)
ADM2302 /Rim Jaber 10

Changes in the Objective Functions Coefficients (contd) What about the value of Z?
Zero value decision variable (unused activity)
The value of the objective function, Z, will not change.

Non Zero value decision variables


The value of Z, will change.

ADM2302 /Rim Jaber

11

Changes in the Objective Functions Coefficients (contd)

Question : What happen if the estimate of the unit profit of ONE of the RMCs chemical based product is inaccurate? Answer: The range of values is wide for both objective function coefficients. Thus we can still be confident that we have obtained the correct optimal solution.
ADM2302 /Rim Jaber 12

Changes in the Right-Hand Sides (RHS) values


= Changes in constraints quantity values = changes in resources = changes in the managerial policy decisions qi = the RHS value for constraint i
q1 = 20 , q2 = 5 and q3 = 21

Graphically
A change in qi usually affect the size of the feasible region and often the value of the optimal solution
ADM2302 /Rim Jaber 13

50

x2

40 30 20 10 0

(0, 25)

(18.75,25) Optimal Solution Point (25,20)

(0,0) 10
20

(35,0)
x1

30

40

50
14

ADM2302 /Rim Jaber

Changes in the Right-Hand Sides (RHS) values (contd)


Shadow Price
The value of an additional unit of resources (e.g. an additional 1 ton of Material 1 will increase my profit by how much? Is it worth it?)

How to find the value of the shadow price?


Increase the RHS by one in the spreadsheet, and solve Look at the sensitivity report
ADM2302 /Rim Jaber 15

Changes in the Right-Hand Sides (RHS) values (contd)


Shadow Prices (Definitions): Is the marginal profit of additional unit of resource The maximum price you are willing to pay for 1 unit of the resources Amount by which profit will decrease if the availability of the resource associated with the constraint is reduced by 1 unit.
ADM2302 /Rim Jaber 16

Changes in the Right-Hand Sides (RHS) values (contd)


Purpose
The effect on the optimal value if a small change were to be made in the RHS constraint (done for each constraint, one at a time). We want to find the range value of qi over which the shadow prices remain valid the range over which the solution remain feasible (x1 and x2 remain different than zero, however their values may change)
ADM2302 /Rim Jaber 17

Changes in the Right-Hand Sides (RHS) values (contd)


Range of Feasibility
The range of values over which the value of the right hand side (qi) can vary without causing a change in the corresponding shadow price value 14 <= q1 <= 21.5 4 <= q2 18.75 <= q3 <= 30
ADM2302 /Rim Jaber 18

Changes in the Right-Hand Sides (RHS) values (contd)


Changes in the right hand side of a binding constraint causes the value of the optimal solution to change. What about changes in the right hand side of a none binding constraint?

ADM2302 /Rim Jaber

19

Simultaneous Changes in the Objective Function Coefficients


The range of values discussed is based on the assumption that only one coefficient changes, and all the other coefficients remain the same (accurate) Purpose
determine, without solving the problem, whether optimal solution may change if certain changes occur simultaneously in the coefficients of the objective function (given that those changes are within the corresponding range)
ADM2302 /Rim Jaber 20

100 Percent Rule for Objective Function Coefficients


For all objective function coefficients that are changed, sum the percentage of the allowable increases and the allowable decreases represented by the changes. If the sum of the percentage changes does not exceed 100%, the optimal solution will not change

(Change / Allowable change)*100 100


ADM2302 /Rim Jaber 21

100 Percent Rule for Objective Function Coefficients


Example (optimal solution does not change)
c1: 40 50 and c2: 30 35 Percentage allowable increase for c1 = (50-40)/ (allowable increase for c1) % = (50-40)/20 (100) = 50 %
Percentage allowable increase for c2 = (35-30)/ (allowable increase for c2) % = (5)/20 (100) = 25 % Sum= 50% + 25% = 75% <100%
ADM2302 /Rim Jaber 22

100 Percent Rule for Objective Function Coefficients


Example: c1: 40 25 and c2: 30 40
Percentage allowable decrease for c1 = (40-25)/ (allowable decrease for c1) % = (15/16) (100) = 93.5 % Percentage allowable increase for c2 = (40-30)/ (allowable increase for c2) % = 10/20 (100) = 50 % Sum= 93.5 % + 50% = 143.5% >100% Since the 100% is not satisfied we must resolve the problem to determine if the optimal solution will change or not
ADM2302 /Rim Jaber 23

Simultaneous changes in the RHS


The 100 percent Rule:
For all right hand side that are changed (with their corresponding range of feasibility), sum percentages of allowable increases and allowable decreases. If the sum does not exceed 100% then the shadow prices will not change

Example:
Q1: 20 21 , 14 <= Q1 <=21.5 Q2: 5 4, 4 <= Q2 Q3: 21 24 , 18.75 <= Q3 <= 30
ADM2302 /Rim Jaber 24

100 Percent Rule for the RHS


Example solution:

Shadow prices remain valid, did not change (refer to the excel file)
ADM2302 /Rim Jaber 25

Reduced Cost
Reduced cost of an unused activity is:
The amount by which the profit contribution of an activity (objective function coefficient) needs to be increased before producing this activity the amount by which the profit will decrease if 1 unit of this activity is forced into the solution

ADM2302 /Rim Jaber

26

Multiple optimal solution


Zero in the final value and reduced cost column zero in the allowable increase or allowable decrease columns of the adjustable cell table

ADM2302 /Rim Jaber

27

Pricing-out New Variables


Information given in sensitivity report can be used to study impact of introduction of new decision variables (products). For example:
If problem is re-solved with a new product in model, will it be recommend that a new product be made? Or, will it be recommend that a new product not be made, and continue making same products (that is, Fuel additive and solvent base)?
ADM2302 /Rim Jaber 28

Pricing-out New Variables (contd) Suppose RMC, Inc. wants to produce another chemical based products ultra base (x3).
0.1 ton 0.1 ton 0.8 ton $35 for of material 1 of material 2 of material 3 each ton

ADM2302 /Rim Jaber

29

Pricing-out New Variables (contd) 1. Checking validity of the shadow price using 100% rule 2. Compute the opportunity cost: marginal worth of the resources that would be diverted from existing product + cost of making the product (it is 0 in this example) 3. Opportunity cost > profit contribution
Do not produce the product
ADM2302 /Rim Jaber 30

Pricing-out New Variables (contd)

ADM2302 /Rim Jaber

31

Final Value

Summary of Sensitivity Report for Changes in the Objective Function Coefficients


The value of the decision variables (changing cells) in the optimal solution.

Reduced Cost
Increase in the objective function value per unit increase in the value of a zero value variable (for small increases)may be interpreted as the shadow price for the nonnegativity constraint.

Objective Coefficient
The current value of the objective coefficient.

Allowable Increase/Decrease
Defines the range of the coefficients in the objective function for which the current solution (value of the decision variables 32 or changing cells in the optimal solution) will not change.

Summary of Sensitivity Report for Changes in the Right-Hand-Sides


Final Value
The usage of the resource (or level of benefit achieved) in the optimal solutionthe left-hand side of the constraint. The change in the value of the objective function per unit increase in the right-hand-side of the constraint (RHS): Z = (Shadow Price)(RHS) (Note: only valid if change is within the allowable rangesee below.) The current value of the right-hand-side of the constraint. Defines the range of values for the RHS for which the shadow price is valid and hence for which the new objective function value can be calculated. (NOT the range for which the current solution will33 not change.)

Shadow Price

Constraint R.H. Side


Allowable Increase/Decrease

You might also like