50

Chapter 4

Linear Programming: A Graphical

Approach

We will look at how to solve linear programming problems with two variables using
a graphical approach. We begin by looking at how to set up a linear programming
problem with two or more variables. Then we look at how to interpret systems of
linear inequalities with two variables in a graphical way. Finally, we look at how to
solve linear programming problems with two variables.

4.1 Modeling Linear Programming Problems

Exercise 4.1 (Modeling Linear Programming Problems)

1. Acme’s Party Service1 . Acme’s Party service gives birthday parties for $30 and
Halloween parties for $40. Birthday party requires 25 balloons and 20 noise-
makers, while Halloween party requires 15 balloons and 10 noisemakers. Only
250 balloons and 175 noisemakers are available. Acme would like to schedule 𝑥
birthday parties and 𝑦 Halloween parties to make as much money as possible.

(a) If Acme gives 4 birthday parties and 3 Halloween parties, it makes

𝑃 = $30 × 4 + $40 × 3 = 200 / 240 / 300 dollars
(b) If Acme gives 10 birthday parties and 10 Halloween parties, it makes
𝑃 = $30 × 10 + $40 × 10 = 500 / 600 / 700 dollars
(c) If Acme gives 𝑥 birthday parties and 𝑦 Halloween parties, it makes
i. 𝑃 = 𝑥 + 𝑦 dollars
ii. 𝑃 = 30𝑥 + 40𝑦 dollars
This is the objective function.
1
Spence and Vanden Eynden, 4, p 124, 1990.

51
52 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

(d) Since Acme needs 25 balloons for each of 𝑥 birthday parties and 15 balloons
for each of 𝑦 Halloween parties and, furthermore, total number of balloons
cannot exceed 250, this means
i. 25𝑥 + 15𝑦 ≥ 250
ii. 15𝑥 + 25𝑦 < 250
iii. 25𝑥 + 15𝑦 ≤ 250
iv. 25𝑥 + 15𝑦 < 250
This is a structural constraint.
(e) Also, Acme needs 20 noisemakers for each of 𝑥 birthday party and 10
noisemakers for each 𝑦 Halloween party and total number of noisemakers
cannot exceed 175, this means
i. 20𝑥 + 10𝑦 ≤ 175
ii. 10𝑥 + 10𝑦 < 175
iii. 20𝑥 + 10𝑦 > 175
iv. 10𝑥 + 20𝑦 ≤ 175
This is another structural constraint.
(f) Since impossible to have less than zero birthday parties or zero Halloween
parties, this means
i. 𝑥<0 and 𝑦≥0
ii. 𝑥≥0 and 𝑦<0
iii. 𝑥≤0 and 𝑦≤0
iv. 𝑥≥0 and 𝑦≥0
These are nonnegativity constraints.
(g) Linear programming (LP) problem for this situation is
i. LP Candidate 1
Maximize x + y
subject to 25x + 15y ≤ 250
20x + 10y ≤ 175
x ≥ 0
y ≥ 0
ii. LP Candidate 2
Minimize 30x + 40y
subject to 25x + 15y ≤ 250
20x + 10y ≤ 175
x ≥ 0
y ≥ 0
iii. LP Candidate 3
Section 1. Modeling Linear Programming Problems (LECTURE NOTES 4) 53

Maximize 30x + 40y

subject to 25x + 15y ≤ 250
20x + 10y ≤ 175
x ≥ 0
y ≥ 0

2. Trout and Perch2 Lake is stocked with trout and perch, which feed on two
types of food (A and B) that grow in lake at rates of 6000 units and 2400 units
per day, respectively. Each trout consumes 6 units of first food and 2 units of
second food per day, and each perch consumes 8 units of first food and 4 units
of second food per day. Average trout weighs 1 pound and average perch weighs
1.5 pounds. Lake is stocked so total weight of fish in lake is as great as possible.

(a) If there are 4 trout and 3 perch, total weight is

𝑊 = 1 × 4 + 1.5 × 3 = 6.5 / 7.5 / 8.5 pounds
(b) If lake is stocked with 𝑥 trout and 𝑦 perch, total weight of fish is
i. 𝑊 = 6000𝑥 + 2400𝑦 pounds
ii. 𝑊 = 6𝑥 + 8𝑦 pounds
iii. 𝑊 = 6𝑥 + 1.5𝑦 pounds
iv. 𝑊 = 𝑥 + 1.5𝑦 pounds
(c) Since each of 𝑥 trout consume 6 units of food A per day and each of 𝑦 perch
consume 8 units of food A per day and total amount of food A available
is 6000 units per day,
i. 6𝑥 + 8𝑦 ≤ 6000
ii. 6𝑥 + 8𝑦 ≥ 6000
iii. 8𝑥 + 6𝑦 ≥ 6000
iv. 8𝑥 + 6𝑦 ≤ 6000
(d) Since each of 𝑥 trout consume 2 units of food B per day and each of 𝑦 perch
consume 4 units of food B per day and total amount of food B available
is 2400 units per day,
i. 4𝑥 + 2𝑦 ≤ 2400
ii. 4𝑥 + 4𝑦 ≤ 2400
iii. 2𝑥 + 4𝑦 ≥ 2400
iv. 2𝑥 + 4𝑦 ≤ 2400
(e) Since impossible to have less than zero trout or zero perch,
i. 𝑥 < 0 and 𝑦 ≥ 0
ii. 𝑥 ≥ 0 and 𝑦 < 0
2
Grosnick, Chapter 3 Exam, March 1996.
54 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

iii. 𝑥 ≤ 0 and 𝑦 ≤ 0
iv. 𝑥 ≥ 0 and 𝑦 ≥ 0
(f) LP problem for this situation is
i. LP Candidate 1
Maximize x + 1.5y
subject to 6x + 8y ≤ 6000
2x + 4y ≤ 2400
x ≥ 0
y ≥ 0
ii. LP Candidate 2
Maximize x + 1.5y
subject to 6x + 8y ≥ 6000
2x + 4y ≤ 2400
x ≥ 0
y ≥ 0
iii. LP Candidate 3
Maximize x + 1.5y
subject to 6x + 8y ≤ 2400
2x + 4y ≤ 6000
x ≥ 0
y ≥ 0

3. Insurance. Westville insurance has a total of $25 million earmarked for home-
owner and business loans. Homeowner loans have a 10% annual rate of return,
whereas business loans yield a 12% annual rate of return. As a matter of policy,
total amount of homeowner loans will be greater than or equal to four times
total amount of business loans. State linear programming problem to determine
total amount of loans of each type in order to maximize returns.

(a) For 𝑥 dollar amount of homeowner loans and 𝑦 dollar amount of business
loans, total return is
i. 𝑅 = 25𝑥 + 𝑦 dollars
ii. 𝑅 = 0.12𝑥 + 0.10𝑦 dollars
iii. 𝑅 = 0.10𝑥 + 0.12𝑦 dollars
iv. 𝑅 = 0.90𝑥 + 0.88𝑦 dollars
(b) Since Westville insurance has $25 million earmarked for 𝑥 homeowner and
𝑦 business loans,
i. 𝑥 + 𝑦 ≥ 25, 000, 000
ii. 𝑥 + 𝑦 ≤ 25, 000, 000
iii. 𝑥 − 𝑦 ≤ 25, 000, 000
Section 1. Modeling Linear Programming Problems (LECTURE NOTES 4) 55

iv. 𝑥 − 𝑦 ≥ 25, 000, 000

(c) Since 𝑥 homeowner loans is greater than or equal to four times 𝑦 business
loans, 𝑥 ≥ 4𝑦,
i. 𝑥 − 4𝑦 ≤ 0
ii. 𝑥 − 4𝑦 ≥ 0
iii. 𝑥 + 4𝑦 ≥ 0.10
iv. 𝑥 + 4𝑦 ≤ 0.12
(d) Since impossible to have less than zero homeowner or business loans,
i. 𝑥<0 and 𝑦≥0
ii. 𝑥≥0 and 𝑦<0
iii. 𝑥≤0 and 𝑦≤0
iv. 𝑥≥0 and 𝑦≥0
(e) LP problem for this situation is
i. LP Candidate 1
Maximize x - y
subject to x + y ≤ 0.10
x - 4y ≤ 0.12
x ≥ 0
y ≥ 0
ii. LP Candidate 2
Minimize 0.10x + 0.12y
subject to x + y ≤ 25,000,000
x - 4y ≤ 0
x ≥ 0
y ≥ 0
iii. LP Candidate 3
Maximize 0.10x + 0.12y
subject to x + y ≤ 25,000,000
x - 4y ≥ 0
x ≥ 0
y ≥ 0
4. Loudspeaker Systems3 Shipping costs per loudspeaker system in dollars from
two plants of Zoom Loudspeaker Company to three warehouses is given below.

Warehouse
Plant A B C
I 20 8 10
II 12 22 18
3
Tan, Example 4, p 187, 1997.
56 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

Maximum output of plants I and II are 400 and 600 units per month. In order for
warehouses A, B and C to meet orders, they need to be shipped, at a minimum,
200, 300 and 400 units per month respectively. Letting 𝑥1 represent number
of loudspeaker systems shipped from plant I to warehouse A and 𝑥2 represent
number of loudspeaker systems shipped from plant I to warehouse B, and so
on. Information summarized in following table.

Warehouse
Plant A B C max plant output
I 𝑥1 𝑥2 𝑥3 400
II 𝑥4 𝑥5 𝑥6 600
min warehouse require 200 300 400

What is shipping schedule to minimize shipping costs?

(a) Total shipping costs of sending units from plants to warehouses is

i. 𝐶 = 𝑥1 + 𝑥2 + 𝑥3 + 𝑥4 + 𝑥5 + 𝑥6 dollars
ii. 𝐶 = 400𝑥1 + 600𝑥2 + 200𝑥3 + 300𝑥4 + 400𝑥5 + 400𝑥6 dollars
iii. 𝐶 = 10𝑥1 + 8𝑥2 + 10𝑥3 − 12𝑥4 + 22𝑥5 − 18𝑥6 dollars
iv. 𝐶 = 20𝑥1 + 8𝑥2 + 10𝑥3 + 12𝑥4 + 22𝑥5 + 18𝑥6 dollars
(b) Since plant I can only produce 400 units per month,
i. 𝑥4 + 𝑥5 + 𝑥6 ≤ 400
ii. 𝑥1 + 𝑥2 + 𝑥3 ≥ 400
iii. 𝑥1 + 𝑥2 ≤ 400
iv. 𝑥1 + 𝑥2 + 𝑥3 ≤ 400
(c) Since plant II can only produce 600 units per month,
i. 𝑥4 + 𝑥5 + 𝑥6 ≤ 600
ii. 𝑥1 + 𝑥2 + 𝑥3 ≥ 600
iii. 𝑥1 + 𝑥2 ≤ 600
iv. 𝑥1 + 𝑥2 + 𝑥3 ≤ 600
(d) Since warehouse A must be shipped minimum of 200 units per month,
i. 𝑥2 + 𝑥4 ≥ 200
ii. 𝑥2 + 𝑥5 ≥ 200
iii. 𝑥1 + 𝑥4 ≤ 200
iv. 𝑥1 + 𝑥4 ≥ 200
(e) Since warehouse B must be shipped minimum of 300 units per month,
i. 𝑥2 + 𝑥4 ≥ 300
ii. 𝑥2 + 𝑥5 ≥ 300
Section 2. Linear Inequalities in Two Variables (LECTURE NOTES 4) 57

iii. 𝑥3 + 𝑥6 ≥ 300
iv. 𝑥1 + 𝑥4 ≥ 300
(f) Since warehouse C must be shipped minimum of 400 units per month,
i. 𝑥2 + 𝑥4 ≥ 400
ii. 𝑥2 + 𝑥5 ≥ 400
iii. 𝑥3 + 𝑥6 ≥ 400
iv. 𝑥1 + 𝑥4 ≥ 400
(g) Since impossible to send less than zero loudspeakers from any plant to any
warehouse
i. 𝑥1 ≤ 0, 𝑥2 ≤ 0, 𝑥3 ≤ 0, 𝑥4 ≤ 0, 𝑥5 ≤ 0, 𝑥6 ≤ 0,
ii. 𝑥1 ≥ 0, 𝑥2 ≤ 0, 𝑥3 ≥ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0,
iii. 𝑥1 ≥ 0, 𝑥2 ≥ 0, 𝑥3 ≥ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0,
iv. 𝑥1 ≥ 0, 𝑥2 ≥ 0, 𝑥3 ≤ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0,
(h) Consequently, LP problem for this situation is
i. LP Candidate 1

Maximize 20𝑥1 + 8𝑥2 + 10𝑥3 + 12𝑥4 + 22𝑥5 + 18𝑥6

subject to 𝑥1 + 𝑥2 + 𝑥3 ≤ 6000
𝑥4 + 𝑥5 + 𝑥6 ≤ 400
𝑥1 + 𝑥4 ≥ 200
𝑥2 + 𝑥5 ≥ 300
𝑥3 + 𝑥6 ≥ 400
𝑥1 ≥ 0, 𝑥2 ≥ 0, 𝑥3 ≥ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0
ii. LP Candidate 2

Minimize 20𝑥1 + 8𝑥2 + 10𝑥3 + 12𝑥4 + 22𝑥5 + 18𝑥6

subject to 𝑥1 + 𝑥2 + 𝑥3 ≤ 400
𝑥4 + 𝑥5 + 𝑥6 ≤ 600
𝑥1 + 𝑥4 ≥ 200
𝑥2 + 𝑥5 ≥ 300
𝑥3 + 𝑥6 ≥ 400
𝑥1 ≥ 0, 𝑥2 ≥ 0, 𝑥3 ≥ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0
iii. LP Candidate 3

Maximize 20𝑥1 + 8𝑥2 + 10𝑥3 + 12𝑥4 + 22𝑥5 + 18𝑥6

subject to 𝑥1 + 𝑥2 + 𝑥3 ≤ 600
𝑥4 + 𝑥5 + 𝑥6 ≤ 400
𝑥1 + 𝑥4 ≤ 200
𝑥2 + 𝑥5 ≤ 300
𝑥3 + 𝑥6 ≤ 400
𝑥1 ≥ 0, 𝑥2 ≥ 0, 𝑥3 ≥ 0, 𝑥4 ≥ 0, 𝑥5 ≥ 0, 𝑥6 ≥ 0

4.2 Linear Inequalities in Two Variables

We look at graphical interpretation of inequalities.

Exercise 4.2 (Linear Inequalities in Two Variables)

1. Which way to shade: a first look.
58 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

y x=3
y y
x=3 x=3

x<3 x>3

(0,0) 3 3 x
x (0,0) (0,0) 3 x

(a) x = 3 (b) x < 3 (c) x > 3

Figure 4.1 (Inequalities related to 𝑥 = 3)

(a) Line 𝑥 = 3 is a horizontal / vertical line. See figure (a).

(b) Inequality 𝑥 < 3 is region left / right of line 𝑥 = 3. See figure (b).
Point (𝑥, 𝑦) = (0, 0) left of 𝑥 = 3, satisfies / violates inequality 𝑥 = 0 < 3.
(c) Inequality 𝑥 > 3 is region left / right of line 𝑥 = 3. See figure (c).
Point (𝑥, 𝑦) = (0, 0) left of 𝑥 = 3, satisfies / violates inequality 𝑥 > 3.
(d) Using A Test Point To Identify Inequality Region. True / False
i. Draw (graph) equality part of inequality.
ii. Determine if test point (often (0,0)) satisfies inequality or not.
iii. If test point does satisfy inequality, point must be inside inequality
region and so shade from equality towards test point. If test point
violates inequality, point must be outside inequality region and so
shade from equality away from test point.

2. Which way to shade: a second look.

y y y
y>5

y=5 y=5 y=5

5 5 5
(0,4) y<5

(0,0) x (0,0) x (0,0) x

(a) y = 5 (b) y < 5 (c) y > 5

Figure 4.2 (Inequalities related to 𝑦 = 5)

(a) Line 𝑦 = 5 is a horizontal / vertical line. See figure (a).

Press Y=, clear everything. Type 5 beside 𝑌1 =, then WINDOW -10 10 1 0 10 1 1, then GRAPH

(b) Point (𝑥, 𝑦) = (0, 0) below 𝑦 = 5, satisfies / violates inequality 𝑦 = 0 < 5.

So inequality 𝑦 < 5 is region below / above line 𝑦 = 5. See figure (b).
Press Y=, choose “shade down” symbol to left of 𝑌1 , then GRAPH
Section 2. Linear Inequalities in Two Variables (LECTURE NOTES 4) 59

(c) Point (𝑥, 𝑦) = (0, 4) below 𝑦 = 5, satisfies / violates inequality 𝑦 = 4 < 5.

So inequality 𝑦 < 5 is region below / above line 𝑦 = 5. See figure (b).
Comment: Any test point like (0, 4) can be used, although point (0, 0) often easiest to use.

(d) Point (𝑥, 𝑦) = (0, 0) below 𝑦 = 5, satisfies / violates inequality 𝑦 > 5.

So inequality 𝑦 > 5 is region below / above line 𝑦 = 5. See figure (c).
Press Y=, choose “shade up” symbol to left of 𝑌1 , then GRAPH

(e) Inequality 𝑦 ≥ 5 is region

below / below and equal / above / above and equal to line 𝑦 = 5.
(f) Inequality 𝑦 ≤ 5 is region
below / below and equal / above / above and equal to line 𝑦 = 5.

3. Which way to shade: another example.

y y
10 10

8 8

6 5x + 4y > 22 6
5x + 4y = 22 or y = 22/4 - (5/4) x
4 4
5x + 4y = 22
2
or y = 22/4 - (5/4) x 2

(0,0) (0,0) x
x 00
00 2 4 6 8 10 2 4 6 8 10

5x + 4y < 22

(a) 5x + 4y > 22 (b) 5x + 4y < 22

or y > 22/4 - (5/4) x or y < 22/4 - (5/4) x
22
Figure 4.3 (Inequalities related to 5𝑥 + 4𝑦 = 22 or equivalently 𝑦 = 4
− 54 𝑥)

(a) The line 5𝑥 + 4𝑦 = 22

intersects 𝑦-axis at 𝑦 = 22
4
4
/ 22 / 22
5
. Hint: when 𝑥 = 0, 5(0) + 4𝑦 = 22, so 𝑦 = ?
22
and intersects 𝑥-axis at 𝑥 = 4 / 22 / 22
4
5
. Hint: when 𝑦 = 0, 5𝑥 + 4(0) = 22, so 𝑥 = ?

(b) Point (𝑥, 𝑦) = (0, 0), below 5𝑥 + 4𝑦 = 22,

satisfies / violates inequality 5𝑥 + 4𝑦 > 22 because 5(0) + 4(0) = 0.
So inequality 5𝑥 + 4𝑦 > 22 is region, excluding (0, 0),
below / above line 5𝑥 + 4𝑦 = 22. See figure (a).
(c) Point (𝑥, 𝑦) = (0, 0), below 5𝑥 + 4𝑦 = 22,
satisfies / violates inequality 5(0) + 4(0) = 0 < 22.
So inequality 5𝑥 + 4𝑦 < 22 is region, including (0, 0),
below / above line 5𝑥 + 4𝑦 = 22. See figure (b).

4. Region 5𝑥 + 4𝑦 ≥ 22 and 3𝑥 + 7𝑦 ≤ 27.

60 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

y
10

_ 22,
5x + 4y > _ 22/4 - (5/4) x
y>
4 or
(2,3) _ 27
3x + 7y < _ 27/7 - (3/7) x
y<
C
2

(0,0)
0 x
0 2 4 6 8 10
B
A

Figure 4.4 (Region 5𝑥 + 4𝑦 ≥ 22 and 3𝑥 + 7𝑦 ≤ 27)

(a) Point (𝑥, 𝑦) = (0, 0), below 5𝑥 + 4𝑦 = 22,
satisfies / violates inequality 5𝑥 + 4𝑦 ≥ 22 because 5(0) + 4(0) = 0.
So inequality 5𝑥 + 4𝑦 ≥ 22 is region, excluding (0, 0),
above / above and equal to line 5𝑥 + 4𝑦 = 22. See figure.
(b) Point (𝑥, 𝑦) = (0, 0), below 3𝑥 + 7𝑦 = 27,
satisfies / violates inequality 3(0) + 7(0) = 0 ≤ 27.
So inequality 3𝑥 + 7𝑦 ≤ 27 is region, including (0, 0),
below / below and equal to line 3𝑥 + 7𝑦 = 27. See figure.
(c) Inequality 5𝑥 + 4𝑦 ≥ 22 and 3𝑥 + 7𝑦 ≤ 27 is region A / B / C in figure.
(d) Region B is
i. feasible and bounded
ii. feasible and unbounded
iii. unfeasible and bounded
iv. unfeasible and unbounded
(e) Intersection point of equalities, 5𝑥 + 4𝑦 = 22 and 3𝑥 + 7𝑦 = 27, is
(1, 3) / (0, 3) / (2, 3) / (3, 2).
22
Let 𝑌1 = 4
− 54 𝑋 and 𝑌2 = 27
7
− 37 𝑋, then GRAPH, then 2nd CALC intersect.

5. Region 5𝑥 + 4𝑦 ≥ 22, 3𝑥 + 7𝑦 ≤ 27, 𝑦 ≥ 0 and 𝑥 ≥ 0.

y
10

5x + 4y >_ 22,
6
3x + 7y <_ 27,
4
x _> 0,
(2,3) y> _ 0
2

(22/5, 0) (27/3, 0)
0 x
0 2 4 6 8 10
Section 2. Linear Inequalities in Two Variables (LECTURE NOTES 4) 61

Figure 4.5 (Region 5𝑥 + 4𝑦 ≥ 22, 3𝑥 + 7𝑦 ≤ 27, 𝑦 ≥ 0 and 𝑥 ≥ 0)

(a) Region 5𝑥 + 4𝑦 ≥ 22, 3𝑥 + 7𝑦 ≤ 27, 𝑦 ≥ 0 and 𝑥 ≥ 0 is

i. feasible and bounded
ii. feasible and unbounded
iii. unfeasible and bounded
iv. unfeasible and unbounded
(b) The three vertices of feasible bounded region are
i. (2,3), ( 27
3
,1) and ( 22
5
,0)
27 22
ii. (2,3), ( 3 ,0) and ( 5 ,0)
iii. (2,3), (0, 27
3
) and ( 22
5
,0)
22
iv. (2,3), (5,1) and ( 5 ,0)
(c) One (unnecessary) inequality not bounding the feasible region is
i. 5𝑥 + 4𝑦 ≥ 22
ii. 3𝑥 + 7𝑦 ≤ 27
iii. 𝑦≥0
iv. 𝑥≥0

6. Choose system of linear inequalities which describes shaded region.

Figure 4.6 (Linear inequalities and shaded region)

(a) system A
𝑦 ≥4+𝑥
𝑦≥4
𝑦 ≤ 10 − 𝑥
(b) system B
𝑦 ≤4+𝑥
𝑦≥4
𝑦 ≥ 10 − 𝑥
62 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

(c) system C
𝑦 ≤4+𝑥
𝑦≥4
𝑦 ≤ 10 − 𝑥

7. Choose system of linear inequalities which describes shaded region.

y = 4 + 2x

y = 10 - x

y = 7 - (1/2)x

x
Figure 4.7 (Linear inequalities and shaded region)

(a) system A
𝑦 ≤ 4 + 2𝑥
𝑦 ≤ 10 − 𝑥
𝑦 ≥ 7 − 21 𝑥
(b) system B
𝑦 ≤ 4 + 2𝑥
𝑦 ≥ 10 − 𝑥
𝑦 ≥ 7 − 21 𝑥
(c) system C
𝑦 ≥ 4 + 2𝑥
𝑦 ≤ 10 − 𝑥
𝑦 ≤ 7 − 21 𝑥

8. Choose system of linear inequalities which describes shaded region.

Section 3. Solving Linear Programming Problems Graphically (LECTURE NOTES 4)63

y
y = 4 + 2x

y = 10 - x

y = 7 - (1/2)x

x
Figure 4.8 (Linear inequalities and shaded region)

(a) system A
𝑦 ≤ 4 + 2𝑥
𝑦 ≤ 10 − 𝑥
𝑦 ≥ 7 − 21 𝑥
(b) system B
𝑦 ≤ 4 + 2𝑥
𝑦 ≥ 10 − 𝑥
𝑦 ≥ 7 − 21 𝑥
(c) system C
𝑦 ≥ 4 + 2𝑥
𝑦 ≤ 10 − 𝑥
𝑦 ≤ 7 − 21 𝑥

4.3 Solving Linear Programming Problems

Graphically
Graphical method of solving linear programming problems involves identifying all
possible vertices of feasible set (corner points), calculating value of objective function
at these points, then choosing point which optimizes objective function.

Exercise 4.3 (Solving Linear Programming Problems Graphically)

1. Erasers and pen kits. L&R company sells Sally kits at 6 dollars a piece and
Tommy kits at 4 dollars a piece. Sally kit requires 1 eraser and 2 pens, while
Tommy kit requires 2 erasers and 1 pen. Only 8 erasers and 10 pen are left.
How many Sally kits (𝑥) and Tommy kits (𝑦) should be sold to maximize profit?
64 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

Maximize 6x + 4y
subject to x + 2y ≤ 8
2x + y ≤ 10
x ≥ 0
y ≥ 0

10
y
x + 2y _ 8,
<
8
2x + y <
_ 10,
2x + y = 10 x _> 0,
6 y _> 0

4
(0,4)
(4,2) optimal solution
2
x + 2y = 8
0 x
(0,0) 0 2 4 6 8 10
(5,0)

Figure 4.9 (Solving Linear Programming Problems Graphically)

Determine y-intercept corner point: when 𝑥 = 0, 𝑥 + 2𝑦 = 0 + 2𝑦 = 8, so 𝑦 = 4

and x-intercept corner point: when 𝑦 = 0, 2𝑥 + 𝑦 = 2𝑥 + 0 = 10, so 𝑥 = 5
and then: GRAPH Y1 = 4 − (1/2)𝑋 and Y2 = 10 − 2𝑥 using WINDOW 0 10 1 0 10 1 1 and
find intersection (corner) point using GRAPH TRACE, 2nd CALC intersect.

(a) Complete table: calculate value of objective function at four corners.

corner 𝑅 = 6𝑥 + 4𝑦
(0,0) 𝑅 = 6(0) + 4(0) = 0
(5,0) 𝑅 = 6(5) + 4(0) = 30
(4,2) 𝑅 = 6(4) + 4(2) = 32
(0,4) 𝑅 = 6(0) + 4(4) =
(b) Optimal (maximal) solution is at (𝑥, 𝑦) = (0, 4) / (5, 0) / (4, 2)
where value of objective function is a maximum of 0 / 16 / 30 / 32.

2. Acme’s party service. Acme would like to schedule 𝑥 birthday parties and 𝑦
Halloween parties to make as much money as possible, taking into account
limited number of balloons and noisemakers as given in following LP.

Maximize x + y
subject to 25x + 15y ≤ 250
20x + 10y ≤ 175
x ≥ 0
y ≥ 0
Section 3. Solving Linear Programming Problems Graphically (LECTURE NOTES 4)65

y
20
20x + 10y = 175
(0, 50/3)
16 20x + 10y <_ 175,
(5/2, 25/2) 25x + 15y <_ 250,
12 x >_ 0,
y >_0
8

4 25x + 15y = 250

0 x
0 2 4 6 8 10
(0,0)
(35/4, 0)

Figure 4.10 (Solving System of Linear Inequalities Graphically)

Determine x-intercept and y-intercept corner points, then

GRAPH Y1 = 250/15 − (25/15)𝑋 and Y2 = 175/10 − (20/10)𝑥 using WINDOW 0 10 1 0 20 1 1,
find intersection points using GRAPH TRACE, 2nd CALC intersect.

(a) Complete following table.

vertex 𝑧 =𝑥+𝑦
(0,0) 𝑧 =0+0=
(0, 503
) 𝑧 = 0 + 50
3
=
5 25 5 25
(2, 2 ) 𝑧 =2+ 2 =
( 35
4
, 0) 𝑧 =
In STAT EDIT, type 𝑥 coordinates into 𝐿1 , 𝑦 coordinates into 𝐿2 , then define 𝐿3 = 𝐿1 + 𝐿2 .

(b) Optimal (maximal) solution (𝑥, 𝑦) = (0, 0) / ( 35

4
, 0) / ( 25 , 25
2
) / (0, 50
3
)
where value of objective function is a maximum of 0 / 8.75 / 15 / 16.7.
(c) If objection function minimum (rather than maximum),
optimal solution (𝑥, 𝑦) = (0, 0) / ( 35
4
, 0) / ( 52 , 25
2
) / (0, 50
3
)
where value of objective function is a minimum of 0 / 8.75 / 15 / 16.7.
(d) True / False Any LP problem with nonempty bounded feasible region 𝑅
and objective function 𝑓 has both minimum at corner of 𝑅 and maximum
at (possible same, most likely different) corner of 𝑅.

3. Edge (rather than corner) of feasible region.

Maximize 6x + 3y
subject to x + 2y ≤ 8
2x + y ≤ 10
x ≥ 0
y ≥ 0
66 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

y
10

8
2x + y = 10
6

(0, 4)
4

(4, 2)
2
x + 2y = 8
0 x
0 2 4 6 8 10
(0,0)
(5, 0)

Figure 4.11 (Optimal Solution: Edge of Feasible Region)

GRAPH Y1 = 4 − (1/2)𝑋 and Y2 = 10 − 2𝑥 using WINDOW 0 10 1 0 10 1 1

and find intersection points using GRAPH TRACE, 2nd CALC intersect.

(a) Complete table.

corner 𝑧 = 6𝑥 + 3𝑦
(0, 0) 𝑧=
(0, 4) 𝑧=
(4, 2) 𝑧=
(5, 0) 𝑧=
In STAT EDIT, type 𝑥 coordinates into 𝐿1 , 𝑦 coordinates into 𝐿2 , then define 𝐿3 = 6𝐿1 + 3𝐿2 .

(b) Optimal solution is

i. either corner (𝑥, 𝑦) = (4, 2) or (𝑥, 𝑦) = (5, 0)
ii. edge of points between and including (𝑥, 𝑦) = (4, 2) and (𝑥, 𝑦) = (5, 0)
where objective function has value 0 / 12 / 30.
(c) If objection function is 𝑧 = 5𝑥 + 4𝑦 instead of 𝑧 = 6𝑥 + 3𝑦,
optimal (maximal) solution is at (𝑥, 𝑦) = (0, 0) / (0, 4) / (4, 2) / (5, 0)
where objective function has value 16 / 28 / 30.

4. Another problem.

Minimize 3x + 2y
subject to − 32 x + y ≤ 52
2
3
x + y ≤ 9
3
2
x + y ≥ 11
2
− 43 x + y ≥ −3
x ≥ 0
y ≥ 0
Section 3. Solving Linear Programming Problems Graphically (LECTURE NOTES 4)67

y = (3/2)x + 5/2 y = -(2/3)x +9

y = (4/3)x -3
y = -(3/2)x + 11/2

Figure 4.12 (Another graphical linear programming problem)

(a) Complete table.

corner 𝑧 = 3𝑥 + 2𝑦
(1, 4) 𝑧=
(3, 7) 𝑧=
(6, 5) 𝑧=
(3, 1) 𝑧=
In STAT EDIT, type 𝑥 coordinates into 𝐿1 , 𝑦 coordinates into 𝐿2 , then define 𝐿3 = 3𝐿1 + 2𝐿2 .

(b) Optimal (minimal) solution is

i. (𝑥, 𝑦) = (1, 4)
ii. (𝑥, 𝑦) = (3, 1)
iii. either corner (𝑥, 𝑦) = (1, 4) or (𝑥, 𝑦) = (3, 1)
iv. edge of points between and including (𝑥, 𝑦) = (1, 4) and (𝑥, 𝑦) = (3, 1)
where objective function has value 0 / 11 / 28.

5. And another problem.

y = 4 + 2x

R
y = 10 - x

y = 7 - (1/2)x

x
Figure 4.13 (Another graphical linear programming problem)
68 Chapter 4. Linear Programming: A Graphical Approach (LECTURE NOTES 4)

GRAPH Y1 = 4 + 2𝑋, Y2 = 10 − 𝑋 and Y3 = 7 − 0.5𝑋, then ZOOM ZoomFit ENTER TRACE,

find intersection points using 2nd CALC intersect.

(a) Corner points are (choose three)

(1.2, 6.4) / (2, 8) / (6, 4) / (7, 5) / (5, 10)
(b) Corner point which minimizes 𝑍 = 𝑥 − 5𝑦 on feasible set 𝑅 is
i. point (1.2, 6.4) with value -30.8.
ii. point (2,8) with value -38.
iii. point (6,4) with value -14.
iv. point (7,5) with value -18.
v. point (5,10) with value -45.
In STAT EDIT, type 𝑥 coordinates into 𝐿1 , 𝑦 coordinates into 𝐿2 , then define 𝐿3 = 𝐿1 − 5𝐿2 .

6. Last problem.

y
y = 4 + 2x

y = 10 - x

y = 7 - (1/2)x

x
Figure 4.14 (Another graphical linear programming problem)

GRAPH Y1 = 4 + 2𝑋, Y2 = 10 − 𝑋 and Y3 = 7 − 0.5𝑋, then ZOOM ZoomFit ENTER TRACE,

find intersection points using 2nd CALC intersect.

(a) Corner points are (choose two)

(1.2, 6.4) / (2, 8) / (6, 4) / (7, 5) / (5, 10)
(b) Corner point which minimizes 𝑍 = 4𝑥 + 3𝑦 on feasible set 𝑅 is
i. point (1.2, 6.4) with value 24.
ii. point (2,8) with value 32.
iii. point (6,4) with value 36.
iv. point (7,5) with value 43.
v. point (5,10) with value 50.
In STAT EDIT, type 𝑥 coordinates into 𝐿1 , 𝑦 coordinates into 𝐿2 , then define 𝐿3 = 4𝐿1 + 3𝐿2 .
Section 3. Solving Linear Programming Problems Graphically (LECTURE NOTES 4)69

(c) True / False. There is no maximum since feasible set 𝑅 is unbounded.

Optimal solution does not occur at a corner point in this case. Optimal
maximum and minimum solutions always occur for bounded regions but
do not necessarily occur for unbounded regions.