Calculus 7th Edition - Ch01
Calculus 7th Edition - Ch01
F UNCTIONS
8 Functions
In this section we will define and develop the concept of a function. Functions are
used by mathematicians and scientists to describe the relationships between variable
quantities and hence play a central role in calculus and its applications.
••••••••••••••••••••••••••••••••••••••
Many scientific laws are discovered by collecting, organizing, and analyzing experimental
SCATTER PLOTS AND TABULAR data. Since graphs play a major role in studying data, we will begin by discussing the kinds
DATA
of information that a graph can convey.
To start, we will focus on paired data. For example, Table 1.1.1 shows the top qualifying
speed by year in the Indianapolis 500 auto race from 1980 to 1999. This table pairs up each
Table 1.1.1 year t between 1980 and 1999 with the top qualifying speed S for that year. These paired
indianapolis 500 data can be represented graphically in a number of ways:
qualifying speeds
year t speed S • One possibility is to plot the paired data points in a rectangular tS-coordinate system
(mi/h) (t horizontal and S vertical), in which case we obtain a scatter plot of S versus t
1980 192.256 (Figure 1.1.1a).
1981 200.546
• A second possibility is to enhance the scatter plot visually by joining successive points
1982 207.004
1983 207.395
with straight-line segments, in which case we obtain a line graph (Figure 1.1.1b).
1984 210.029 • A third possibility is to represent the paired data by a bar graph (Figure 1.1.1c).
1985 212.583
1986 216.828 All three graphical representations reveal an upward trend in the data, as one would expect
1987 215.390 with improvements in automotive technology.
1988 219.198
1989 223.885 235 235
1990 225.301
225 225
Speed S (mi/h)
Speed S (mi/h)
1991 224.113
1992 232.482 215 215
1993 223.967
205 205
1994 228.011
1995 231.604 195 195
1996 233.100 185 185
1997 218.263 1980 1985 1990 1995 2000 1980 1985 1990 1995 2000
1998 223.503
Year t Year t
1999 225.179
(a) (b)
235
225
Speed S (mi/h)
215
205
195
185
1980 1985 1990 1995 2000
Year t
(c)
Figure 1.1.1
••••••••••••••••••••••••••••••••••••••
One of the first books to use graphs for representing numerical data was The Commercial
EXTRACTING INFORMATION FROM and Political Atlas, published in 1786 by the Scottish political economist William Play-
GRAPHS
fair (1759–1823). Figure 1.1.2a shows an engraving from that work that compares exports
and imports by England to Denmark and Norway (combined). In spite of its antiquity, the
January 12, 2001 11:10 g65-ch1 Sheet number 3 Page number 9 cyan magenta yellow black
engraving is modern in spirit and provides a wealth of information. You should be able to
extract the following information from Playfair’s graphs:
• In the year 1700 imports were valued at about 70,000 pounds and exports at about
35,000 pounds.
• During the period from 1700 to about 1754 imports exceeded exports (a trade deficit
for England).
• In the year 1754 the imports and exports were equal (a trade balance in today’s economic
terminology).
• From 1754 to 1780 exports exceeded imports (a trade surplus for England). The greatest
surplus occurred in 1780, at which time exports exceeded imports by about 95,000
pounds.
• During the period from 1700 to 1725 imports were rising. They peaked in 1725, and
then slowly fell until about 1760, at which time they bottomed out and began to rise
again slowly until 1780.
• During the period from 1760 to 1780 exports and imports were both rising, but exports
were rising more rapidly than imports, resulting in an ever-widening trade surplus for
England.
Fairness
3,000 Doctrine
Early reports
Great linking smoking Nonsmokers begin
Depression and cancer to demand rights
2,000
Rotating
First surgeon package
World War II general's report warnings
1,000
0
1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 2000
Playfair's Graph of 1786: The horizontal scale is in years from 1700 Source: U.S. Department of Health and Human Services.
to 1780 and the vertical scale is in units of 1,000 pounds sterling
from 0 to 200.
(a) (b)
Figure 1.1.2
Figure 1.1.2b is a more contemporary graph; it describes the per capita consumption of
cigarettes in the United States between 1925 and 1995.
•
•
•
•
•
•
When did the maximum annual cigarette consumption per adult occur and how many
•
•
•
•
•
were consumed?
•
•
•
•
•
•
•
What factors are likely to cause sharp decreases in cigarette consumption?
•
•
•
•
•
•
• What factors are likely to cause sharp increases in cigarette consumption?
•
•
•
•
•
•
•
•
What were the long- and short-term effects of the first surgeon general’s report on the
•
•
•
• health risks of smoking?
January 12, 2001 11:10 g65-ch1 Sheet number 4 Page number 10 cyan magenta yellow black
10 Functions
••••••••••••••••••••••••••••••••••••••
Graphs can be used to describe mathematical equations as well as physical data. For example,
GRAPHS OF EQUATIONS consider the equation
y = x 9 − x2 (1)
For each value of x in the interval −3 ≤ x ≤ 3, this equation produces a corresponding
real value of y, which is obtained by substituting the value of x into the right side of the
equation. Some typical values are shown in Table 1.1.2.
Table 1.1.2
x –3 –2 –1 0 1 2 3
y 0 –2 √ 5 ≈ –4.47214 –2 √ 2 ≈ –2.82843 0 2 √ 2 ≈ 2.82843 2 √ 5 ≈ 4.47214 0
y The set of all points in the xy-plane whose coordinates satisfy an equation in x and
6 y is called the graph of that equation in the xy-plane. Figure 1.1.3 shows the graph of
y = x √ 9 – x2
Equation (1) in the xy-plane. Notice that the graph extends only over the interval [−3, 3].
This is because values of x outside of this interval produce complex values of y, and in these
cases the ordered pairs (x, y) do not correspond to√ points in the
√ xy-plane. For example, if
x x =√ 8, then the corresponding value of y is y = 8 −55 = 8 55 i, and the ordered pair
-6 6 (8, 8 55 i) is not a point in the xy-plane.
Example 1 Figure 1.1.4 shows the graph of an unspecified equation that was used to
obtain the values that appear in the shaded parts of the accompanying tables. Examine the
-6 graph and confirm that the values in the tables are reasonable approximations.
Figure 1.1.3
y
3
2 x y x y
1
–3 0 None –3
–2 –1 –2.8, –2.3 –2
x –1 0.9 –2.9, –2, 2.4, 2.9 –1
0
0 0.7 –3, –1.7, 2.1, 3 0
-1 1 2 0.3, 1.8 1
2 0.4 1, 1.4 2
-2 3 0 None 3
-3
-3 -2 -1 0 1 2 3
Figure 1.1.4
••••••••••••••••••••••••••••••••••••••
Tables, graphs, and equations provide three methods for describing how one quantity de-
FUNCTIONS pends on another—numerical, visual, and algebraic. The fundamental importance of this
idea was recognized by Leibniz in 1673 when he coined the term function to describe the
dependence of one quantity on another. The following examples illustrate how this term is
used:
• The area A of a circle depends on its radius r by the equation A = πr 2 , so we say that
A is a function of r.
January 12, 2001 11:10 g65-ch1 Sheet number 5 Page number 11 cyan magenta yellow black
• The velocity v of a ball falling freely in the Earth’s gravitational field increases with
time t until it hits the ground, so we say that v is a function of t.
• In a bacteria culture, the number n of bacteria present after 1 hour of growth depends
on the number n0 of bacteria present initially, so we say that n is a function of n0 .
1.1.2 DEFINITION. A function f is a rule that associates a unique output with each
input. If the input is denoted by x, then the output is denoted by f(x) (read “f of x”).
225 • REMARK.
• In this definition the term unique means “exactly one.” Thus, a function cannot
Weight W (pounds)
•
200 •
•
•
175 •
•
•
assign two different outputs to the same input. For example, Figure 1.1.6 shows a scatter
•
150 •
•
• plot of weight versus age for a random sample of 100 college students. This scatter plot
•
125 •
•
100
•
•
•
•
does not describe the weight W as a function of the age A because there are some values of
•
75 •
•
•
•
A with more than one corresponding value of W . This is to be expected, since two people
50 •
•
•
•
•
with the same age need not have the same weight. In contrast, Table 1.1.1 describes S as
•
10 15 20 25 30 •
•
• a function of t because there is only one top qualifying speed in a given year; similarly,
•
•
∗
LEONHARD EULER (1707–1783). Euler was probably the most prolific mathematician who ever lived. It has
been said that “Euler wrote mathematics as effortlessly as most men breathe.” He was born in Basel, Switzerland,
and was the son of a Protestant minister who had himself studied mathematics. Euler’s genius developed early.
He attended the University of Basel, where by age 16 he obtained both a Bachelor of Arts degree and a Master’s
degree in philosophy. While at Basel, Euler had the good fortune to be tutored one day a week in mathematics by
a distinguished mathematician, Johann Bernoulli. At the urging of his father, Euler then began to study theology.
The lure of mathematics was too great, however, and by age 18 Euler had begun to do mathematical research.
Nevertheless, the influence of his father and his theological studies remained, and throughout his life Euler was
a deeply religious, unaffected person. At various times Euler taught at St. Petersburg Academy of Sciences (in
Russia), the University of Basel, and the Berlin Academy of Sciences. Euler’s energy and capacity for work were
virtually boundless. His collected works form more than 100 quarto-sized volumes and it is believed that much
of his work has been lost. What is particularly astonishing is that Euler was blind for the last 17 years of his life,
and this was one of his most productive periods! Euler’s flawless memory was phenomenal. Early in his life he
memorized the entire Aeneid by Virgil and at age 70 could not only recite the entire work, but could also state the
first and last sentence on each page of the book from which he memorized the work. His ability to solve problems
in his head was beyond belief. He worked out in his head major problems of lunar motion that baffled Isaac Newton
and once did a complicated calculation in his head to settle an argument between two students whose computations
differed in the fiftieth decimal place.
Following the development of calculus by Leibniz and Newton, results in mathematics developed rapidly in a
disorganized way. Euler’s genius gave coherence to the mathematical landscape. He was the first mathematician to
bring the full power of calculus to bear on problems from physics. He made major contributions to virtually every
branch of mathematics as well as to the theory of optics, planetary motion, electricity, magnetism, and general
mechanics.
January 12, 2001 11:10 g65-ch1 Sheet number 6 Page number 12 cyan magenta yellow black
12 Functions
••••••••••••••••••••••••••••••••••••••
Functions can be represented in various ways:
WAYS TO DESCRIBE FUNCTIONS
• Numerically by tables
• Geometrically by graphs
• Algebraically by formulas
• Verbally
The method of representation often depends on how the function arises. For example:
Time in minutes
0 10 20 30 40 50 60 70 80 t
Figure 1.1.7
Table 1.1.3 30 30
Population P (millions)
Population P (millions)
U.S. population 25 25
20 20
year t population P
(millions) 15 15
10 10
1790 3.9
1800 5.3 5 5
1810 7.2 1790 1810 1830 1850 1870 1790 1810 1830 1850 1870
1820 9.6
Year t Year t
1830 12
(a) (b)
1840 17
1850 23 Figure 1.1.8
Source: The World Almanac.
••••••••••••••••••••••••••••••••••••••
Engineers and physicists distinguish between continuous data and discrete data. Contin-
DISCRETE VERSUS CONTINUOUS uous data have values that vary continuously over an interval, whereas discrete data have
DATA
values that make discrete jumps. For example, for the seismic data in Figure 1.1.7 both
the time and intensity vary continuously, whereas in Table 1.1.3 and Figure 1.1.8a both
the year and population make discrete jumps. As a rule, continuous data lead to graphs
that are continuous, unbroken curves, whereas discrete data lead to scatter plots consisting
of isolated points. Sometimes, as in Figure 1.1.8b, it is desirable to approximate a scatter
plot by a continuous curve. This is useful for making conjectures about the values of the
quantities between the recorded data points.
••••••••••••••••••••••••••••••••••••••
Sometimes a function is buried in the statement of a problem, and it is up to the problem
GRAPHS AS PROBLEM-SOLVING solver to uncover it and use it in an appropriate way to solve the problem. Here is an example
TOOLS
that illustrates the power of graphical representations of functions as a problem-solving tool.
Example 2 Figure 1.1.9a shows an offshore oil well located at a point W that is 5 km from
the closest point A on a straight shoreline. Oil is to be piped from W to a shore point B that
is 8 km from A. It costs $1,000,000/km to lay pipe under water and $500,000/km over land.
In your role as project manager you receive three proposals for piping the oil from W to B.
Proposal 1 claims that it is cheapest to pipe directly from W to B, since the shortest distance
between two points is a straight line. Proposal 2 claims that it is cheapest to pipe directly
to point A and then along the shoreline to B, thereby using the least amount of expensive
underwater pipe. Proposal 3 claims that it is cheapest to compromise by piping under water
10
Cost c (millions of dollars)
9
8
W W 7
6
5 km 5 km 5
Shoreline 4
3
A B A P B 2
x 8–x 1
8 km 8 km 0 1 2 3 4 5 6 7 8
Distance x from point A (km)
(a) (b) (c)
Figure 1.1.9
January 12, 2001 11:10 g65-ch1 Sheet number 8 Page number 14 cyan magenta yellow black
14 Functions
to some well-chosen point between A and B, and then piping along the shoreline to B.
Which proposal is correct?
Solution. Let P be any point between A and B (Figure 1.1.9b), and let
x = distance (in kilometers) between A and P
c = cost (in millions of dollars) for the entire pipeline
Proposal 1 claims that x = 8 results in the least cost, Proposal 2 claims that it is x = 0, and
Proposal 3 claims it is some value of x between 0 and 8. From Figure 1.1.9b the length of
pipe along the shore is
8−x (3)
and from the Theorem of Pythagoras, the length of pipe under water is
x 2 + 25 (4)
Thus, from (3) and (4) the total cost c (in millions of dollars) for the pipeline is
c=1 x 2 + 25 + 0.5(8 − x) = x 2 + 25 + 0.5(8 − x) (5)
where 0 ≤ x ≤ 8. The graph of Equation (5), shown in Figure 1.1.9c, makes it clear that
Proposal 3 is correct—the most cost-effective strategy is to pipe to a point a little less than
3 km from point A.
1. Use the cigarette consumption graph in Figure 1.1.2b to an- rapidly during the first 2 years or the second 2 years of
swer the following questions, making reasonable approxi- that period? Explain your reasoning.
mations where needed.
median u.s. household income in
(a) When did the annual cigarette consumption reach 3000 thousands of constant 1995 dollars
per adult for the first time? $36
(b) When did the annual cigarette consumption per adult
$35
reach its peak, and what was the peak value?
$34
(c) Can you tell from the graph how many cigarettes were
consumed in a given year? If not, what additional infor- $33
mation would you need to make that determination? $32
(d) What factors are likely to cause a sharp increase in an- $31
nual cigarette consumption per adult? $30
1975 1979 1983 1987 1991 1995
(e) What factors are likely to cause a sharp decline in annual
cigarette consumption per adult? Source: Census Bureau, March 1996
[1996 measures 1995 income].
2. The accompanying graph shows the median income in U.S. Figure Ex-2
households (adjusted for inflation) between 1975 and 1995.
Use the graph to answer the following questions, making 3. Use the accompanying graph to answer the following ques-
reasonable approximations where needed. tions, making reasonable approximations were needed.
(a) When did the median income reach its maximum value, (a) For what values of x is y = 1?
and what was the median income when that occurred? (b) For what values of x is y = 3?
(b) When did the median income reach its minimum value, (c) For what values of y is x = 3?
and what was the median income when that occurred? (d) For what values of x is y ≤ 0?
(c) The median income was declining during the 4-year (e) What are the maximum and minimum values of y and
period between 1989 and 1993. Was it declining more for what values of x do they occur?
January 12, 2001 11:10 g65-ch1 Sheet number 9 Page number 15 cyan magenta yellow black
y
over a 1-week period, would you expect the graph of
3
the number of boxes on the shelf versus time to be a
2 continuous (unbroken) curve? Explain your reasoning.
9. A construction company wants to build a rectangular en-
1 closure with an area of 1000 square feet by fencing in three
x sides and using its office building as the fourth side. Your ob-
0 jective as supervising engineer is to design the enclosure so
that it uses the least amount of fencing. Proceed as follows.
-1
(a) Let x and y be the dimensions of the enclosure, where
x is measured parallel to the building, and let L be
-2
the length of fencing required for those dimensions.
-3
Since the area must be 1000 square feet, we must have
-3 -2 -1 0 1 2 3 xy = 1000. Find a formula for L in terms of x and y,
Figure Ex-3 and then express L in terms of x alone by using the area
equation.
(b) Are there any restrictions on the value of x? Explain.
4. Use the accompanying table to answer the questions posed
(c) Make a graph of L versus x over a reasonable interval,
in Exercise 3.
and use the graph to estimate the value of x that results
in the smallest value of L.
x –2 –1 0 2 3 4 5 6 (d) Estimate the smallest value of L.
y 5 1 –2 7 –1 1 0 9
10. A manufacturer constructs open boxes from sheets of card-
board that are 6 inches square by cutting small squares from
Table Ex-4 the corners and folding up the sides (as shown in the ac-
companying figure). The Research and Development De-
partment asks you to determine the size of the square that
5. Use the equation y = x 2 − 6x + 8 to answer the following
produces a box of greatest volume. Proceed as follows.
questions.
(a) Let x be the length of a side of the square to be cut,
(a) For what values of x is y = 0?
and let V be the volume of the resulting box. Show that
(b) For what values of x is y = −10?
V = x(6 − 2x)2 .
(c) For what values of x is y ≥ 0?
(d) Does y have a minimum value? A maximum value? If (b) Are there any restrictions on the value of x? Explain.
so, find them. (c) Make a graph of V versus x over an appropriate inter-
√ val, and use the graph to estimate the value of x that
6. Use the equation y = 1 + x to answer the following ques- results in the largest volume.
tions.
(d) Estimate the largest volume.
(a) For what values of x is y = 4?
(b) For what values of x is y = 0?
(c) For what values of x is y ≥ 6? x x
(d) Does y have a minimum value? A maximum value? If x x x
so, find them.
6 in
7. (a) If you had a device that could record the Earth’s pop- 6 – 2x
x x
ulation continuously, would you expect the graph of x x
population versus time to be a continuous (unbroken) 6 in 6 – 2x
curve? Explain what might cause breaks in the curve.
(b) Suppose that a hospital patient receives an injection of Figure Ex-10
an antibiotic every 8 hours and that between injections
the concentration C of the antibiotic in the bloodstream 11. A soup company wants to manufacture a can in the shape
decreases as the antibiotic is absorbed by the tissues. of a right circular cylinder that will hold 500 cm3 of liquid.
What might the graph of C versus the elapsed time t The material for the top and bottom costs 0.02 cent/cm2 ,
look like? and the material for the sides costs 0.01 cent/cm2 .
8. (a) If you had a device that could record the temperature (a) Use the method of Exercises 9 and 10 to estimate the ra-
of a room continuously over a 24-hour period, would dius r and height h of the can that costs the least to man-
you expect the graph of temperature versus time to be a ufacture. [Suggestion: Express the cost C in terms of r.]
continuous (unbroken) curve? Explain your reasoning. (b) Suppose that the tops and bottoms of radius r are
(b) If you had a computer that could track the number of punched out from square sheets with sides of length
boxes of cereal on the shelf of a market continuously 2r and the scraps are waste. If you allow for the cost of
January 12, 2001 11:10 g65-ch1 Sheet number 10 Page number 16 cyan magenta yellow black
16 Functions
the waste, would you expect the can of least cost to be football field and a straightaway. Make a graph of L
taller or shorter than the one in part (a)? Explain. versus x.
(c) Estimate the radius, height, and cost of the can in part (c) Use the graph to estimate the value of x that produces
(b), and determine whether your conjecture was correct. the shortest straightaways, and then find this value of x
12. The designer of a sports facility wants to put a quarter- exactly.
mile (1320 ft) running track around a football field, oriented (d) Use the graph to estimate the length of the longest pos-
as in the accompanying figure. The football field is 360 ft sible straightaways, and then find that length exactly.
long (including the end zones) and 160 ft wide. The track
consists of two straightaways and two semicircles, with the
straightaways extending at least the length of the football
field. 160′
(a) Show that it is possible to construct a quarter-mile track
around the football field. [Suggestion: Find the shortest
track that can be constructed around the field.]
(b) Let L be the length of a straightaway (in feet), and let 360′
x be the distance (in feet) between a sideline of the Figure Ex-12
In this section we will explore properties of functions in more detail. We will assume
that you are familiar with the standard notation for intervals and the basic properties
of absolute value. Reviews of these topics are provided in Appendices A and B.
••••••••••••••••••••••••••••••••••••••
Recall from the last section that a function f is a rule that associates a unique output f(x)
INDEPENDENT AND DEPENDENT with each input x. This output is sometimes called the value of f at x or the image of x
VARIABLES
under f . Sometimes we will want to denote the output by a single letter, say y, and write
y = f(x)
This equation expresses y as a function of x; the variable x is called the independent
variable (or argument) of f , and the variable y is called the dependent variable of f . This
terminology is intended to suggest that x is free to vary, but that once x has a specific value a
corresponding value of y is determined. For now we will only consider functions in which the
Table 1.2.1 independent and dependent variables are real numbers, in which case we say that f is a real-
valued function of a real variable. Later, we will consider other kinds of functions as well.
x 0 1 2 3 Table 1.2.1 can be viewed as a numerical representation of a function of f . For this
y 3 4 –1 6 function we have
f(0) = 3 f associates y = 3 with x = 0.
• REMARK.
• Although f, x, and y are the most common notations for functions and variables,
•
•
•
•
•
•
•
•
any letters can be used. For example, to indicate that the area A of a circle is a function
•
•
•
•
•
of the radius r, it would be more natural to write A = f(r) [where f(r) = πr 2 ]. Similarly,
•
•
•
•
•
to indicate that the circumference C of a circle is a function of the radius r, we might
write C = g(r) [where g(r) = 2πr]. The area function and the circumference function are
•
•
•
•
•
•
•
•
• different, which is why we denoted them by different letters, f and g.
••••••••••••••••••••••••••••••••••••••
If y = f(x), then the set of all possible inputs (x-values) is called the domain of f , and the
DOMAIN AND RANGE set of outputs (y-values) that result when x varies over the domain is called the range of f .
For example, consider the equations
y = x2 and y = x2, x≥2
In the first equation there is no restriction on x, so we may assume that any real value
of x is an allowable input. Thus, the equation defines a function f(x) = x 2 with domain
−⬁ < x < +⬁. In the second equation, the inequality x ≥ 2 restricts the allowable inputs
to be greater than or equal to 2, so the equation defines a function g(x) = x 2 , x ≥ 2 with
domain 2 ≤ x < +⬁.
As x varies over the domain of the function f(x) = x 2 , the values of y = x 2 vary over the
interval 0 ≤ y < +⬁, so this is the range of f . By comparison, as x varies over the domain
of the function g(x) = x 2 , x ≥ 2, the values of y = x 2 vary over the interval 4 ≤ y < +⬁,
so this is the range of g.
It is important to understand here that even though f(x) = x 2 and g(x) = x 2 , x ≥ 2
involve the same formula, we regard them to be different functions because they have
different domains. In short, to fully describe a function you must not only specify the rule
that relates the inputs and outputs, but you must also specify the domain, that is, the set of
allowable inputs.
••••••••••••••••••••••••••••••••••••••
If f is a real-valued function of a real variable, then the graph of f in the xy-plane is defined
GRAPHS OF FUNCTIONS to be the graph of the equation y = f(x). For example, the graph of the function f(x) = x
is the graph of the equation y = x, shown in Figure 1.2.1. That figure also shows the graphs
y y=x y y = x2 y y = x3
4 7 8
3 6 6
2 4
5
1 2
x x
0 4 0
-1 3 -2
-2 -4
2
-3 -6
-4 1 -8
-4 -3 -2 -1 0 1 2 3 4 x -8 -6 -4 -2 0 2 4 6 8
0
-1
-3 -2 -1 0 1 2 3
3
y y = 1/x y y = √x y y = √x
4 4 4
3 3 3
2 2
1 2
1 x x
0 0 1
x
-1 -1 0
-2 -2 -1
-3 -3
-4 -2
-4
-5-4-3-2-1 0 1 2 3 4 5 –1 0 1 2 3 4 5 6 7 8 9 -3
-4
-8 -6 -4 -2 0 2 4 6 8
Figure 1.2.1
January 12, 2001 11:10 g65-ch1 Sheet number 12 Page number 18 cyan magenta yellow black
18 Functions
of some other basic functions that may already be familiar to you. Later in this chapter we
will discuss techniques for graphing functions using graphing calculators and computers.
Graphs can provide useful visual information about a function. For example, because
the graph of a function f in the xy-plane consists of all points whose coordinates satisfy
the equation y = f(x), the points on the graph of f are of the form (x, f(x)); hence each
y-coordinate is the value of f at the x-coordinate (Figure 1.2.2a). Pictures of the domain
and range of f can be obtained by projecting the graph of f onto the coordinate axes
(Figure 1.2.2b). The values of x for which f(x) = 0 are the x-coordinates of the points
where the graph of f intersects the x-axis (Figure 1.2.2c); these values of x are called the
zeros of f , the roots of f(x) = 0, or the x-intercepts of y = f(x).
y y y
y = f (x)
y = f (x)
Range
(x, f (x))
f (x)
y = f (x)
x x x
x Domain x1 0 x2 x3
••••••••••••••••••••••••••••••••••••••
Not every curve in the xy-plane is the graph of a function. For example, consider the curve
THE VERTICAL LINE TEST in Figure 1.2.3, which is cut at two distinct points, (a, b) and (a, c), by a vertical line. This
curve cannot be the graph of y = f(x) for any function f ; otherwise, we would have
y
f(a) = b and f(a) = c
which is impossible, since f cannot assign two different values to a. Thus, there is no
function f whose graph is the given curve. This illustrates the following general result,
which we will call the vertical line test.
(a, c)
1.2.1 THE VERTICAL LINE TEST. A curve in the xy-plane is the graph of some function
(a, b) x
f if and only if no vertical line intersects the curve more than once.
a
Figure 1.2.3 Example 1 The graph of the equation
x 2 + y 2 = 25 (3)
is a circle of radius 5, centered at the origin (see Appendix D for a review of circles), and
hence there are vertical lines that cut the graph more than once. This can also be seen
algebraically by solving (3) for y in terms of x:
y = ± 25 − x 2
This equation does not define y as a function of x because the right side is “multiple valued”
in the sense that values of x in the interval (−5, 5) produce two corresponding values of y.
For example, if x = 4, then y = ±3, and hence (4, 3) and (4, −3) are two points on the
circle that lie on the same vertical line (Figure 1.2.4a). However, we can regard the circle
as the union of two semicircles:
y = 25 − x 2 and y = − 25 − x 2
(Figure 1.2.4b), each of which defines y as a function of x.
January 12, 2001 11:10 g65-ch1 Sheet number 13 Page number 19 cyan magenta yellow black
y y y
6 6 y = √ 25 – x 2 6 y = – √ 25 – x 2
(4, 3)
x x x
-6 6 -6 6 -6 6
(4, –3)
-6 -6 -6
(a) (b)
Figure 1.2.4
••••••••••••••••••••••••••••••••••••••
Recall that the absolute value or magnitude of a real number x is defined by
THE ABSOLUTE VALUE FUNCTION
x, x ≥ 0
|x| =
−x, x < 0
The effect of taking the absolute value of a number is to strip away the minus sign if the
number is negative and to leave the number unchanged if it is nonnegative. Thus,
|5| = 5, − 4 = 4 , |0| = 0
7 7
A more detailed discussion of the properties of absolute value is given in Appendix B.
However, for convenience we provide the following summary of its algebraic properties.
• REMARK.
• Symbols such as +x and −x are deceptive, since it is tempting to conclude
•
•
that +x is positive and −x is negative. However, this need not be so, since x itself can be
•
•
•
•
•
•
•
•
•
•
•
positive or negative. For example, if x is negative, say x = −3, then −x = 3 is positive
and +x = −3 is negative.
•
•
•
The graph of the function f(x) = |x| can be obtained by graphing the two parts of the
equation
y y = |x| x, x ≥ 0
5 y=
4 −x, x < 0
3 separately. For x ≥ 0, the graph of y = x is a ray of slope 1 with its endpoint at the origin,
2 and for x < 0, the graph of y = −x is a ray of slope −1 with its endpoint at the origin.
1 Combining the two parts produces the V-shaped graph in Figure 1.2.5.
x
0 Absolute values have important relationships to square roots. To see why this is so, recall
-1
from algebra that every positive real√number x has two square roots, one positive and one
-2
negative. By definition, the symbol x denotes
√ the positive square root of x. To denote the
-3
-5 -4 -3 -2 -1 0 1 2 3 4 5 negative
√ square root you must write − x.
√ For example, the positive square root of 9 is
Figure 1.2.5 √ 9 = 3, and the negative square root is − 9 = −3. (Do not make the mistake of writing
9 = ±3.) √
Care must 2
√ be exercised in simplifying expressions of the form x , since it is not always
true that x = x. This equation is correct if x is nonnegative, but it is false for negative x.
2
20 Functions
• FOR THE READER. Verify this relationship by using a graphing utility to show that the
•
•
•
•
√
•
•
•
• equations y = x 2 and y = |x| have the same graph.
••••••••••••••••••••••••••••••••••••••
The absolute value function f(x) = |x| is an example of a function that is defined piecewise
FUNCTIONS DEFINED PIECEWISE in the sense that the formula for f changes, depending on the value of x.
Example 2 Sketch the graph of the function defined piecewise by the formula
0, x ≤ −1
√
f(x) = 1 − x , −1 < x < 1
2
x, x≥1
y Solution. The formula for f changes at the points x = −1 and x = 1. (We call these the
2 breakpoints for the formula.) A good procedure for graphing functions defined piecewise is
to graph the function separately over the open intervals determined by the breakpoints, and
1 then graph f at the breakpoints themselves. For the function f in this example
√ the graph is
the horizontal ray y = 0 on the interval (−⬁, −1), it is the semicircle y = 1 − x 2 on the
x
interval (−1, 1), and it is the ray y = x on the interval (1, +⬁). The formula for f specifies
-2 -1 1 2
that the equation y = 0 applies at the breakpoint −1 [so y = f(−1) = 0], and it specifies
that the equation y = x applies at the breakpoint 1 [so y = f(1) = 1]. The graph of f is
Figure 1.2.6
shown in Figure 1.2.6.
• REMARK.
• In Figure 1.2.6 the solid dot and open circle at the breakpoint x = 1 serve
•
•
•
•
•
•
•
to emphasize that the point on the graph lies on the ray and not the semicircle. There is
•
•
•
•
•
•
no ambiguity at the breakpoint x = −1 because the two parts of the graph join together
•
•
• continuously there.
Example 3 Increasing the speed at which air moves over a person’s skin increases the rate
of moisture evaporation and makes the person feel cooler. (This is why we fan ourselves
in hot weather.) The windchill index is the temperature at a wind speed of 4 mi/h that
would produce the same sensation on exposed skin as the current temperature and wind
speed combination. An empirical formula (i.e., a formula based on experimental data) for
the windchill index W at 32 ◦ F for a wind speed of v mi/h is
32, 0 ≤ v ≤ 4
√
W = 91.4 + 59.4(0.0203v − 0.304 v − 0.474), 4 < v < 45
−3.8, v ≥ 45
A computer-generated graph of W(v) is shown in Figure 1.2.7.
Windchill Versus Wind Speed at 32°F
35
30
25
Windchill W (°F)
20
15
10
5
0
-5
-10
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
Figure 1.2.7 Wind speed v (mi/h)
January 12, 2001 11:10 g65-ch1 Sheet number 15 Page number 21 cyan magenta yellow black
••••••••••••••••••••••••••••••••••••••
Sometimes, restrictions on the allowable values of an independent variable result from a
THE NATURAL DOMAIN mathematical formula that defines the function. For example, if f(x) = 1/x, then √x =0
must be excluded from the domain to avoid division by zero, and if f(x) = x, then
negative values of x must be excluded from the domain, since we are only considering
real-valued functions of a real variable for now. We make the following definition.
Solution (a). The function f has real values for all real x, so its natural domain is the
interval (−⬁, +⬁).
Solution (b). The function f has real values for all real x, except x = 1 and x = 3, where
divisions by zero occur. Thus, the natural domain is
{x : x = 1 and x = 3} = (−⬁, 1) ∪ (1, 3) ∪ (3, +⬁)
Solution (c). Since f(x) = tan x = sin x / cos x, the function f has real values except
where cos x = 0, and this occurs when x is an odd integer multiple of π/2. Thus, the natural
domain consists of all real numbers except
π 3π 5π
x = ± ,± ,± ,...
2 2 2
Solution (d ). The function f has real values, except when the expression inside the radical
is negative. Thus the natural domain consists of all real numbers x such that
x 2 − 5x + 6 = (x − 3)(x − 2) ≥ 0
This inequality is satisfied if x ≤ 2 or x ≥ 3 (verify), so the natural domain of f is
(−⬁, 2] ∪ [3, +⬁)
• REMARK.
• In some problems we will want to limit the domain of a function by imposing
•
•
•
•
•
•
•
•
specific restrictions. For example, by writing
•
•
f(x) = x 2 , x≥0
•
•
•
•
•
•
•
•
•
•
• we can limit the domain of f to the positive x-axis (Figure 1.2.8).
y y
x x
y = x2 y = x 2, x ≥ 0
Figure 1.2.8
January 12, 2001 11:10 g65-ch1 Sheet number 16 Page number 22 cyan magenta yellow black
22 Functions
••••••••••••••••••••••••••••••••••••••
Algebraic expressions are frequently simplified by canceling common factors in the nu-
THE EFFECT OF ALGEBRAIC merator and denominator. However, care must be exercised when simplifying formulas for
OPERATIONS ON THE DOMAIN
functions in this way, since this process can alter the domain.
y y
6 6
5 5 x2 – 4
y = x+2 y=
4 4 x–2
3 3
2 2
1 x 1 x
-3-2-1 1 2 3 4 5 -3-2-1 1 2 3 4 5
Figure 1.2.9
Solution (b). The given function f is defined for all real x, except x = 1, so the natural
domain of f is
{x : x = 1} = (−⬁, 1) ∪ (1, +⬁)
To determine the range it will be convenient to introduce a dependent variable
x+1
y= (4)
x−1
Although the set of possible y-values is not immediately evident from this equation, the
graph of (4), which is shown in Figure 1.2.10b, suggests that the range of f consists of all
January 12, 2001 11:10 g65-ch1 Sheet number 17 Page number 23 cyan magenta yellow black
y y
y = 2 + √x – 1 y=
x+1
5 5 x–1
4 4
3 3
2
2
1 x 1
x
1 2 3 4 5 6 7 8 9 10
-3 -2 -1-1 1 2 3 4 5 6
-2
(a) (b)
Figure 1.2.10
••••••••••••••••••••••••••••••••••••••
In applications, physical considerations often impose restrictions on the domain and range
DOMAIN AND RANGE IN APPLIED of a function.
PROBLEMS
(a) Let V be the volume of the box that results when the squares have sides of length x.
Find a formula for V as a function of x.
(b) Find the domain of V .
(c) Use the graph of V given in Figure 1.2.11c to estimate the range of V .
(d) Describe in words what the graph tells you about the volume.
800
Volume V of box (in3 )
x x
x 700
x x
600
16 in 500
16 – 2x 400
x x 300
x x 200
30 in 30 – 2x 100
0 1 2 3 4 5 6 7 8 9
Side x of square cut (in)
(a) (b) (c)
Figure 1.2.11
January 12, 2001 11:10 g65-ch1 Sheet number 18 Page number 24 cyan magenta yellow black
24 Functions
Solution (a). As shown in Figure 1.2.11b, the resulting box has dimensions 16 − 2x by
30 − 2x by x, so the volume V (x) is given by
V (x) = (16 − 2x)(30 − 2x)x = 480x − 92x 2 + 4x 3
Solution (b). The domain is the set of x-values and the range is the set of V -values.
Because x is a length, it must be nonnegative, and because we cannot cut out squares whose
sides are more than 8 in long (why?), the x-values in the domain must satisfy
0≤x≤8
Solution (c). From the graph of V versus x in Figure 1.2.11c we estimate that the V -values
in the range satisfy
0 ≤ V ≤ 725
Note that this is an approximation. Later we will show how to find the range exactly.
Solution (d ). The graph tells us that the box of maximum volume occurs for a value of x
that is between 3 and 4 and that the maximum volume is approximately 725 in3 . Moreover,
the volume decreases toward zero as x gets closer to 0 or 8.
In applications involving time, formulas for functions are often expressed in terms of a
variable t whose starting value is taken to be t = 0.
Example 8 At 8:05 A.M. a car is clocked at 100 ft/s by a radar detector that is positioned
at the edge of a straight highway. Assuming that the car maintains a constant speed between
8:05 A.M. and 8:06 A.M., find a function D(t) that expresses the distance traveled by the car
Radar Tracking during that time interval as a function of the time t.
6000
Distance D (ft)
5000
4000 Solution. It would be clumsy to use clock time for the variable t, so let us agree to measure
3000 the elapsed time in seconds, starting with t = 0 at 8:05 A.M. and ending with t = 60 at
2000 8:06 A.M. At each instant, the distance traveled (in ft) is equal to the speed of the car (in
1000
ft/s) multiplied by the elapsed time (in s). Thus,
0 10 20 30 40 50 60
8:05 a.m. Time t (s) 8:06 a.m.
D(t) = 100t, 0 ≤ t ≤ 60
Figure 1.2.12 The graph of D versus t is shown in Figure 1.2.12.
••••••••••••••••••••••••••••••••••••••
In geometric problems where you want to preserve the “true” shape of a graph, you must
ISSUES OF SCALE AND UNITS use units of equal length on both axes. For example, if you graph a circle in a coordinate
system in which 1 unit in the y-direction is smaller than 1 unit in the x-direction, then the
y
circle will be squashed vertically into an elliptical shape (Figure 1.2.13). You must also use
units of equal length when you want to apply the distance formula
x
d = (x2 − x1 )2 + (y2 − y1 )2
to calculate the distance between two points (x1 , y1 ) and (x2 , y2 ) in the xy-plane.
However, sometimes it is inconvenient or impossible to display a graph using units of
equal length. For example, consider the equation
The circle is squashed because 1
unit on the y -axis has a smaller y = x2
length than 1 unit on the x -axis.
If we want to show the portion of the graph over the interval −3 ≤ x ≤ 3, then there is
Figure 1.2.13 no problem using units of equal length, since y only varies from 0 to 9 over that interval.
However, if we want to show the portion of the graph over the interval −10 ≤ x ≤ 10, then
there is a problem keeping the units equal in length, since the value of y varies between 0
and 100. In this case the only reasonable way to show all of the graph that occurs over the
interval −10 ≤ x ≤ 10 is to compress the unit of length along the y-axis, as illustrated in
Figure 1.2.14.
January 12, 2001 11:10 g65-ch1 Sheet number 19 Page number 25 cyan magenta yellow black
y y
9
8 100
7
80
6
5 60
4
3 40
2
20
1
x x
-3 -2 -1 1 2 3 -10 -5 5 10
Figure 1.2.14
• REMARK.
• In applications where the variables on the two axes have unrelated units (say,
•
•
•
•
•
•
•
centimeters on the y-axis and seconds on the x-axis), then nothing is gained by requiring
•
•
•
• the units to have equal lengths; choose the lengths to make the graph as clear as possible.
1
3. (a) f(x) = (b) g(x) = x2 − 3
x−3
x
(c) G(x) = x 2 − 2x + 5 (d) f(x) = (a) (b)
|x|
1 y y
(e) h(x) =
1 − sin x
1
4. (a) f(x) = (b) h(x) = x − 3x 2 x x
+7
5x
x2 − 4 x2 − 1
(c) G(x) = (d) f(x) =
x−4 x+1
3 (c) (d)
(e) h(x) =
2 − cos x Figure Ex-7
January 12, 2001 11:10 g65-ch1 Sheet number 20 Page number 26 cyan magenta yellow black
26 Functions
Figure Ex-8
Figure Ex-15
9. As shown in the accompanying figure, a pendulum of con-
stant length L makes an angle θ with its vertical position. 16. As shown in the accompanying figure, a camera is mounted
Express the height h as a function of the angle θ . at a point 3000 ft from the base of a rocket launching pad.
The shuttle rises vertically when launched, and the camera’s
elevation angle is continually adjusted to follow the bottom
of the rocket.
u
(a) Choose letters to represent the height of the rocket and
L
the elevation angle of the camera, and express the height
h as a function of the elevation angle.
(b) Find the natural domain and the range of the function,
Figure Ex-9 ignoring any physical restrictions on the values of the
variables.
(c) Modify the domain and range appropriately to account
10. A cup of hot coffee sits on a table. You pour in some cool
for the physical restrictions on the values of the vari-
milk and let it sit for an hour. Sketch a rough graph of the
ables.
temperature of the coffee as a function of time.
(d) Generate the graph of height versus the elevation on
11. A boat is bobbing up and down on some gentle waves. Sud- a graphing utility, and use it to estimate the height of
denly it gets hit by a large wave and sinks. Sketch a rough the rocket when the elevation angle is π/4 ≈ 0.7854
graph of the height of the boat above the ocean floor as a radian. Compare this estimate to the exact height. [Sug-
function of time. gestion: If you are using a graphing calculator, the trace
12. Make a rough sketch of your weight as a function of time and zoom features will be helpful here.
from birth to the present.
Rocket
In Exercises 13 and 14, express the function in piecewise
form without using absolute values. [Suggestion: It may help
to generate the graph of the function.]
x
13. (a) f(x) = |x| + 3x + 1 (b) g(x) = |x| + |x − 1|
14. (a) f(x) = 3 + |2x − 5| (b) g(x) = 3|x − 2| − |x + 1| u
15 inches, by cutting out squares with sides of length x from Figure Ex-16
each corner and bending up the sides.
(a) Express the volume V as a function of x. In Exercises 17 and 18: (i) Explain why the function f has
(b) Find the natural domain and the range of the function, one or more holes in its graph, and state the x-values at which
ignoring any physical restrictions on the values of the those holes occur. (ii) Find a function g whose graph is iden-
variables. tical to that of f, but without the holes.
(c) Modify the domain and range appropriately to account √
for the physical restrictions on the values of V and x. (x + 2)(x 2 − 1) x+ x
17. f(x) = 18. f(x) = √
(d) In words, describe how the volume V of the box varies (x + 2)(x − 1) x
with x, and discuss how one might construct boxes of 19. For a given outside temperature T and wind speed v, the
maximum volume and minimum volume. windchill index (WCI) is the equivalent temperature that
January 12, 2001 11:10 g65-ch1 Sheet number 21 Page number 27 cyan magenta yellow black
In this section we will discuss issues that relate to generating graphs of equations
and functions with graphing utilities (graphing calculators and computers). Because
graphing utilities vary widely, it is difficult to make general statements about them.
Therefore, at various places in this section we will ask you to refer to the documenta-
tion for your own graphing utility for specific details about the way it operates.
••••••••••••••••••••••••••••••••••••••
The development of new technology has significantly changed how and where mathemati-
GRAPHING CALCULATORS AND cians, engineers, and scientists perform their work, as well as their approach to problem
COMPUTER ALGEBRA SYSTEMS
solving. Not only have portable computers and handheld calculators with graphing capa-
bilities become standard tools in the scientific community, but there have been major new
innovations in computer software. Among the most significant of these innovations are
programs called Computer Algebra Systems (abbreviated CAS), the most common be-
∗
ing Mathematica, Maple, and Derive. Computer algebra systems not only have powerful
graphing capabilities, but, as their name suggests, they can perform many of the symbolic
computations that occur in algebra, calculus, and branches of higher mathematics. For
example, it is a trivial task for a CAS to perform the factorization
x 6 + 23x 5 + 147x 4 − 139x 3 − 3464x 2 − 2112x + 23040 = (x + 5)(x − 3)2 (x + 8)3
or the exact numerical computation
3
63456 43907 2251912457164208291259320230122866923
− =
3177295 22854377 382895955819369204449565945369203764688375
Technology has also made it possible to generate graphs of equations and functions in
seconds that in the past might have taken hours to produce. Graphing technology includes
handheld graphing calculators, computer algebra systems, and software designed for that
purpose. Figure 1.3.1 shows the graphs of the function f(x) = x 4 − x 3 − 2x 2 produced with
various graphing utilities; the first two were generated with the CAS programs, Mathematica
and Maple, and the third with a graphing calculator. Graphing calculators produce coarser
graphs than most computer programs but have the advantage of being compact and portable.
∗
Mathematica is a product of Wolfram Research, Inc.; Maple is a product of Waterloo Maple Software, Inc.; and
Derive is a product of Soft Warehouse, Inc.
January 12, 2001 11:10 g65-ch1 Sheet number 22 Page number 28 cyan magenta yellow black
28 Functions
4 4
3
2 y 2
1
-3 -2 -1 1 2 3 -3 -2 -1 1 2 3
-1
-2 -2 x
-3
-4 -4
Generated by Mathematica Generated by Maple Generated by a graphing calculator
Figure 1.3.1
••••••••••••••••••••••••••••••••••••••
Graphing utilities can only show a portion of the xy-plane in the viewing screen, so the
VIEWING WINDOWS first step in graphing an equation is to determine which rectangular portion of the xy-plane
you want to display. This region is called the viewing window (or viewing rectangle).
For example, in Figure 1.3.1 the viewing window extends over the interval [−3, 3] in the
x-direction and over the interval [−4, 4] in the y-direction, so we say that the viewing
window is [−3, 3] × [−4, 4] (read “[−3, 3] by [−4, 4]”). In general, if the viewing window
is [a, b] × [c, d], then the window extends between x = a and x = b in the x-direction
and between y = c and y = d in the y-direction. We will call [a, b] the x-interval for the
window and [c, d] the y-interval for the window (Figure 1.3.2).
(a, d) (b, d)
[c, d]
Figure 1.3.2
Different graphing utilities designate viewing windows in different ways. For example,
the first two graphs in Figure 1.3.1 were produced by the commands
and the last graph was produced on a graphing calculator by pressing the GRAPH button after
setting the following values for the variables that determine the x-interval and y-interval:
xMin = −3, xMax = 3, yMin = −4, yMax = 4
• FOR THE READER.
• Use your own graphing utility to generate the graph of the function
•
•
•
•
•
•
• f(x) = x 4 − x 3 − 2x 2 in the window [−3, 3] × [−4, 4].
••••••••••••••••••••••••••••••••••••••
To help locate points in a viewing window visually, graphing utilities provide methods for
TICK MARKS AND GRID LINES drawing tick marks (also called scale marks) on the coordinate axes or at other locations in
the viewing window. With computer programs such as Mathematica and Maple, there are
specific commands for designating the spacing between tick marks, but if the user does not
January 12, 2001 11:10 g65-ch1 Sheet number 23 Page number 29 cyan magenta yellow black
specify the spacing, then the programs make certain default choices. For example, in the
first two parts of Figure 1.3.1, the tick marks shown were the default choices.
On some graphing calculators the spacing between tick marks is determined by two scale
variables (also called scale factors), which we will denote by
xScl and yScl
(The notation varies among calculators.) These variables specify the spacing between the
tick marks in the x- and y-directions, respectively. For example, in the third part of Fig-
ure 1.3.1 the window and tick marks were designated by the settings
xMin = −3 xMax = 3
yMin = −4 yMax = 4
xScl = 1 yScl = 1
Most graphing utilities allow for variations in the design and positioning of tick marks.
For example, Figure 1.3.3 shows two variations of the graphs in Figure 1.3.1; the first was
generated on a computer using an option for placing the ticks and numbers on the edges of
a box, and the second was generated on a graphing calculator using an option for drawing
grid lines to simulate graph paper.
-2
-4
-3 -2 -1 0 1 2 3
Generated by Mathematica Generated by a graphing calculator
Figure 1.3.3
Example 1 Figure 1.3.4a shows the window [−5, 5] × [−5, 5] with the tick marks
spaced .5 unit apart in the x-direction and 10 units apart in the y-direction. Note that no
tick marks are actually visible in the y-direction because the tick mark at the origin is
covered by the x-axis, and all other tick marks in the y-direction fall outside of the viewing
window.
Example 2 Figure 1.3.4b shows the window [−10, 10] × [−10, 10] with the tick marks
spaced .1 unit apart in the x- and y-directions. In this case the tick marks are so close together
that they create the effect of thick lines on the coordinate axes. When this occurs you will
usually want to increase the scale factors to reduce the number of tick marks and make them
legible.
January 12, 2001 11:10 g65-ch1 Sheet number 24 Page number 30 cyan magenta yellow black
30 Functions
••••••••••••••••••••••••••••••••••••••
When the graph of a function extends indefinitely in some direction, no single viewing
CHOOSING A VIEWING WINDOW window can show the entire graph. In such cases the choice of the viewing window can
drastically affect one’s perception of how the graph looks. For example, Figure 1.3.5 shows
a computer-generated graph of y = 9 − x 2 , and Figure 1.3.6 shows four views of this graph
generated on a calculator.
• In part (a) the graph falls completely outside of the window, so the window is blank
(except for the ticks and axes).
• In part (b) the graph is broken into two pieces because it passes in and out of the window.
• In part (c) the graph appears to be a straight line because we have zoomed in on such a
small segment of the curve.
• In part (d ) we have a more complete picture of the graph shape because the window
encompasses all of the important points, namely the high point on the graph and the
intersections with the x-axis.
y = 9 – x2
Four views of y = 9 – x 2
For a function whose graph does not extend indefinitely in either the x- or y-directions,
the domain and range of the function can be used to obtain a viewing window that contains
the entire graph.
January 12, 2001 11:10 g65-ch1 Sheet number 25 Page number 31 cyan magenta yellow black
Example 4 Graph the equation y = x 3 −12x 2 +18 in the following windows and discuss
the advantages and disadvantages of each window.
Solution (a). The window in Figure 1.3.8a has chopped off the portion of the graph that
intersects the y-axis, and it shows only two of three possible real roots for the given cubic
polynomial. To remedy these problems we need to widen the window in both the x- and
y-directions.
Solution (b). The window in Figure 1.3.8b shows the intersection of the graph with the
y-axis and the three real roots, but it has chopped off the portion of the graph between
the two positive roots. Moreover, the ticks in the y-direction are nearly illegible because
they are so close together. We need to extend the window in the negative y-direction and
increase yScl. We do not know how far to extend the window, so some experimentation will
be required to obtain what we want.
Solution (c). The window in Figure 1.3.8c shows all of the main features of the graph.
However, we have some wasted space in the x-direction. We can improve the picture by
shortening the window in the x-direction appropriately.
Solution (d ). The window in Figure 1.3.8d shows all of the main features of the graph
without a lot of wasted space. However, the window does not provide a clear view of the
roots. To get a closer view of the roots we must forget about showing all of the main features
of the graph and choose windows that zoom in on the roots themselves.
Solution (e). The window in Figure 1.3.8e displays very little of the graph, but it clearly
shows that the root in the interval [1, 2] is slightly less than 1.3.
• FOR THE READER.
• Sometimes you will want to determine the viewing window by choos-
•
•
•
•
•
•
•
ing the x-interval for the window and allowing the graphing utility to determine a y-interval
•
•
•
•
•
•
that encompasses the maximum and minimum values of the function over the x-interval.
•
•
•
•
•
Most graphing utilities provide some method for doing this, so check your documentation to
•
•
•
•
•
determine how to use this feature. Allowing the graphing utility to determine the y-interval
•
•
•
•
•
of the window takes some of the guesswork out of problems like that in part (b) of the
•
•
•
• preceding example.
January 12, 2001 11:10 g65-ch1 Sheet number 26 Page number 32 cyan magenta yellow black
32 Functions
[–10, 10] × [–10, 10] [–20, 20] × [–20, 20] [–20, 20] × [–300, 20]
x Scl = 1, yScl = 1 x Scl = 1, yScl = 1 x Scl = 1, yScl = 20
(a) (b) (c)
••••••••••••••••••••••••••••••••••••••
The process of enlarging or reducing the size of a viewing window is called zooming. If you
ZOOMING reduce the size of the window, you see less of the graph as a whole, but more detail of the
part shown; this is called zooming in. In contrast, if you enlarge the size of the window, you
see more of the graph as a whole, but less detail of the part shown; this is called zooming
out. Most graphing calculators provide menu items for zooming in or zooming out by fixed
factors. For example, on one calculator the amount of enlargement or reduction is controlled
by setting values for two zoom factors, xFact and yFact. If
xFact = 10 and yFact = 5
then each time a zoom command is executed the viewing window is enlarged or reduced by a
factor of 10 in the x-direction and a factor of 5 in the y-direction. With computer programs
such as Mathematica and Maple, zooming is controlled by adjusting the x-interval and
y-interval directly; however, there are ways to automate this by programming.
••••••••••••••••••••••••••••••••••••••
Enlarging the viewing window for a graph has the geometric effect of compressing the
COMPRESSION graph, since more of the graph is packed into the calculator screen. If the compression is
sufficiently great, then some of the detail in the graph may be lost. Thus, the choice of the
viewing window frequently depends on whether you want to see more of the graph or more
of the detail. Figure 1.3.9 shows two views of the equation
y = x 5 (x − 2)
In part (a) of the figure the y-interval is very large, resulting in a vertical compression that
obscures the detail in the vicinity of the x-axis. In part (b) the y-interval is smaller, and
consequently we see more of the detail in the vicinity of the x-axis but less of the graph in
the y-direction.
January 12, 2001 11:10 g65-ch1 Sheet number 27 Page number 33 cyan magenta yellow black
Example 5 Describe the graph of the function f(x) = x + 0.01 sin 50πx; then graph
the function in the following windows and explain why the graphs do or do not differ from
your description.
(a) [−10, 10] × [−10, 10] (b) [−1, 1] × [−1, 1]
(c) [−.1, .1] × [−.1, .1] (d) [−.01, .01] × [−.01, .01]
[–5, 5] × [–1000, 1000] Solution. The formula for f is the sum of the function x (whose graph is a straight line)
x Scl = 1, yScl = 500 and the function 0.01 sin 50πx (whose graph is a sinusoidal curve with an amplitude of 0.01
(a) and a period of 2π/50π = 0.04). Intuitively, this suggests that the graph of f will follow the
general path of the line y = x but will have small bumps resulting from the contributions
of the sinusoidal oscillations.
∗
To generate the four graphs, we first set the calculator to the radian mode. Because
the windows in successive parts of this example are decreasing in size by a factor of 10,
it will be convenient to use the zoom in feature of the calculator with the zoom factors set
to 10 in the x- and y-directions. In Figure 1.3.10a the graph appears to be a straight line
because compression has hidden the small sinusoidal oscillations. (Keep in mind that the
amplitude of the sinusoidal portion of the function is only 0.01.) In part (b) the oscillations
[–5, 5] × [–10, 10] have begun to appear since the y-interval has been reduced, and in part (c) the oscillations
x Scl = 1, yScl = 1
have become very clear because the vertical scale is more in keeping with the amplitude of
(b) the oscillations. In part (d ) the graph appears to be a line segment because we have zoomed
Figure 1.3.9 in on such a small portion of the curve.
[–10, 10] × [–10, 10] [–1, 1] × [–1, 1] [–.1, .1] × [–.1, .1] [–.01, .01] × [–.01, .01]
x Scl = 1, yScl = 1 x Scl = .1, yScl = .1 x Scl = .01, yScl = .01 x Scl = .001, yScl = .001
(a) (b) (c) (d)
Figure 1.3.10
••••••••••••••••••••••••••••••••••••••
Figure 1.3.11a shows a circle of radius 5 and two perpendicular lines graphed in the window
ASPECT RATIO DISTORTION [−10, 10] × [−10, 10] with xScl = 1 and yScl = 1. However, the circle is distorted and
the lines do not appear perpendicular because the calculator has not used the same length
for 1 unit on the x-axis and 1 unit on the y-axis. (Compare the spacing between the ticks on
the axes.) This is called aspect ratio distortion. Many calculators provide a menu item for
automatically correcting the distortion by adjusting the viewing window appropriately. For
example, one calculator makes this correction to the viewing window [−10, 10] × [−10, 10]
by changing it to
[−16.9970674487, 16.9970674487] × [−10, 10]
(Figure 1.3.11b). With computer programs such as Mathematica and Maple, aspect ratio
distortion is controlled with adjustments to the physical dimensions of the viewing window
on the computer screen, rather than altering the x- and y-intervals of the viewing window.
∗
In this text we follow the convention that angles are measured in radians unless degree measure is specified.
January 12, 2001 11:10 g65-ch1 Sheet number 28 Page number 34 cyan magenta yellow black
34 Functions
••••••••••••••••••••••••••••••••••••••
Sometimes graphing utilities produce unexpected results. For example, Figure 1.3.12 shows
PIXELS AND RESOLUTION the graph of y = cos(10πx), which was generated on a graphing calculator in four differ-
ent windows. (Your own calculator may produce different results.) The first graph has the
correct shape, but the remaining three do not. To explain what is happening here we need
to understand more precisely how graphing utilities generate graphs.
[–1, 1] × [–1, 1] [–12.6, 12.6] × [–1, 1] [–12.5, 12.6] × [–1, 1] [–6, 6] × [–1, 1]
x Scl = .5, yScl = .5 x Scl = 1, yScl = .5 x Scl = 1, yScl = .5 x Scl = 1, yScl = .5
(a) (b) (c) (d)
Figure 1.3.12
Screen displays for graphing utilities are divided into rows and columns of rectangular
blocks, called pixels. For black-and-white displays each pixel has two possible states—an
63
activated (or dark) state and a deactivated (or light) state. Since graphical elements are
Pixels produced by activating pixels, the more pixels that a screen has to work with, the greater
the amount of detail it can show. For example, one calculator has a resolution of 63 × 127,
meaning that there are 63 rows with 127 pixels per row (Figure 1.3.13). In contrast, a
computer screen may have a resolution of 1024 × 1280 (1024 rows with 1280 pixels per
127 Pixels
row), so the computer screen is capable of displaying much smoother graphs than the
A viewing window with calculator.
resolution 63 × 127
(63 rows of 127 pixels) • FOR THE READER.
•
•
•
•
If you are using a graphing calculator, check the documentation to
•
Figure 1.3.13 •
•
• determine its resolution.
••••••••••••••••••••••••••••••••••••••
The procedure that a graphing utility follows to generate a graph is similar to the proce-
SAMPLING ERROR dure for plotting points by hand. When a viewing window is selected and an equation is
entered, the graphing utility determines the x-coordinates of certain pixels on the x-axis
and computes the corresponding points (x, y) on the graph. It then activates the pixels
whose coordinates most closely match those of the calculated points and uses some built-in
algorithm to activate additional intermediate pixels to create the curve shape. The point to
keep in mind here is that changing the window changes the points plotted by the graphing
utility. Thus, it is possible that a particular window will produce a false impression about
the graph shape because significant characteristics of the graph occur between the plotted
pixels. This is called sampling error. This is exactly what occurred in Figure 1.3.12 when
we graphed y = cos(10πx). In part (b) of the figure the plotted pixels happened to fall at
January 12, 2001 11:10 g65-ch1 Sheet number 29 Page number 35 cyan magenta yellow black
the peaks of the cosine curve, giving the false impression that the graph is a horizontal line
at y = 1. In part (c) the plotted pixels fell at successively higher points along the graph,
and in part (d ) the plotted pixels fell in a strange way that created yet another misleading
impression of the graph shape.
• REMARK.
• Figure 1.3.12 suggests that for trigonometric graphs with rapid oscillations,
•
•
•
•
•
•
•
•
restricting the x-interval to a few periods is likely to produce a more accurate representation
•
•
• about the graph shape.
••••••••••••••••••••••••••••••••••••••
Sometimes graphs that are continuous appear to have gaps when they are generated on a
FALSE GAPS calculator. These false gaps typically occur where the graph rises so rapidly that vertical
space is opened up between successive pixels.
√
Example 6 Figure 1.3.14 shows the graph of the semicircle y = 9 − x 2 in two viewing
windows. Although this semicircle has x-intercepts at the points x = ±3, part (a) of the
figure shows false gaps at those points because there are no pixels with x-coordinates ±3 in
the window selected. In part (b) no gaps occur because there are pixels with x-coordinates
x = ±3 in the window being used.
••••••••••••••••••••••••••••••••••••••
In addition to creating false gaps in continuous graphs, calculators can err in the opposite
FALSE LINE SEGMENTS direction by placing false line segments in the gaps of discontinuous curves.
Example 7 Figure 1.3.15a shows the graph of y = 1/(x − 1) in the default window on
a calculator. Although the graph appears to contain vertical line segments near x = 1, they
should not be there. There is actually a gap in the curve at x = 1, since a division by zero
occurs at that point (Figure 1.3.15b).
y
4
x
-5 5
(a) (b)
Figure 1.3.15
January 12, 2001 11:10 g65-ch1 Sheet number 30 Page number 36 cyan magenta yellow black
36 Functions
••••••••••••••••••••••••••••••••••••••
Most graphing utilities
√ use logarithms to evaluate functions with fractional exponents such as
ERRORS OF OMISSION 3
f(x) = x 2/3 = x 2 . However, because logarithms are only defined for positive numbers,
many (but not all) graphing utilities will omit portions of the graphs of functions with
fractional exponents. For example, one calculator graphs y = x 2/3 as in Figure 1.3.16a,
whereas the actual graph is as in Figure 1.3.16b. (See the discussion preceding Exercise 29
for a way of circumventing this problem.)
4
3
2
1
x
-4 -3 -2 -1 1 2 3 4
[–4, 4] × [–1, 4]
x Scl = 1, yScl = 1
(a) (b)
Figure 1.3.16
••••••••••••••••••••••••••••••••••••••
Although graphing utilities are powerful tools for generating graphs quickly, they can pro-
WHAT IS THE TRUE SHAPE OF A duce misleading graphs as a result of compression, sampling error, false gaps, and false line
GRAPH?
segments. In short, graphing utilities can suggest graph shapes, but they cannot establish
them with certainty. Thus, the more you know about the functions you are graphing, the
easier it will be to choose good viewing windows, and the better you will be able to judge
the reasonableness of the results produced by your graphing utility.
••••••••••••••••••••••••••••••••••••••
The main source of information about your graphing utility is its own documentation,
MORE INFORMATION ON and from time to time we will suggest that you refer to that documentation to learn some
GRAPHING AND CALCULATING
UTILITIES particular technique.
1. Use a graphing utility to generate the graph of the function the window that you think gives the best view of the graph.
f(x) = x 4 − x 2 in the given viewing windows, and specify (a) [−1, 1] × [13, 15] (b) [−2, 2] × [11, 15]
the window that you think gives the best view of the graph. (c) [−4, 4] × [10, 28] (d) A window of your choice
(a) [−50, 50] × [−50, 50] (b) [−5, 5] × [−5, 5]
4. Use a graphing utility to generate the graph of the function
(c) [−2, 2] × [−2, 2] (d) [−2, 2] × [−1, 1]
f(x) = −12 − x 2 in the given viewing windows, and spec-
(e) [−1.5, 1.5] × [−0.5, 0.5] ify the window that you think gives the best view of the
2. Use a graphing utility to generate the graph of the function graph.
f(x) = x 5 − x 3 in the given viewing windows, and specify (a) [−1, 1] × [−15, −13] (b) [−2, 2] × [−15, −11]
the window that you think gives the best view of the graph. (c) [−4, 4] × [−28, −10] (d) A window of your choice
(a) [−50, 50] × [−50, 50] (b) [−5, 5] × [−5, 5]
(c) [−2, 2] × [−2, 2] (d) [−2, 2] × [−1, 1]
In Exercises 5 and 6, use the domain and range of f to deter-
(e) [−1.5, 1.5] × [−0.5, 0.5]
mine a viewing window that contains the entire graph, and
3. Use a graphing utility to generate the graph of the function generate the graph in that window.
f(x) = x 2 + 12 in the given viewing windows, and specify
January 12, 2001 11:10 g65-ch1 Sheet number 31 Page number 37 cyan magenta yellow black
38 Functions
29. (a) Generate the graphs of f(x) = x 2/5 and g(x) = |x|2/5 , (b) Find the x-intercepts of the bipartite cubic
and determine whether your graphing utility missed part y 2 = x(x − a)(x − b)
of the graph of f .
and make a conjecture about how changes in the values
(b) Generate the graphs of the functions f(x) = x 1/5 and of a and b would affect the graph. Test your conjecture
g(x) = (|x|/x)|x|1/5 , and then determine whether your by graphing the bipartite cubic for various values of a
graphing utility missed part of the graph of f . and b.
(c) Generate a graph of the function f(x) = (x − 1)4/5 that
shows all of its important features. y
(d) Generate a graph of the function f(x) = (x + 1)3/4 that
shows all of its important features.
2/3 1/3
30. The graphs of y = x 2 − 4 and y = (x 2 − 4)2
should be the same. Does your graphing utility produce the x
same graph for both equations? If not, what do you think is
happening?
31. In each part, graph the function for various values of c, and
write a paragraph or two that describes how changes in c
affect the graph in each case.
(a) y = cx 2 Bipartite cubic
Figure Ex-32
(b) y = x 2 + cx
(c) y = x 2 + x + c 33. Based on your knowledge of the graphs of y = x and
32. The graph of an equation of the form y = x(x − a)(x − b)
2 y = sin x, make a sketch of the graph of y = x sin x. Check
(where 0 < a < b) is called a bipartite cubic. The accom- your conclusion using a graphing utility.
panying figure shows a typical graph of this type. 34. What do you think the graph of y = sin(1/x) looks like?
(a) Graph the bipartite cubic y 2 = x(x − 1)(x − 2) by solv- Test your conclusion using a graphing utility. [Suggestion:
ing for y in terms of x and graphing the two resulting Examine the graph on a succession of smaller and smaller
functions. intervals centered at x = 0.]
Just as numbers can be added, subtracted, multiplied, and divided to produce other
numbers, so functions can be added, subtracted, multiplied, and divided to produce
other functions. In this section we will discuss these operations and some others that
have no analogs in ordinary arithmetic.
••••••••••••••••••••••••••••••••••••••
Two functions, f and g, can be added, subtracted, multiplied, and divided in a natural way
ARITHMETIC OPERATIONS ON to form new functions f + g, f − g, f g, and f /g. For example, f + g is defined by the
FUNCTIONS
formula
(f + g)(x) = f(x) + g(x) (1)
which states that for each input the value of f + g is obtained by adding the values of f
and g. For example, if
f(x) = x and g(x) = x 2
then
(f + g)(x) = f(x) + g(x) = x + x 2
Equation (1) provides a formula for f + g but does not say anything about the domain of
f + g. However, for the right side of this equation to be defined, x must lie in the domain
of f and in the domain of g, so we define the domain of f + g to be the intersection of
those two domains. More generally, we make the following definition:
January 12, 2001 11:10 g65-ch1 Sheet number 33 Page number 39 cyan magenta yellow black
• REMARK.
• If f is a constant function, say f(x) = c for all x, then the product of f and g
•
•
•
•
•
•
• is cg, so multiplying a function by a constant is a special case of multiplying two functions.
Example 1 Let
√
f(x) = 1 + x − 2 and g(x) = x − 3
Find (f + g)(x), (f − g)(x), (f g)(x), (f /g)(x), and (7f )(x); state the domains of f + g,
f − g, f g, f /g, and 7f .
Solution. First, we will find formulas for the functions and then the domains. The formulas
are
√ √
(f + g)(x) = f(x) + g(x) = (1 + x − 2 ) + (x − 3) = x − 2 + x−2 (2)
√ √
(f − g)(x) = f(x) − g(x) = (1 + x − 2 ) − (x − 3) = 4 − x + x−2 (3)
√
(f g)(x) = f(x)g(x) = (1 + x − 2 )(x − 3) (4)
√
1+ x−2
(f /g)(x) = f(x)/g(x) = (5)
x−3
√
(7f )(x) = 7f(x) =7+7 x−2 (6)
In all five cases the natural domain determined by the formula is the same as the domain
specified in Definition 1.4.1, so there is no need to state the domain explicitly in any of
these cases. For example, the domain of f is [2, +⬁), the domain of g is (−⬁, +⬁), and the
natural domain for f(x) + g(x) determined by Formula (2) is [2, +⬁), which is precisely
the intersection of the domains of f and g.
• REMARK.
• There are situations in which the natural domain associated with the formula
•
•
•
•
•
•
•
resulting from an operation
√ on two functions
√ is not the correct domain for the new function.
•
•
•
•
•
•
For example, if f(x) = x and g(x) = x, then according to Definition √ 1.4.1
√ the domain
•
•
•
•
•
of f g should be [0, +⬁) ∩ [0, +⬁) = [0, +⬁). However, (f g)(x) = x x = x, which
has a natural domain of (−⬁, +⬁). Thus, to be precise in describing the formula for f g,
•
•
•
•
•
••••••••••••••••••••••••••••••••••••••
Multiplying a function f by a positive constant c has the geometric effect of stretching or
STRETCHES AND COMPRESSIONS compressing the graph of y = f(x) in the y-direction. For example, examine the graphs of
y = f(x), y = 2f(x), and y = 12 f(x) shown in Figure 1.4.1a. Multiplying by 2 doubles
each y-coordinate, thereby stretching the graph, and multiplying by 12 cuts each y-coordinate
in half, thereby compressing the graph. In general, if c > 0, then the graph of y = cf(x) can
be obtained from the graph of y = f(x) by compressing the graph of y = f(x) vertically
by a factor of 1/c if 0 < c < 1, or stretching it by a factor of c if c > 1.
Analogously, multiplying x by a positive constant c has the geometric effect of stretching
or compressing the graph of y = f(x) in the x-direction. For example, examine the graphs
January 12, 2001 11:10 g65-ch1 Sheet number 34 Page number 40 cyan magenta yellow black
40 Functions
y y
y = 2 f (x)
y = f (x) y = f (x / 2)
1
y= 2
f (x) y = f (x)
y = f (2x)
x x
(a) (b)
Figure 1.4.1
√x x 1/x x x
Figure 1.4.3
••••••••••••••••••••••••••••••••••••••
We now consider an operation on functions, called composition, which has no direct analog
COMPOSITION OF FUNCTIONS in ordinary arithmetic. Informally stated, the operation of composition is performed by
substituting some function for the independent variable of another function. For example,
suppose that
f(x) = x 2 and g(x) = x + 1
January 12, 2001 11:10 g65-ch1 Sheet number 35 Page number 41 cyan magenta yellow black
• REMARK.
• Although the domain of f ◦ g may seem complicated at first glance, it makes
•
•
•
•
•
•
•
•
sense intuitively: To compute f(g(x)) one needs x in the domain of g to compute g(x),
•
•
•
• then one needs g(x) in the domain of f to compute f(g(x)).
••••••••••••••••••••••••••••••••••••••
In Section 1.1 we noted that a function f can be viewed as a computer program that takes
COMPOSITIONS VIEWED AS an input x, operates on it, and produces an output f(x). From this viewpoint composition
COMPUTER PROGRAMS
can be viewed as two programs, g and f , operating in succession: An input x is fed first to a
program g, which produces the output g(x); then this output is fed as input to a program f ,
which produces the output f(g(x)) (Figure 1.4.4). However, rather than have two separate
programs operating in succession, we could create a single program that takes the input
x and directly produces the output f(g(x)). This program is the composition f ◦ g since
(f ◦ g)(x) = f(g(x)).
g f
Computer Computer
Program Program
Input x Output g(x) Output f (g(x))
fg
Computer
Program
Input x Output f (g(x))
Figure 1.4.4
√
Example 3 Let f(x) = x 2 + 3 and g(x) = x. Find
(a) (f ◦ g)(x) (b) (g ◦ f )(x)
42 Functions
√ is no need to indicate that the domain is (−⬁, +⬁), since this is the natural domain
There
of x 2 + 3.
• REMARK.
• Note that the functions f ◦ g and g ◦ f in the preceding example are not the
•
•
•
•
•
•
•
same. Thus, the order in which functions are composed can (and usually will) make a
•
•
•
• difference in the end result.
Compositions can also be defined for three or more functions; for example, (f ◦ g ◦ h)(x)
is computed as
(f ◦ g ◦ h)(x) = f(g(h(x)))
In other words, first find h(x), then find g(h(x)), and then find f(g(h(x))).
Solution.
(f ◦ g ◦ h)(x) = f(g(h(x))) = f(g(x 3 )) = f(1/x 3 ) = 1/x 3 = 1/x 3/2
••••••••••••••••••••••••••••••••••••••
Many problems in mathematics are attacked by “decomposing” functions into compositions
EXPRESSING A FUNCTION AS A of simpler functions. For example, consider the function h given by
COMPOSITION
h(x) = (x + 1)2
To evaluate h(x) for a given value of x, we would first compute x + 1 and then square the
result. These two operations are performed by the functions
g(x) = x + 1 and f(x) = x 2
We can express h in terms of f and g by writing
h(x) = (x + 1)2 = [g(x)]2 = f(g(x))
so we have succeeded in expressing h as the composition h = f ◦ g.
The thought process in this example suggests a general procedure for decomposing a
function h into a composition h = f ◦ g:
• Think about how you would evaluate h(x) for a specific value of x, trying to break the
evaluation into two steps performed in succession.
• The first operation in the evaluation will determine a function g and the second a func-
tion f .
• The formula for h can then be written as h(x) = f(g(x)).
For descriptive purposes, we will refer to g as the “inside function” and f as the “outside
function” in the expression f(g(x)). The inside function performs the first operation and
the outside function performs the second.
Solution. To evaluate h(x) for a given value of x we would first compute x − 4 and then
raise the result to the fifth power. Therefore, the inside function (first operation) is
g(x) = x − 4
and the outside function (second operation) is
f(x) = x 5
so h(x) = f(g(x)). As a check,
f(g(x)) = [g(x)]5 = (x − 4)5 = h(x)
Example 6 Express sin(x 3 ) as a composition of two functions.
Solution. To evaluate sin(x 3 ), we would first compute x 3 and then take the sine, so
g(x) = x 3 is the inside function and f(x) = sin x the outside function. Therefore,
Table 1.4.1
g(x) f (x)
function inside outside composition
• REMARK.
• It should be noted that there is always more than one way to express a function
•
•
as a composition. For example, here are two ways to express (x 2 + 1)10 as a composition
•
•
•
•
•
•
•
•
•
•
that differ from that in Table 1.4.1:
•
•
5
•
•
•
•
2
(x 2 + 1)10 = (x 2 + 1)2 = f(g(x))
•
•
•
•
g(x) = x 2 + 1 and f(x) = x 5
•
•
10/3
•
•
3
•
•
•
•
•
•
(x 2 + 1)10 = (x 2 + 1)3 = f(g(x)) g(x) = x 2 + 1 and f(x) = x 10/3
••••••••••••••••••••••••••••••••••••••
Figure 1.4.5 shows the graphs of three curves that have certain obvious symmetries. The
SYMMETRY graph in part (a) is symmetric about the x-axis in the sense that for each point (x, y) on the
graph the point (x, −y) is also on the graph; the graph in part (b) is symmetric about the
y-axis in the sense that for each point (x, y) on the graph the point (−x, y) is also on the
graph; and the graph in part (c) is symmetric about the origin in the sense that for each point
(x, y) on the graph the point (−x, −y) is also on the graph. Geometrically, symmetry about
the origin occurs if rotating the graph 180 ◦ about the origin leaves the graph unchanged.
Symmetries can often be detected from the equation of a curve. For example, the graph of
y = x3 (7)
must be symmetric about the origin because for any point (x, y) whose coordinates satisfy
January 12, 2001 11:10 g65-ch1 Sheet number 38 Page number 44 cyan magenta yellow black
44 Functions
y y y
x x x
(7), the coordinates of the point (−x, −y) also satisfy (7), since substituting these coordi-
nates in (7) yields
−y = (−x)3
which simplifies to (7). This suggests the following symmetry tests (Figure 1.4.6).
y y y y
(–x, y) (x, y) (–x, y) (x, y) (x, y) (x, y)
x x x x
Figure 1.4.6
••••••••••••••••••••••••••••••••••••••
For the graph of a function f to be symmetric about the y-axis, the equations y = f(x) and
EVEN AND ODD FUNCTIONS y = f(−x) must be equivalent; for this to happen we must have
f(x) = f(−x)
A function with this property is called an even function. Some examples are x 2 , x 4 , x 6 ,
and cos x. Similarly, for the graph of a function f to be symmetric about the origin, the
equations y = f(x) and −y = f(−x) must be equivalent; for this to happen we must have
f(x) = −f(−x)
A function with this property is called an odd function. Some examples are x, x 3 , x 5 , and
sin x.
••••••••••••••••••••••••••••••••••••••
Once you know the graph of an equation y = f(x), there are some techniques that can be
TRANSLATIONS used to help visualize the graphs of the equations
y = f(x) + c, y = f(x) − c, y = f(x + c), y = f(x − c)
where c is any positive constant.
If a positive constant is added to or subtracted from f(x), the geometric effect is to
translate the graph of y = f(x) parallel to the y-axis; addition translates the graph in the
positive direction and subtraction translates it in the negative direction. This is illustrated in
Table 1.4.2. Similarly, if a positive constant is added to or subtracted from the independent
variable x, the geometric effect is to translate the graph of the function parallel to the x-axis;
subtraction translates the graph in the positive direction, and addition translates it in the
negative direction. This is also illustrated in Table 1.4.2.
Table 1.4.2
geometric Translates the graph of Translates the graph of Translates the graph of Translates the graph of
effect y = f (x) up c units y = f (x) down c units y = f (x) left c units y = f (x) right c units
y y y y
y = x2 + 2 y = (x + 2) 2 y = (x – 2) 2
y = x2 y = x2 y = x2
2 y = x2 y = x2 – 2
example x x x x
-2 2
-2
Before proceeding to the following examples, it will be helpful to review the graphs in
Figures 1.2.1 and 1.2.5.
x x x
9 3 12 –3 6
y = √x y = √x – 3 y = √x + 3
Figure 1.4.7
Solution. The graph can be obtained by two translations: first translate the graph of
y = |x| right 3 units to obtain the graph of y = |x − 3|, then translate this graph up 2 units
to obtain the graph of y = |x − 3| + 2 (Figure 1.4.8).
January 12, 2001 11:10 g65-ch1 Sheet number 40 Page number 46 cyan magenta yellow black
46 Functions
y y y
2
x x x
1 3 3
y = |x| y = |x – 3| y = |x – 3| + 2
Figure 1.4.8
• REMARK.
• The graph in the preceding example could also have been obtained by perform-
•
•
ing the translations in the opposite order: first translating the graph of y = |x| up 2 units to
•
•
•
•
•
•
obtain the graph of y = |x| + 2, then translating this graph right 3 units to obtain the graph
•
•
•
•
•
of y = |x − 3| + 2.
•
•
•
•
Solution. We can build up Equation (8) in three steps from the equation y = x 2 . First,
we can multiply by a to obtain y = ax 2 . If a > 0, this operation has the geometric effect
of stretching or compressing the graph of y = x 2 ; and if a < 0, it has the geometric
effect of reflecting the graph about the x-axis, in addition to stretching or compressing it.
Since stretching or compressing does not alter the general parabolic shape of the original
curve, the graph of y = ax 2 looks roughly like one of those in Figure 1.4.10a. Next, we can
subtract h from x to obtain the equation y = a(x − h)2 , and then we can add k to obtain
y = a(x − h)2 + k. Subtracting h causes a horizontal translation (right or left, depending
on the sign of h), and adding k causes a vertical translation (up or down, depending on the
sign of k). Thus, the graph of (8) looks roughly like one of those in Figure 1.4.10b, which
are shown with h > 0 and k > 0 for simplicity.
y y y y
x x k x
k x
h
h
(a) (b)
Figure 1.4.10
January 12, 2001 11:10 g65-ch1 Sheet number 41 Page number 47 cyan magenta yellow black
••••••••••••••••••••••••••••••••••••••
The graph of y = f(−x) is the reflection of the graph of y = f(x) about the y-axis, and
REFLECTIONS the graph of y = −f(x) [or equivalently, −y = f(x)] is the reflection of the graph of
y = f(x) about the x-axis. Thus, if you know what the graph of y = f(x) looks like, you
can obtain the graphs of y = f(−x) and y = −f(x) by making appropriate reflections.
This is illustrated in Table 1.4.3.
Table 1.4.3
operation on
y = f (x) Replace x by – x Multiply f (x) by –1
y y
y = √–x 3 y = √x 3 y = √x
x x
example
-6 6 -6 6
-3 -3
y = –√x
√
Example 12 Sketch the graph of y = 3
2 − x.
Solution. The
√ graph can be obtained by a reflection and a translation:
√ first reflect the
graph of y = 3 x about the y-axis to obtain the graph of y√= 3 −x, then √translate this
graph right 2 units to obtain the graph of the equation y = 3 −(x − 2) = 3 2 − x (Fig-
ure 1.4.11).
y y y
6 6 6
x x x
–10 10 –10 10 –10 10
–6 –6 –6
3 3 3
y = √x y = √–x y = √2 – x
Figure 1.4.11
Solution. The graph can be obtained by a reflection and two translations: first translate
the graph of y = |x| right 2 units to obtain the graph of y = |x − 2|; then reflect this graph
about the x-axis to obtain the graph of y = −|x − 2|; and then translate this graph up 4
units to obtain the graph of the equation y = −|x − 2| + 4 = 4 − |x − 2| (Figure 1.4.12).
January 12, 2001 11:10 g65-ch1 Sheet number 42 Page number 48 cyan magenta yellow black
48 Functions
y y y y
8 8 8 8
x x x x
-8 8 -6 10 -6 10 -6 10
-8 -8 -8 -8
y = | x| y = | x – 2| y = – |x – 2| y = 4 – | x – 2|
Figure 1.4.12
y
In Exercises 39–44, find formulas for f ◦ g and g ◦ f , and
3
state the domains of the functions.
f
39. f(x) = 2x + 1, g(x) = x 2 − x
x
40. f(x) = 2 − x 2 , g(x) = x 3
√ -3 3
41. f(x) = x 2 , g(x) = 1 − x
√ √ g
42. f(x) = x − 3, g(x) = x 2 + 3
1+x x -3
43. f(x) = , g(x) =
1−x 1−x
Figure Ex-55
x 1
44. f(x) = , g(x) =
1+x 2 x
56. Sketch the graph of y = g(f(x)) for the functions graphed
In Exercises 45 and 46, find a formula for f ◦ g ◦ h.
in Exercise 55.
1 57. Use the graphs of f and g in Exercise 55 to estimate the
45. f(x) = x 2 + 1, g(x) =, h(x) = x 3 solutions of the equations f(g(x)) = 0 and g(f(x)) = 0.
x
1 √ 1 58. Use the table in Exercise 53 to solve the equations
46. f(x) = , g(x) = 3 x, h(x) = 3
1+x x f(g(x)) = 0 and g(f(x)) = 0.
January 12, 2001 11:10 g65-ch1 Sheet number 44 Page number 50 cyan magenta yellow black
50 Functions
y
In Exercises 59–62, find
f(x + h) − f(x) f(w) − f(x)
and
h w−x
Simplify as much as possible. x
y y
x x
x x
(a) (b)
(a) (b)
y y
y y
x x
x x
(c) (d)
Figure Ex-67
(c) (d)
Figure Ex-63
68. Classify the functions whose values are given in the follow-
ing table as even, odd, or neither.
64. The accompanying figure shows a portion of a graph. Com-
plete the graph so that the entire graph is symmetric about
(a) the x-axis (b) the y-axis (c) the origin. x –3 –2 –1 0 1 2 3
f (x) 5 3 2 3 1 –3 5
y
g(x) 4 1 –2 0 2 –1 –4
h(x) 2 –5 8 –2 8 –5 2
x
1.5 Lines 51
1.5 LINES
This section includes a quick review of precalculus material on lines. Readers who
want to review this material in more depth are referred to Appendix C.
••••••••••••••••••••••••••••••••••••••
An equation that is expressible in the form
EQUATIONS OF LINES
Ax + By + C = 0 (1)
where A and B are not both zero, is called a first-degree equation or a linear equation in
x and y. It is shown in precalculus that every first-degree equation in x and y has a straight
line as its graph and, conversely, every straight line can be represented by a first-degree
equation in x and y. For this reason (1) is sometimes called the general equation of a line.
Recall that equations of lines may be written in several different forms:
y = mx + b Slope-intercept form (2)
y − y1 = m(x − x1 ) Point-slope form (3)
x y
+ =1 Double-intercept form (4)
a b
January 12, 2001 11:10 g65-ch1 Sheet number 46 Page number 52 cyan magenta yellow black
52 Functions
In these equations m is the slope of the line, a is the x-intercept, b is the y-intercept, and
(x1 , y1 ) is any point on the line (Figure 1.5.1). Keep in mind that these equations do not
apply to vertical lines. For vertical lines the slope is undefined, or stated informally, a vertical
line has infinite slope. Vertical and horizontal lines have particularly simple equations:
x=a The vertical line with x-intercept a (5)
y y y y
(x1, y1) x=a
y=b b
b Slope = m Slope = m b
x x x x
a a
Figure 1.5.1
y = mx + b
Equation (2) is especially useful because the slope and the y-intercept of the line can
be determined by inspection: the slope is the coefficient of x, and the y-intercept is the
Slope y -intercept
constant term (Figure 1.5.2). This equation expresses y as a function of x, the function
Figure 1.5.2 being f(x) = mx + b. A function of this form is called a linear function of x.
••••••••••••••••••••••••••••••••••••••
The slope m of a nonvertical line y = mx + b has two important interpretations (which are
INTERPRETATIONS OF SLOPE related but different in viewpoint):
The steepness interpretation has an analog in surveying. Surveyors measure the grade or
slope of a hill as the ratio of its rise over its run (Figure 1.5.3a). The same idea applies
to lines. Consider a particle that moves left to right along a nonvertical line from a point
P1 (x1 , y1 ) to a point P2 (x2 , y2 ). In the course of its travel the point moves y2 − y1 units
vertically as it travels x2 − x1 units horizontally (Figure 1.5.3b). The vertical change, which
is denoted by -y = y2 − y1 , is called the rise, and the horizontal change, which is denoted
by -x = x2 − x1 , is called the run. The ratio of the rise over the run is always equal to the
y
rise 2 ft 1 P2(x2, y2 )
Slope = = =
run 20 ft 10
y2 – y1
Expressed as a percentage,
this is a 10% grade. (Rise)
P1(x1, y1)
Rise = 2 ft x 2 – x1
Run = 20 ft (Run) x
(a) (b)
Figure 1.5.3
January 12, 2001 11:10 g65-ch1 Sheet number 47 Page number 53 cyan magenta yellow black
1.5 Lines 53
slope, regardless of where the points P1 and P2 are located on the line; that is,
-y y2 − y1
m= = (7)
-x x2 − x1
• REMARK.
• The symbols -x and -y should not be interpreted as products; rather, -x
•
•
•
•
•
•
•
should be viewed as a single entity representing the change in the value of x, and -y as a
•
•
•
•
•
•
single entity representing the change in the value of y. In general, if v is any variable whose
value changes from an initial value of v1 to a final value of v2 , then we call -v = v2 − v1
•
•
•
•
•
•
•
•
•
•
(final value minus initial value) an increment in v. Increments can be positive or negative,
•
•
•
• depending on whether the final value is larger or smaller than the initial value.
••••••••••••••••••••••••••••••••••••••
The slope of a nonvertical line L is related to the angle that L makes with the positive x-axis.
ANGLE OF INCLINATION If φ is the smallest positive angle measured counterclockwise from the x-axis to L, then
the slope of the line can be expressed as
m = tan φ (8)
(Figure 1.5.4a). The angle φ, which is called the angle of inclination of the line, satisfies
0 ◦ ≤ φ < 180 ◦ in degree measure (or, equivalently, 0 ≤ φ < π in radian measure). If φ is
an acute angle, then m = tan φ is positive and the line slopes up to the right, and if φ is an
obtuse angle, then m = tan φ is negative and the line slopes down to the right. For example,
a line whose angle of inclination is 45 ◦ has slope m = tan 45 ◦ = 1, and a line whose angle
y m=3 of inclination is 135 ◦ has a slope of m = tan 135 ◦ = −1 (Figure 1.5.4b). Figure 1.5.5
m=2 shows a convenient way of using the line x = 1 as a “ruler” for visualizing the relationship
4 Positive between lines of various slopes.
slope
3
m=1 y y
2
1
m = –1 m=1
m=0
1 Rise
-1
135°
f x x
-2 45°
m = –1 Run
-3 Negative
slope
-4
m = rise = tan f
run
m = –2
m = –3 (a) (b)
x=1
Figure 1.5.4
Figure 1.5.5
••••••••••••••••••••••••••••••••••••••
In applied problems, changing the units of measurement can change the slope of a line,
SLOPES OF LINES IN APPLIED so it is essential to include the units when calculating the slope. The following example
PROBLEMS
illustrates this.
54 Functions
25 25
Temperature T (°C)
Temperature T (°C)
20 20
25°C 5°C
Thermal insulation 15 15
10 10
5 5
Rod x
0 40 0 10 20 30 40 0 0.1 0.2 0.3 0.4
Distance x (cm) Distance x (m)
(a) (b) (c)
Figure 1.5.6
Solution (a). The slope is m = −0.5 and the intercept on the T -axis is 25, so
T = −0.5x + 25, 0 ≤ x ≤ 40
where the restriction on x is required because the rod is 40 cm in length. The graph of this
equation with the restriction is a line segment rather than a line.
Solution (b). The slope is m = −50, the intercept on the T -axis is 25, and the restriction
on x is 0 ≤ x ≤ 0.4. Thus, the equation is
T = −50x + 25, 0 ≤ x ≤ 0.4
••••••••••••••••••••••••••••••••••••••
If y is a linear function of x, say y = mx + b, then it follows from (7) that
SLOPES AS RATE OF CHANGE
-y = m-x
Thus, a 1-unit increase in x (-x = 1) produces an m-unit change in y (-y = m). Moreover,
y
this is true at every point on the line (Figure 1.5.7), so we say that y changes at a constant
rate with respect to x, and we call m the rate of change of y with respect to x. This idea
y = mx + b m
can be summarized as follows.
1
m
1 1.5.1 CONSTANT RATE OF CHANGE. If a variable y is related to a variable x in such
m
a way that the rate of change of y with respect to x is constant, say m, then y is a linear
1
x function of x of the form
y = mx + b
A 1-unit increase in x always Conversely, if y is a linear function of x whose graph has slope m, then the rate of change
produces an m -unit change in y.
of y with respect to x is constant and equal to m.
Figure 1.5.7
It follows from this that linear functions are appropriate whenever experimentation or theory
suggests that the rate of change of y with respect to x is constant.
January 12, 2001 11:10 g65-ch1 Sheet number 49 Page number 55 cyan magenta yellow black
1.5 Lines 55
••••••••••••••••••••••••••••••••••••••
One of the important themes in calculus is the study of motion. To describe the motion of
UNIFORM RECTILINEAR MOTION an object completely, one must specify its speed (how fast it is going) and the direction in
which it is moving. The speed and the direction of motion together comprise what is called
the velocity of the object. For example, knowing that the speed of an aircraft is 500 mi/h tells
us how fast it is going, but not which way it is moving. In contrast, knowing that the velocity
of the aircraft is 500 mi/h due south pins down the speed and the direction of motion.
Later, we will study the motion of particles that move along curves in two- or three-
dimensional space, but for now we will focus on motion along a line; this is called rectilinear
motion. In general rectilinear motion, a particle can move back and forth along the line (as
with a piston moving up and down in a cylinder); however, for now we will only consider
the simple case in which the particle moves in just one direction along a line (as with a car
traveling on a straight road).
For simplicity, we will assume that the motion is along a coordinate line, such as an x-axis
or y-axis, and that the particle is moving in the positive direction. In general discussions
we will usually name the coordinate line the s-axis to avoid being specific. A graphical
description of rectilinear motion along an s-axis can be obtained by making a plot of the
s-coordinate of the particle versus the elapsed time t. This is called the position versus time
curve for the particle. Figure 1.5.8a shows a typical position versus time curve for a car
moving in the positive direction along an s-axis.
0
t e
av
s s
e =v
Position versus lop
time curve (t1, s1) S
s1 – s0
Elapsed time (t0, s0)
t1 – t0
s t t
0 t0 t1
(a) (b)
Figure 1.5.8
Because we are assuming that the particle is moving in the positive direction of the s-
axis, there is no ambiguity about the direction of motion, and hence the terms “speed” and
“velocity” can be used interchangeably. However, later, when we consider general rectilinear
motion or motion along a curved path, it will be necessary to distinguish between these terms,
since the direction of motion may vary.
For a particle in rectilinear motion along a coordinate axis, we define the average velocity
vave of the particle during the time interval from t0 to t1 to be
s1 − s0 -s
vave = = (11)
t1 − t0 -t
where s0 and s1 are the s-coordinates of the particle at times t0 and t1 , respectively. Ge-
ometrically, this is the slope of the secant line connecting the points (t0 , s0 ) and (t1 , s1 )
on the position versus time curve (Figure 1.5.8b). The quantity -s = s1 − s0 is called
the displacement or change in position of the particle during the time interval from t0 to
t1 . With this terminology, Formula (11) states that for a particle in rectilinear motion the
average velocity over a time interval is the displacement during the time interval divided
by the length of the time interval. For example, if a car moving in one direction along a
straight road travels 75 miles in 3 hours, then its average velocity is 75/3 = 25 mi/h.
January 12, 2001 11:10 g65-ch1 Sheet number 50 Page number 56 cyan magenta yellow black
56 Functions
In the special case where the average velocity of a particle in rectilinear motion is the
s (ft)
1000 same over every time interval, the particle is said to have constant velocity or uniform
s = 100 + 88t rectilinear motion. If the average velocity over every time interval is v, then we will refer
to v as the velocity of the particle (dropping the adjective “average”).
500 For a particle with uniform rectilinear motion the displacement over any time interval is
given by the formula
100 t (s) displacement = velocity × elapsed time (12)
1 2 3 4 5 6 7 8 9 10
Figure 1.5.9 Example 3 Suppose that a car moves with a constant velocity of 88 ft/s in the positive
direction of an s-axis. Given that the s-coordinate of the car at time t = 0 is s = 100, find
s an equation for s as a function of t, and graph the position versus time curve.
s = s0 + vt
Solution. It follows from (12) that in a period of t seconds, the car will move 88t feet
v
p e= from its starting point, so its coordinate s at time t will be
Slo
s = 100 + 88t
s0
t
The graph of this equation is the line in Figure 1.5.9.
It is not accidental that the position versus time curve turned out to be a line in the last
Position versus time curve for example; this will always be the case for uniform rectilinear motion. To see why this is so,
a particle with coordinate s0 suppose that a particle moves with constant velocity v in the positive direction along an
at time t = 0 and moving with
constant velocity v
s-axis, starting at the point s0 at time t = 0. It follows from (12) that in t units of time the
particle will move vt units from its starting point s0 , so its coordinate s at time t will be
Figure 1.5.10
s = s0 + vt
s (m) which is a line with s-intercept s0 and slope v (Figure 1.5.10). It follows from this equation
and 1.5.1 that we can view the velocity v as the rate of change of s with respect to t, that is,
4 the rate of change of position with respect to time.
3 Example 4 Figure 1.5.11 shows the position versus time curve for a particle moving
along an s-axis. Describe the motion of the particle in words.
2
1
Solution. At time t = 0 the particle is at the origin. From time t = 0 to t = 2 the slope
of the line segment is 12 , so the particle is moving with a constant velocity of 12 = 0.5 m/s.
t (s)
At time t = 2 the particle is at the point s = 1 (i.e., 1 meter from the origin). From time
1 2 3 4
t = 2 to t = 4 the slope of the line segment is 32 , so the particle is moving with a constant
Figure 1.5.11 velocity of 32 = 1.5 m/s. At time t = 4 it is at the point s = 4.
••••••••••••••••••••••••••••••••••••••
In everyday language we say that an object is “accelerating” if it is speeding up and “decel-
CONSTANT ACCELERATION erating” if it is slowing down. Mathematically, the acceleration of a particle in rectilinear
motion is defined to be the rate of change of velocity with respect to time, where the accel-
eration is positive if the velocity is increasing and negative if it is decreasing. Thus, for a
particle that moves in the positive direction of an s-axis, negative acceleration means the
v
particle is “decelerating” in everyday language. Acceleration, like velocity, can be variable
v = v 0 + at or constant. For example, by pressing the gas pedal of a car toward the floor smoothly, the
driver can make the car’s velocity increase at a constant rate (a constant acceleration); how-
a ever, if the driver suddenly slams the pedal to the floor, the car will lurch forward, reflecting
p e=
Slo a nonconstant acceleration. Later in the text we will study acceleration in more depth, but
for now we will only consider the case in which acceleration is constant.
v0
t
• REMARK.
• The units of acceleration are units of velocity divided by units of time. For
•
•
•
•
1.5 Lines 57
Graphical information about the acceleration of a particle can be obtained from the graph
of velocity versus time; this is called the velocity versus time curve. In the case where the
particle has constant acceleration, the velocity versus time curve will be linear, and its slope,
which is the rate of change of velocity with time, will be the acceleration (Figure 1.5.12).
Example 5 Suppose that a car moves in the positive direction of an s-axis in such a way
that its velocity v increases at a constant rate of 2 ft/s2 .
v (ft/s)
125 (a) Assuming that the velocity of the car is 88 ft/s at time t = 0, find an equation for v as
100 v = 88 + 2t a function of t.
75 (b) Make a graph of velocity versus time, and mark the point on the graph at which the
50 car attains a velocity of 100 ft/s.
25
t (s)
2 4 6 8 10 12 14 16 18 20
Solution (a). Since the rate of change of v with respect to t is 2 ft/s2 , and since v = 88
ft/s if t = 0, the equation for velocity as a function of time is
Velocity versus time curve for v = 88 + 2t (13)
a particle with a velocity of 88 ft/s
at time t = 0 and moving with a Solution (b). To find the time it takes for the car to reach a velocity of 100 ft/s, we
constant acceleration of 2 ft/s2
substitute v = 100 in (13) and solve for t. This yields t = 6. The graph of (13) and the
Figure 1.5.13 point at which the velocity reaches 100 ft/s is shown in Figure 1.5.13.
••••••••••••••••••••••••••••••••••••••
Recall that a variable y is said to be directly proportional to a variable x if there is a positive
DIRECT PROPORTION constant k, called the constant of proportionality, such that
y = kx (14)
The graph of this equation is a line through the origin whose slope k is the constant of
proportionality. Thus, linear functions are appropriate in physical problems where one
variable is directly proportional to another.
∗
Hooke’s law in physics provides a nice example of direct proportion. It follows from
this law that if a weight of x units is suspended from a spring, then the spring will be
y stretched by an amount y that is directly proportional to x, that is, y = kx (Figure 1.5.14).
The constant k depends on the stiffness of the spring: the stiffer the spring, the smaller the
value of k (why?).
x
y is directly proportional to x.
Example 6 Figure 1.5.15 shows an old-fashioned spring scale that is calibrated in pounds.
Figure 1.5.14 (a) Given that the pound scale marks are 0.5 in apart, find an equation that expresses the
length y that the spring is stretched (in inches) in terms of the suspended weight x (in
pounds).
(b) Graph the equation obtained in part (a).
y = 0.5x
6
Length stretched y (in)
5
4
3
2
1
0 2 4 6 8 10 12
Weight x (lb)
Figure 1.5.15
∗
Hooke’s law, named for the English physicist Robert Hooke (1635–1703), applies only for small displacements
that do not stretch the spring to the point of permanently distorting it.
January 12, 2001 11:10 g65-ch1 Sheet number 52 Page number 58 cyan magenta yellow black
58 Functions
Solution (a). It follows from Hooke’s law that y is related to x by an equation of the form
y = kx. To find k we rewrite this equation as k = y /x and use the fact that a weight of
x = 1 lb stretches the spring y = 0.5 in. Thus,
y 0.5
k= = = 0.5 and hence y = 0.5x
x 1
Solution (b). The graph of the equation y = 0.5x is shown in Figure 1.5.15.
••••••••••••••••••••••••••••••••••••••
One method for determining whether n points
LINEAR DATA
(x1 , y1 ), (x2 , y2 ), . . . , (xn , yn )
lie on a line is to compare the slopes of the line segments joining successive points. The
points lie on a line if and only if those slopes are equal (Figure 1.5.16).
y
yn
...
y3
y2
y1
x
x1 x2 x3 ... xn
y
Solution (b). A linear equation relating x and y can be obtained from the point-slope form
8
of the line using the slope m = 0.8 calculated in part (a) and any one of the five data points.
y = 0.8x – 0.9
If we use the first data point, (1.5, 0.3), we obtain
6
y − 0.3 = 0.8(x − 1.5)
4
or in slope-intercept form,
2
x y = 0.8x − 0.9
-4 -2 2 4 6 8 10 The graph of this equation together with the given data are shown in Figure 1.5.17.
-2
• REMARK.
• Sometimes, data points that should theoretically lie on a line do not because of
-4 •
•
•
•
•
•
•
•
experimental error and other factors. In such cases curve-fitting techniques are used to find
Figure 1.5.17 •
•
•
• a line that most closely fits the data. Such techniques will be discussed later in the text.
••••••••••••••••••••••••••••••••••••••
Linear functions arise in a variety of practical problems. Here is a typical example.
OTHER APPLICATIONS OF LINEAR
FUNCTIONS
Example 8 A university parking lot charges $3.00 per day but offers a $40.00 monthly
sticker with which the student pays only $0.25 per day.
January 12, 2001 11:10 g65-ch1 Sheet number 53 Page number 59 cyan magenta yellow black
1.5 Lines 59
(a) Find equations for the cost C of parking for x days per month under both payment
methods, and graph the equations for 0 ≤ x ≤ 30. (Treat C as a continuous function
of x, even though x only assumes integer values.)
C (b) Find the value of x for which the graphs intersect, and discuss the significance of this
100 C = 3x value.
90
Cost in dollars
80
70
60 Solution (a). The cost in dollars of parking for x days at $3.00 per day is C = 3x, and the
50
40
cost for the $40.00 sticker plus x days at $0.25 per day is C = 40 + 0.25x (Figure 1.5.18).
30 C = 40 + 0.25x
20
10 x
Solution (b). The graphs intersect at the point where
5 10 15 20 25 30 3x = 40 + 0.25x
Number of parking days which is x = 40/2.75 ≈ 14.5. This value of x is not an option for the student, since x
Figure 1.5.18 must be an integer. However, it is the dividing point at which the monthly sticker method
becomes less expensive than the daily payment method; that is, for x ≥ 15 it is cheaper to
buy the monthly sticker and for x ≤ 14 it is cheaper to pay the daily rate.
x x
III IV
Figure Ex-4
60 Functions
8. Find x and y if the line through (0, 0) and (x, y) has slope
1 In Exercises 25 and 26, use the graph to find the equation of
2
, and the line through (x, y) and (7, 5) has slope 2.
the line in slope-intercept form, and then check your result
9. Find x if the slope of the line through (1, 2) and (x, 0) is the by using a graphing utility to graph the equation.
negative of the slope of the line through (4, 5) and (x, 0).
In Exercises 10 and 11, find the angle of inclination of the line 25. y y
with slope m to the nearest degree. Use a calculating utility,
where needed.
x 1 x
10. (a) m = 1
2
(b) m = −1 1
(c) m = 2 (d) m = −57
11. (a) m = − 12 (b) m = 1
(c) m = −2 (d) m = 57 (a) (b)
Figure Ex-25
In Exercises 12 and 13, find the angle of inclination of the
line to the nearest degree. Use a calculating utility, where
needed. 26. y y
√
12. (a) 3y = 2 − 3x (b) y − 4x + 7 = 0
√ x x
13. (a) y = 3x + 2 (b) y + 2x + 5 = 0 1 1
14. Find equations for the x- and y-axes.
15. Slope = −2, y-intercept = 4 27. The accompanying figure shows the position versus time
curve for a particle moving along an x-axis.
16. m = 5, b = −3 (a) What is the velocity of the particle?
17. The line is parallel to y = 4x − 2 and its y-intercept is 7.
(b) What is the x-coordinate of the particle at time t = 0?
18. The line is parallel to 3x + 2y = 5 and passes through
(c) What is the x-coordinate of the particle at time t = 2?
(−1, 2).
(d) At what time does the particle have an x-coordinate of
19. The line is perpendicular to y = 5x + 9 and has y-intercept
x = 4?
6.
20. The line is perpendicular to x − 4y = 7 and passes through
(3, −4). x (ft)
5
21. The line passes through (2, 4) and (1, −7).
22. The line passes through (−3, 6) and (−2, 1).
23. In each part, classify the lines as parallel, perpendicular, or t (s)
neither. 10
(a) y = 4x − 7 and y = 4x + 9
(b) y = 2x − 3 and y = 7 − 12 x
(c) 5x − 3y + 6 = 0 and 10x − 6y + 7 = 0 Figure Ex-27
(d) Ax + By + C = 0 and Bx − Ay + D = 0
(e) y − 2 = 4(x − 3) and y − 7 = 14 (x − 3)
28. A particle moving along an x-axis with constant velocity is
24. In each part, classify the lines as parallel, perpendicular, or
at the point x = 1 when t = 2 and is at the point x = 5
neither.
when t = 4.
(a) y = −5x + 1 and y = 3 − 5x
(a) Find the velocity of the particle if x is in meters and t
(b) y − 1 = 2(x − 3) and y − 4 = − 12 (x + 7)
is in seconds.
(c) 4x + 5y + 7 = 0 and 5x − 4y + 9 = 0
(b) Find an equation that expresses x as a function of t.
(d) Ax + By + C = 0 and Ax + By + D = 0
(e) y = 12 x and x = 12 y (c) What is the coordinate of the particle at time t = 0?
January 12, 2001 11:10 g65-ch1 Sheet number 55 Page number 61 cyan magenta yellow black
1.5 Lines 61
29. A particle moving along an x-axis with constant acceler- 33. A locomotive travels on a straight track at a constant speed
ation has velocity v = 3 ft/s at time t = 1 and velocity of 40 mi/h, then reverses direction and returns to its starting
v = −1 ft/s at time t = 4. point, traveling at a constant speed of 60 mi/h.
(a) Find the acceleration of the particle. (a) What is the average velocity for the round-trip?
(b) Find an equation that expresses v as a function of t. (b) What is the average speed for the round-trip?
(c) What is the velocity of the particle at time t = 0?
(c) What is the total distance traveled by the train if the
30. The accompanying figure shows the velocity versus time total trip took 5 h?
curve for a particle moving along the x-axis.
34. A ball is tossed straight up at time t = 0 with an initial ve-
(a) What is the acceleration of the particle?
locity of 64 ft/s. We will show later using basic principles
(b) What is the velocity of the particle at time t = 0? of physics that the velocity of the ball as a function of time
(c) What is the velocity of the particle at time t = 2? is v = 64 − 32t.
(d) At what time does the particle have a velocity of v = 3 (a) What direction is the ball traveling 3 s after it is re-
ft/s? leased? Explain your reasoning.
(b) At what time does the ball reach its maximum height
v (ft/s)
5
above the ground? Explain your reasoning.
(c) What can you say about the acceleration of the ball?
35. A car is stopped at a toll booth on a straight highway. Start-
t (s) ing at time t = 0 it accelerates at a constant rate of 10 ft/s2
10 for 10 s. It then travels at a constant speed of 100 ft/s for 90
s. At that time it begins to decelerate at a constant rate of 5
ft/s2 for 20 s, at which point in time it reaches a full stop at
Figure Ex-30 a traffic light.
(a) Sketch the velocity versus time curve.
31. The accompanying figure shows the position versus time (b) Express v as a piecewise function of t.
curve for a particle moving along an x-axis. 36. Make a reasonable sketch of a position versus time curve for
(a) Describe the motion of the particle in words. a particle that moves in the positive x-direction with positive
(b) Find the average velocity of the particle from t = 0 to constant acceleration.
t = 10.
37. A spring with a natural length of 15 in stretches to a length
(c) Find the average speed of the particle from t = 0 to of 20 in when a 45-lb object is suspended from it.
t = 10. (a) Use Hooke’s law to find an equation that expresses the
amount y by which the spring is stretched (in inches)
x (cm)
10
in terms of the suspended weight x (in pounds).
(b) Graph the equation obtained in part (a).
(c) Find the length of the spring when a 100-lb object is
suspended from it.
(d) What is the largest weight that can be suspended from
the spring if the spring cannot be stretched to more than
t (s) twice its natural length?
10 Figure Ex-31 38. The spring in a heavy-duty shock absorber has a natural
length of 3 ft and is compressed 0.2 ft by a load of 1 ton. An
32. The accompanying figure shows the velocity versus time additional load of 5 tons compresses the spring an additional
curve for a particle moving along an x-axis. Describe the 1 ft.
motion of the particle in words. (a) Assuming that Hooke’s law applies to compression as
well as extension, find an equation that expresses the
v (km/h) length y that the spring is compressed from its natural
10
length (in feet) in terms of the load x (in tons).
(b) Graph the equation obtained in part (a).
(c) Find the amount that the spring is compressed from its
natural length by a load of 3 tons.
(d) Find the maximum load that can be applied if safety
t (h) regulations prohibit compressing the spring to less than
10 Figure Ex-32 half its natural length.
January 12, 2001 11:10 g65-ch1 Sheet number 56 Page number 62 cyan magenta yellow black
62 Functions
Functions are often grouped into families according to the form of their defining for-
mulas or other common characteristics. In this section we will discuss some of the
most basic families of functions.
••••••••••••••••••••••••••••••••••••••
A function f whose values are all the same is called a constant function. For example, the
FAMILIES OF LINES formula f(x) = c defines the constant function whose value is c for all x. The graph of the
constant function f(x) = c is the horizontal line y = c (Figure 1.6.1a). If we vary c, then
we obtain a set or family of horizontal lines (Figure 1.6.1b).
y y
c = 4
(0, c) y=c c = 3
c = 2
c = 1
x c = 0 x
c = –1
c = –2
c = –3
c = –4.5
(a) (b)
Figure 1.6.1
• REMARK.
• The expression f(x) = c can be confusing because it can be interpreted either
•
•
Example 1
64 Functions
y y
x x
y = mx + b y = mx + b
(b fixed and m varying) (m fixed and b varying)
(a) (b)
Figure 1.6.2
y y = –2x + b y
y = 12 x + b
5 5
4 4
3 3
2 y = 12 x – 1 2
1 x 1 x
-5 -4 -3 -2 -1 1 2 3 4 5 -5 -4 -3 -2 -1 1 2 3 4 5
-1 -1
-2 -2
-3 -3
-4 -4
-5 -5
(a) (b)
Figure 1.6.3
Solution (c). Since the slopes of perpendicular lines are negative reciprocals, it follows
that the lines perpendicular to those in part (a) have slope −2 and hence are of the form
y = −2x + b
Some typical lines in families (1) and (2) are graphed in Figure 1.6.3b.
••••••••••••••••••••••••••••••••••••••
A function of the form f(x) = x p , where p is constant, is called a power function. Consider
THE FAMILY y = x n the case where p is a positive integer, say p = n. The graphs of the curves y = x n for
n = 1, 2, 3, 4, and 5 are shown in Figure 1.6.4. The first graph is the line y = x with slope 1
that passes through the origin, and the second is a parabola that opens up and has its vertex
at the origin (see Appendix D).
January 12, 2001 11:10 g65-ch1 Sheet number 59 Page number 65 cyan magenta yellow black
y y=x y y = x2 y y = x3 y y = x4 y y = x5
1 1 1 1 1
x x x x x
-1 1 -1 1 -1 1 -1 1 -1 1
-1 -1 -1 -1 -1
Figure 1.6.4
For n > 2 the shape of the graph of y = x n depends on whether n is even or odd
(Figure 1.6.5). For even values of n the graphs have the same general shape as the parabola
y = x 2 (though they are not actually parabolas if n > 2), and for odd values of n greater
than 1 they have the same general shape as y = x 3 . The graphs in the family y = x n share
a number of important characteristics:
• For even values of n the functions f(x) = x n are even, and their graphs are symmetric
about the y-axis; for odd values of n the functions f(x) = x n are odd, and their graphs
are symmetric about the origin.
• For all values of n the graphs pass through the origin and the point (1, 1). For even
values of n the graphs pass through (−1, 1), and for odd values of n they pass through
(−1, −1).
• Increasing n causes the graph to become flatter over the interval −1 < x < 1 and
steeper over the intervals x > 1 and x < −1.
y y = x5 3 y y = x6 4
y=x y=x
y = x7 y=x y= x8
y = x2
1 2
x
1
-1 1
x
-1
-1 1
y = xn y = xn
( n odd) ( n even)
Figure 1.6.5
• REMARK.
• The last characteristic can be explained numerically by considering the effect of
•
•
raising a real number x to successively higher powers. If x is a fraction, that is, −1 < x < 1,
•
•
•
•
•
•
•
•
•
•
•
then the absolute value of x n decreases as n increases (try raising 12 or − 12 to higher and
•
•
•
•
•
higher powers, for example). This explains why successive graphs in Figure 1.6.5 become
flatter over the interval −1 < x < 1. On the other hand, if x > 1 or x < −1, then the
•
•
•
•
•
powers). This explains why successive graphs become steeper if x > 1 or x < −1.
•
•
•
•
••••••••••••••••••••••••••••••••••••••
If p is a negative integer, say p = −n, then the power functions f(x) = x p have the form
THE FAMILY y = x –n f(x) = x −n = 1/x n . Figure 1.6.6a shows the graphs of y = 1/x and y = 1/x 2 , and
Figure 1.6.6b shows how these graphs relate to other members of the family. The graph of
y = 1/x is called an equilateral hyperbola (for reasons to be discussed later).
January 12, 2001 11:10 g65-ch1 Sheet number 60 Page number 66 cyan magenta yellow black
66 Functions
y y
y = 1/x 3
y y y = 1/x 2
y = 1/x y = 1/x 4 y = 1/x y = 1/x 5
y= 1/x 2 y = 1/x 6
1
(1, 1)
x x
1
(–1, –1)
1 (–1, 1) (1, 1)
x x
1
(a) (b)
Figure 1.6.6
For odd values of n the graphs have the same general shape as y = 1/x, and for even
values of n they have the same general shape as y = 1/x 2 . The graphs in the family y = 1/x n
share a number of important characteristics:
• For even values of n the functions f(x) = 1/x n are even, and their graphs are symmetric
about the y-axis; for odd values of n the functions f(x) = x n are odd, and their graphs
are symmetric about the origin.
• For all values of n the graphs pass through the point (1, 1) and have a break (called a
discontinuity) at x = 0. This is caused by the division by zero that occurs when x = 0.
For even values of n the graphs pass through (−1, 1), and for odd values of n they pass
through (−1, −1).
• Increasing n causes the graph to become steeper over the intervals −1 < x < 0 and
0 < x < 1, and flatter over the intervals x > 1 and x < −1.
• REMARK.
• The last characteristic can be explained numerically by considering the effect of
•
•
•
•
•
•
•
raising the reciprocal of a number x to successively higher powers. If x is a nonzero fraction,
then it lies in the interval −1 < x < 1, and its reciprocal satisfies 1/x > 1 or 1/x < −1.
•
•
•
•
•
•
•
•
•
•
•
Thus, as n increases the absolute value of 1/x n also increases. This explains why successive
•
•
•
•
•
graphs in Figure 1.6.6 become successively steeper over the interval −1 < x < 1. On the
other hand, if x > 1 or x < −1, then −1 < 1/x < 1. Thus, as n increases the absolute value
•
•
•
•
•
•
•
•
•
•
of 1/x n decreases. This explains why successive graphs in Figure 1.6.6 get successively
flatter if x > 1 or x < −1.
•
•
•
•
••••••••••••••••••••••••••••••••••••••
If p = 1/n, where n is√a positive integer, then the power functions
√ f(x) = x have the
p
THE FAMILY y = x 1/n
√= x = x. In particular, if n = 2, then f(x) = x, and if n = 3, then
1/ n n
form f(x)
f(x) = 3 x. The graphs of these functions are shown in parts (a) and (b) of Figure 1.6.7.
y y y
y = √x 3 y = √x
y=√ x
(1, 1) x (1, 1) x x
y = –√x
√ √
Observe that the graph of y = 3 x extends over the entire x-axis because f(x) = 3 x is
defined for
√ all real values of x (every real number has a cube root); in contrast, the graph
of y = x only extends over the nonnegative x-axis √ (negative numbers have imaginary
square roots). Observe also that the graph of y = x is the upper half of the parabola
x = y 2 (Figure 1.6.7c). √ √
For even values of n the graphs of y = n x have the same general
√ shape as y = x,
and for odd values of n they have the same general shape as y = 3 x.
••••••••••••••••••••••••••••••••••••••
Power functions can also have fractional or irrational exponents. For example,
POWER FUNCTIONS WITH √ √
5
FRACTIONAL AND IRRATIONAL f(x) = x 2/3 , f(x) = x 3 , f(x) = x −7/8 , and f(x) = x 2 (3)
EXPONENTS
are all power functions of this type; we will discuss power functions of these forms in later
sections.
Read the note preceding Exercise 29 of Section 1.3, and use a graphing utility to generate
complete graphs of the functions in (3).
••••••••••••••••••••••••••••••••••••••
Recall that a variable y is said to be inversely proportional to a variable x if there is a
INVERSE PROPORTIONS positive constant k, called the constant of proportionality, such that
k
y= (4)
x
Since k is assumed to be positive, the graph of this equation has the same basic shape as
y = 1/x but is compressed or stretched in the y-direction.
Observe that in Formula (4) doubling x decreases y by a factor of 1/2, tripling x decreases
y by a factor of 1/3, and, more generally, increasing x by a factor of r decreases y by a
factor of 1/r.
Functions involving inverse proportion arise in various laws of physics. For example,
Boyle’s law in physics states that at a constant temperature the pressure P exerted by a
fixed quantity of an ideal gas is inversely proportional to the volume V occupied by the gas,
that is,
k
P =
V
(Figure 1.6.8).
If y is inversely proportional to x, then it follows from (4) that the product of y and x is
constant, since yx = k. This provides a useful way of identifying inverse proportionality
in experimental data.
68 Functions
Boyle's Law (P = k /V )
Pressure P
0
Temperature Volume V
control unit
y
10 As the volume of the gas changes,
y = 5x the temperature control unit adds or
9
8 removes heat to maintain a constant
temperature.
7
6 Figure 1.6.8
5
4
3 Table 1.6.1
2 experimental data
1
x x 0.8 1 2.5 4 6.25 10
1 2 3 4 5 6 7 8 9 10
y 6.25 5 2 1.25 0.8 0.5
Figure 1.6.9
••••••••••••••••••••••••••••••••••••••
A detailed review of polynomials is given in Appendix F, but for convenience we will review
A QUICK REVIEW OF POLYNOMIALS some of the terminology here.
A polynomial in x is a function that is expressible as a sum of finitely many terms
of the form cx n , where c is a constant and n is a nonnegative integer. Some examples of
polynomials are
√
2x + 1, 3x 2 + 5x − 2, x 3 , 4 (= 4x 0 ), 5x 7 − x 4 + 3
3
The function x 2 − 4 is also a polynomial because it can be expanded by the binomial
formula (see the inside front cover) and expressed as a sum of terms of the form cx n :
2 3 3 2
x − 4 = x 2 − 3 x 2 (4) + 3(x 2 )(42 ) − (43 ) = x 6 − 12x 4 + 48x 2 − 64 (5)
A general polynomial can be written in either of the following forms, depending on
whether one wants the powers of x in ascending or descending order:
c0 + c1 x + c2 x 2 + · · · + cn x n
cn x n + cn−1 x n−1 + · · · + c1 x + c0
The constants c0 , c1 , . . . , cn are called the coefficients of the polynomial. When a polyno-
mial is expressed in one of these forms, the highest power of x that occurs with a nonzero
coefficient is called the degree of the polynomial. Nonzero constant polynomials are con-
sidered to have degree 0, since we can write c = cx 0 . Polynomials of degree 1, 2, 3, 4, and
5 are described as linear, quadratic, cubic, quartic, and quintic, respectively. For example,
3 + 5x Has degree 1 (linear)
√
3 + x3 + x5 Has degree 5 (quintic)
2 3
x −4 Has degree 6 [see (5)]
January 12, 2001 11:10 g65-ch1 Sheet number 63 Page number 69 cyan magenta yellow black
The natural domain of a polynomial in x is (−⬁, +⬁), since the only operations involved
are multiplication and addition; the range depends on the particular polynomial. We already
know that the graphs of polynomials of degree 0 and 1 are lines and that the graphs of
polynomials of degree 2 are parabolas. Figure 1.6.10 shows the graphs of some typical
polynomials of higher degree. Later, we will discuss polynomial graphs in detail, but for
now it suffices to observe that graphs of polynomials are very well behaved in the sense that
they have no discontinuities or sharp corners. As illustrated in Figure 1.6.10, the graphs of
polynomials wander up and down for awhile in a roller-coaster fashion, but eventually that
behavior stops and the graphs steadily rise or fall indefinitely as one travels along the curve
in either the positive or negative direction. We will see later that the number of peaks and
valleys is less than the degree of the polynomial.
y y y y
x x x x
Figure 1.6.10
••••••••••••••••••••••••••••••••••••••
A function that can be expressed as a ratio of two polynomials is called a rational function.
RATIONAL FUNCTIONS If P (x) and Q(x) are polynomials, then the domain of the rational function
P (x)
f(x) =
Q(x)
consists of all values of x such that Q(x) = 0. For example, the domain of the rational
function
x 2 + 2x
f(x) = 2
x −1
consists of all values of x, except x = 1 and x = −1. Its graph is shown in Figure 1.6.11
along with the graphs of two other typical rational functions.
y y
y
4 3
4
x x x
-5 5 -5 7 -4 4
-3
-4
x 2 + 2x x2 – 1 3
y= y= y=
x2 – 1 x 2 – 2x – 3 x2 + 1
Figure 1.6.11
The graphs of rational functions with nonconstant denominators differ from the graphs
of polynomials in some essential ways:
• Unlike polynomials whose graphs are continuous (unbroken) curves, the graphs of
rational functions have discontinuities at the points where the denominator is zero.
January 12, 2001 11:10 g65-ch1 Sheet number 64 Page number 70 cyan magenta yellow black
70 Functions
• Unlike polynomials, rational functions may have numbers at which they are not defined.
Near such points, many (but not all) rational functions have graphs that approximate a
vertical line, called a vertical asymptote. These are represented by the dashed vertical
lines in Figure 1.6.11.
• Unlike the graphs of polynomials, which eventually rise or fall indefinitely, the graphs
of many (but not all) rational functions eventually get closer and closer to some hori-
zontal line, called a horizontal asymptote, as one travels along the curve in either the
positive or negative direction. The horizontal asymptotes are represented by the dashed
horizontal lines in the first two parts of Figure 1.6.11; in the third part of the figure the
x-axis is a horizontal asymptote.
••••••••••••••••••••••••••••••••••••••
Functions that can be constructed from polynomials by applying finitely many algebraic op-
ALGEBRAIC FUNCTIONS erations (addition, subtraction, division, and root extraction) are called algebraic functions.
Some examples are
√
f(x) = x 2 − 4, f(x) = 3 3 x(2 + x), f(x) = x 2/3 (x + 2)2
As illustrated in Figure 1.6.12, the graphs of algebraic functions vary widely, so it is difficult
to make general statements about them. Later in this text we will develop general calculus
methods for analyzing such functions.
4
y y
3
5 15
4 2
10
3
5 1
2 x
1 x
x -3 -2 -1 1 2
-5 -4 -3 -2 -1 1
-5 -4 -3 -2 -1 1 2 3 4 5
y = √x 2 – 4
3
y = 3√x (2 + x) y = x 2/3(x + 2)2
Figure 1.6.12
••••••••••••••••••••••••••••••••••••••
A detailed review of trigonometric functions is given in Appendix E, but for convenience
A QUICK REVIEW OF we will summarize some of the main ideas here.
TRIGONOMETRIC FUNCTIONS
It is often convenient to think of the trigonometric functions in terms of circles rather
than triangles. For this purpose, consider a point that moves either clockwise or counter-
clockwise along the unit circle u2 +v 2 = 1 in the uv-plane, starting at (1, 0) and stopping at
a point P (Figure 1.6.13a). Let x denote the signed arc length traveled by the moving point,
taking x to be positive for counterclockwise motion and negative for clockwise motion.
(We allow for the possibility that the point may traverse the circle more than once.) When
convenient, the variable x can also be interpreted as the angle in radians that is swept out
by the radial line from the origin to P , with the usual convention that angles are positive if
generated by counterclockwise rotations and negative if generated by clockwise rotations.
We can define cos x to be the u-coordinate of P and sin x to be the v-coordinate of P
(Figure 1.6.13b).
The remaining trigonometric functions can be defined in terms of the functions sin x and
cos x:
sin x cos x
tan x = cot x =
cos x sin x
1 1
sec x = csc x =
cos x sin x
The graphs of the six trigonometric functions in Figure 1.6.14 should already be familiar
January 12, 2001 11:10 g65-ch1 Sheet number 65 Page number 71 cyan magenta yellow black
v v v
P P(cos x, sin x)
x
u u u
(1, 0) (1, 0) (1, 0)
x
P(cos x, sin x)
(a) (b)
Figure 1.6.13
y
1
x y
–2π –π π 2π
-1
y = sin x x
I C ^ 6 c i
y
1
x
–2π –π π 2π
y = tan x
-1
y = cos x
y y y
x x x
O C c o OI C ^ 6 c i o O C c o
Figure 1.6.14
to you, but try generating them using a graphing utility, making sure to use radian measure
for x.
• REMARK.
• In this text we will always assume that the independent variable in a trigono-
•
•
•
•
•
•
• metric function is in radians unless specifically stated otherwise.
••••••••••••••••••••••••••••••••••••••
Many of the basic properties of sin x and cos x can be deduced from the circle definitions
PROPERTIES OF sin x, cos x, of these functions. For example:
AND tan x
• As the point P (cos x, sin x) moves around the unit circle, its coordinates vary between
−1 and 1, and hence
−1 ≤ sin x ≤ 1 and −1 ≤ cos x ≤ 1
January 12, 2001 11:10 g65-ch1 Sheet number 66 Page number 72 cyan magenta yellow black
72 Functions
• If x increases or decreases by 2π radians, then the point P (cos x, sin x) makes one
complete revolution around the unit circle, and the coordinates return to their starting
values. Thus, sin x and cos x have period 2π; that is,
sin(x ± 2π) = sin x
cos(x ± 2π) = cos x
• As P (cos x, sin x) moves around the unit circle, sin x is zero when P is on the horizontal
axis (which occurs when x is an integer multiple of π), and cos x is zero when P is on
the vertical axis (which occurs when x is an odd multiple of π/2). Thus,
sin x = 0 if and only if x = 0, ±π, ±2π, ±3π, . . .
cos x = 0 if and only if x = ±π/2, ±3π/2, ±5π/2, . . .
• As P (cos x, sin x) moves around the unit circle u2 + v 2 = 1, its coordinates satisfy this
equation for all x, which produces the fundamental trigonometric identity
v
L cos2 x + sin2 x = 1
P(cos x, sin x) Observe that the graph of y = tan x has vertical asymptotes at the points x = ±π/2,
1 ±3π/2, ±5π/2, . . . . This is to be expected since tan x = sin x / cos x, and these are the
sin x
x u values of x at which cos x is zero. What is less obvious, however, is the fact that tan x repeats
cos x every π radians (i.e., has period π), even though sin x and cos x have period 2π. This can
be explained by interpreting
sin x
tan x =
cos x
as the slope of the line L that passes through the origin and the point P (cos x, sin x) on the
Slope of L =
sin x
= tan x unit circle in the uv-plane (Figure 1.6.15). Each time x increases or decreases by π radians,
cos x
the point P traverses half the circumference, and the line L rotates π radians, so its starting
Figure 1.6.15 and ending slope are the same.
••••••••••••••••••••••••••••••••••••••
The choice of radian measure as opposed to degree measure depends on the nature of the
RADIANS AS A DIMENSIONLESS problem being considered; degree measure is usually chosen in engineering problems in-
UNIT
volving measurements of angles, and radian measure is usually chosen when the function
properties of sin x, cos x, tan x, . . . are the primary focus. Radian measure is also usually
s chosen in problems involving arc lengths on circles because of the basic result in trigonom-
etry which states that the arc length s of a sector with radius r and a central angle of θ
u
(radians) is given by
r
s = rθ (6)
(Figure 1.6.16).
In applications involving angles, radians require special treatment to ensure that quanti-
If u is in radians,
then s = ru.
ties are assigned proper units. To see why this is so, let us rewrite (6) as
s
Figure 1.6.16 θ=
r
s The left side of this equation is in radians, and the right side is the ratio of two lengths, say
meters/meters or feet/feet. However, because these units of length cancel, the right side of
this equation is actually dimensionless (has no units). Thus, to ensure consistency between
the two sides of the equation, we would have to omit the units of radians on the left side
r to make it dimensionless as well. In practical terms this means that units of radians can be
u used in intermediate computations, when convenient, but they need to be omitted in the end
result to ensure consistency of units. This is confusing, to say the least, but the following
example should clarify the idea.
Example 3 Suppose that two satellites circle the equator in an orbit whose radius is
r = 4.23 × 107 m (Figure 1.6.17). Find the arc length s that separates the satellites if they
Figure 1.6.17 have an angular separation of θ = 2.00 ◦ .
January 12, 2001 11:10 g65-ch1 Sheet number 67 Page number 73 cyan magenta yellow black
Solution. To apply Formula (6), we must convert the angular separation to radians:
π
2.00 ◦ = (2.00) ≈ 0.0349 rad
180
Thus, from (6)
s = rθ ≈ (4.23 × 107 m)(0.0349 rad) ≈ 1.48 × 106 m
In this computation the product rθ produces units of meters × radians, but if we treat radians
as dimensionless, we have meters × radians = meters, which correctly produces units of
meters (m) for the arc length s.
••••••••••••••••••••••••••••••••••••••
Many important applications lead to trigonometric functions of the form
THE FAMILIES y = A sin Bx
AND y = A cos Bx f(x) = A sin(Bx − C) and g(x) = A cos(Bx − C) (7)
where A, B, and C are nonzero constants. The graphs of such functions can be obtained by
stretching, compressing, translating, and reflecting the graphs of y = sin x and y = cos x
appropriately. To see why this is so, let us start with the case where C = 0 and consider
how the graphs of the equations
y = A sin Bx and y = A cos Bx
relate to the graphs of y = sin x and y = cos x. If A and B are positive, then the effect of
the constant A is to stretch or compress the graphs of y = sin x and y = cos x vertically
by a factor of A, and the effect of the constant B is to compress or stretch the graphs of
sin x and cos x horizontally by a factor of B. For example, the graph of y = 2 sin 4x can be
obtained by stretching the graph of y = sin x vertically by a factor of 2 and compressing it
horizontally by a factor of 4. (Recall from Section 1.4 that the multiplier of x stretches when
it is less than 1 and compresses when it is greater than 1.) Thus, as shown in Figure 1.6.18,
the graph of y = 2 sin 4x varies between −2 and 2, and repeats every 2π/4 = π/2 units.
2
y = 2 sin 4x
1
y = sin x
x
-2π -π π 2π
-1
-2
Figure 1.6.18
Example 4 Make sketches of the following graphs that show the period and amplitude.
(a) y = 3 sin 2πx (b) y = −3 cos 0.5x (c) y = 1 + sin x
January 12, 2001 11:10 g65-ch1 Sheet number 68 Page number 74 cyan magenta yellow black
74 Functions
Solution (a). The equation is of the form y = A sin Bx with A = 3 and B = 2π,
so the graph has the shape of a sine function, but with amplitude A = 3 and period
2π/B = 2π/2π = 1 (Figure 1.6.19a).
y y y
3 3
2
Amplitude Amplitude Amplitude
x x
1
1 π
x
-3 -3 C c o
Period Period
(a) (b) (c)
Figure 1.6.19
Solution (b). The equation is of the form y = A cos Bx with A = −3 and B = 0.5, so the
graph has the shape of a cosine function that has been reflected about the x-axis (because
A = −3 is negative), but with amplitude |A| = 3 and period 2π/B = 2π/0.5 = 4π
(Figure 1.6.19b).
Solution (c). The graph has the shape of a sine function that has been translated up 1 unit
(Figure 1.6.19c).
••••••••••••••••••••••••••••••••••••••
To investigate the graphs of the more general families
THE FAMILIES y = A sin(Bx – C)
AND y = A cos(Bx – C) y = A sin(Bx − C) and y = A cos(Bx − C)
it will be helpful to rewrite these equations as
C C
y = A sin B x − and y = A cos B x −
B B
In this form we see that the graphs of these equations can be obtained by translating the
graphs of y = A sin Bx and y = A cos Bx to the left or right, depending on the sign of
C /B. For example, if C /B > 0, then the graph of
y = A sin[B(x − C /B)] = A sin(Bx − C)
can be obtained by translating the graph of y = A sin Bx to the right by C /B units
(Figure 1.6.20). If C /B < 0, the graph of y = A sin(Bx − C) is obtained by translating the
graph of y = A sin Bx to the left by |C /B| units
y C/B
Amplitude = A
x
y = Asin (Bx – C )
y = A sin Bx
Figure 1.6.20
January 12, 2001 11:10 g65-ch1 Sheet number 69 Page number 75 cyan magenta yellow black
1. (a) Find an equation for the family of lines whose members 8. Find an equation for the family of lines that pass through
have slope m = 3. the intersection of 5x − 3y + 11 = 0 and 2x − 9y + 7 = 0.
(b) Find an equation for the member of the family that 9. The U.S. Internal Revenue Service uses a 10-year linear de-
passes through (−1, 3). preciation schedule to determine the value of various busi-
(c) Sketch some members of the family, and label them ness items. This means that an item is assumed to have a
with their equations. Include the line in part (b). value of zero at the end of the tenth year and that at inter-
2. Find an equation for the family of lines whose members are mediate times the value is a linear function of the elapsed
perpendicular to those in Exercise 1. time. Sketch some typical depreciation lines, and explain
3. (a) Find an equation for the family of lines with y-intercept the practical significance of the y-intercepts.
b = 2. 10. Find all lines through (6, −1) for which the product of the
(b) Find an equation for the member of the family whose x- and y-intercepts is 3.
angle of inclination is 135 ◦ . 11. In each part, match the equation with one of the accompa-
(c) Sketch some members of the family, and label them nying graphs.
with their equations. Include the line in part (b). √
(a) y = 5 x (b) y = 2x 5
4. Find an equation for
(c) y = −1/x 8 (d) y = x 2 − 1
(a) the family of lines that pass through the origin √
4
√5
(b) the family of lines with x-intercept a = 1 (e) y = x − 2 (f ) y = − x 2
(c) the family of lines that pass through the point (1, −2)
(d) the family of lines parallel to 2x + 4y = 1.
y y y
In Exercises 5 and 6, state a geometric property common to
all lines in the family, and sketch five of the lines.
x x x
76 Functions
12. The accompanying table gives approximate values of three 21. Sketch the graph of y = x 2 + 2x by completing the square
functions: one of the form kx 2 , one of the form kx −3 , and and making appropriate transformations to the graph of
one of the form kx 3/2 . Identify which is which, and estimate y = x2.
k in each case. √
√ graph of y = x to help sketch the graph of
22. (a) Use the
y = |x|.
√
x 0.25 0.37 2.1 4.0 5.8 6.2 7.9 9.3 (b) Use the
√ graph of y = 3 x to help sketch the graph of
f (x) 640 197 1.08 0.156 0.0513 0.0420 0.0203 0.0124 y = 3 |x|.
g(x) 0.0312 0.0684 2.20 8.00 16.8 19.2 31.2 43.2 23. As discussed in this section, Boyle’s law states that at a con-
h(x) 0.250 0.450 6.09 16.0 27.9 30.9 44.4 56.7 stant temperature the pressure P exerted by a gas is related
to the volume V by the equation P = k /V .
Table Ex-12 (a) Find the appropriate units for the constant k if pressure
(which is force per unit area) is in newtons per square
meter (N/m2 ) and volume is in cubic meters (m3 ).
In Exercises 13 and 14, sketch the graph of the equation for
n = 1, 3, and 5 in one coordinate system and for n = 2, 4, (b) Find k if the gas exerts a pressure of 20,000 N/m2 when
and 6 in another coordinate system. Check your work with a the volume is 1 liter (0.001 m3 ).
graphing utility. (c) Make a table that shows the pressures for volumes of
0.25, 0.5, 1.0, 1.5, and 2.0 liters.
13. (a) y = −x n (b) y = 2x −n (c) y = (x − 1)1/n (d) Make a graph of P versus V .
14. (a) y = 2x n (b) y = −x −n 24. A manufacturer of cardboard drink containers wants to con-
(c) y = −3(x + 2)1/n struct a closed rectangular container that has a square base
1
15. (a) Sketch the graph of y = ax 2 for a = ±1, ±2, and ±3 and will hold 10 liter (100 cm3 ). Estimate the dimension of
in a single coordinate system. the container that will require the least amount of material
(b) Sketch the graph of y = x 2 + b for b = ±1, ±2, and for its manufacture.
±3 in a single coordinate system.
(c) Sketch some typical members of the family of curves A variable y is said to be inversely proportional to the square
y = ax 2 + b. of a variable x if y is related to x by an equation of the form
√ y = k /x 2 , where k is a nonzero constant, called the constant
16. (a) Sketch the graph of y = a x for a = ±1, ±2, and ±3
of proportionality. This terminology is used in Exercises 25
in a single coordinate system.
√ and 26.
(b) Sketch the graph of y = x + b for b = ±1, ±2, and
±3 in a single coordinate system. 25. According to Coulomb’s law, the force F of attraction be-
(c) Sketch√some typical members of the family of curves tween positive and negative point charges is inversely pro-
y = a x + b. portional to the square of the distance x between them.
(a) Assuming that the force of attraction between two point
In Exercises 17–20, sketch the graph of the equation by mak- charges is 0.0005 newton when the distance between
ing appropriate transformations to the graph of a basic power them is 0.3 meter, find the constant of proportionality
function. Check your work with a graphing utility. (with proper units).
17. (a) y = 2(x + 1)2 (b) y = −3(x − 2)3 (b) Find the force of attraction between the point charges
−3 1 when they are 3 meters apart.
(c) y = (d) y =
(x + 1)2 (x − 3)5 (c) Make a graph of force versus distance for the two
√ √3 charges.
18. (a) y = 1 − x + 2 (b) y = 1 − x + 2
5 2 (d) What happens to the force as the particles get closer and
(c) y = (d) y = closer together? What happens as they get farther and
(1 − x)3 (4 + x)4
√3 √ farther apart?
19. (a) y = x + 1 (b) y = 1 − x − 2
26. It follows from Newton’s Law of Universal Gravitation that
x+1
(c) y = (x − 1) + 2 5
(d) y = the weight W of an object (relative to the Earth) is inversely
x proportional to the square of the distance x between the
1 1 object and the center of the Earth, that is, W = C /x 2 .
20. (a) y = 1 + (b) y =
x−2 1 + 2x − x 2 (a) Assuming that a weather satellite weighs 2000 pounds
2 on the surface of the Earth and that the Earth is a sphere
(c) y = − (d) y = x 2 + 2x
x7 of radius 4000 miles, find the constant C.
January 12, 2001 11:10 g65-ch1 Sheet number 71 Page number 77 cyan magenta yellow black
π π 2π
(b) Find the weight of the satellite when it is 1000 miles 29. (a) (b) − (c)
above the surface of the Earth. 3 2 3
(c) Make a graph of the satellite’s weight versus its distance 5π 11π
(d) −1 (e) (f )
from the center of the Earth. 4 6
(d) Is there any distance from the center of the Earth at 3π 5π
which the weight of the satellite is zero? Explain your 30. (a) (b) − (c) π
2 4
reasoning. 5π 33π
(d) (e) 4 (f ) −
27. In each part, match the equation with one of the accompa- 2 7
nying graphs, and give the equations for the horizontal and
vertical asymptotes. In Exercises 31 and 32, use a calculating utility set to the ra-
x2 x−1 dian mode to confirm the approximations sin(π/5) ≈ 0.588
(a) y = 2 (b) y = 2
x −x−2 x −x−6 and cos(π/8) ≈ 0.924, and then use these values to approx-
2x 4 4 imate the given expressions by hand calculation. Check your
(c) y = 4 (d) y =
x +1 (x + 2)2 answers using the trigonometric function operations of your
calculating utility.
y y
4π π 11π
31. (a) sin (b) cos − (c) sin
5 8 5
x
x 7π π 2π
(d) cos (e) cos2 (f ) sin2
8 5 5
16π 17π 41π
32. (a) sin (b) cos − (c) sin
I II 5 8 5
π 27π π
(d) sin − (e) cos (f ) tan2
y 16 8 8
y 33. Assuming that sin α = a, cos β = b, and tan γ = c, express
the stated quantities in terms of a, b, and c.
(a) sin(−α) (b) cos(−β) (c) tan(−γ )
x
π
x
(d) sin −α (e) cos(π − β) (f ) sin(α + π)
2
III IV Figure Ex-27 (g) sin(2β) (h) cos(2β) (i) sec(β + 2π)
β
(j) csc(α + π) (k) cot(γ + 5π) (l) sin2
2
28. Find an equation of the form y = k /(x 2 + bx + c) whose
graph is a reasonable match to that in the accompanying 34. A ship travels from a point near Hawaii at 20 ◦ N latitude
figure. Check your work with a graphing utility. directly north to a point near Alaska at 56 ◦ N latitude.
(a) Assuming the Earth to be a sphere of radius 4000 mi,
find the actual distance traveled by the ship.
y
(b) What fraction of the Earth’s circumference did the ship
travel?
35. The Moon completes one revolution around the Earth in ap-
x proximately 27.3 days. Assuming that the Moon’s orbit is a
–1 –1 1 circle with a radius of 0.38 × 109 m from the center of the
Earth, find the arc length traveled by the Moon in 1 day.
36. A spoked wheel with a diameter of 3 ft rolls along a flat road
without slipping. How far along the road does the wheel roll
Figure Ex-28 if the spokes turn through 225 ◦ ?
37. As illustrated in the accompanying figure, suppose that you
hold one quarter flat against a table while you rotate a second
In Exercises 29 and 30, draw a radial line from the origin with
quarter around it without slippage. Through what angle will
the given angle, and determine whether the six trigonometric
the second quarter have turned about its own center when it
functions are positive, negative, or undefined for that angle.
returns to its original location?
January 12, 2001 11:10 g65-ch1 Sheet number 72 Page number 78 cyan magenta yellow black
78 Functions
41. In each part, find an equation for the graph that has the form
y = y0 + A sin(Bx − C).
y y
1 6
x
o
x
-1 9p
Figure Ex-37
(a) (b)
In this section we will introduce some simple mathematical models that are based on
linear, quadratic, and trigonometric functions. The method of “least squares” will be
discussed briefly within the context of linear and quadratic regression.
••••••••••••••••••••••••••••••••••••••
A mathematical model of a physical law or phenomenon is a description of that law or
MATHEMATICAL MODELS phenomenon in the language of mathematics. The modern scientific view about mathe-
matical models was first expressed in the seventeenth century by the Italian mathematician
and scientist Galileo Galilei (1564–1642) who wrote, “The book of nature is written in
mathematics.” This seemingly innocuous statement represented a major departure from the
approach of the ancient Greek scientists and philosophers, who generally described sci-
entific laws in words. For example, the Greek philosopher Aristotle (384 B.C.–322 B.C.)
would have described the motion of a falling body by saying that such bodies “seek their
natural position”, whereas today we would use functions, equations, or other mathematical
means to describe the motion precisely.
Mathematical models allow us to use mathematical models to deduce results about
the physical world that are not evident or have never been observed. For example, the
possibility of placing a satellite in orbit around the Earth was deduced mathematically from
Issac Newton’s model of mechanics nearly 200 years before the launching of Sputnik, and
Albert Einstein (1879–1955) gave a relativistic model of mechanics in 1915 that explained
a precession (position shift) in the perihelion of the planet Mercury that was not confirmed
by physical measurement until 1967.
One of the most important steps in creating a mathematical model of a physical phe-
nomenon is deciding which factors to consider and which to ignore—the more factors one
takes into account the more complicated the formulas and equations of the model tend to
become, so there is always a balance to be struck between keeping a model mathematically
simple and considering enough factors to make the model useful. For example, if a meteo-
rologist were trying to model the relationship between the speed of a raindrop when it hits
the ground and the height of the cloud in which it was formed, then he or she would certainly
have to take air resistance into account, but with equal certainty he or she would ignore
the infinitesimal effect that the Pluto’s gravitational pull has on the raindrop. The danger
is that in trying to keep a mathematical model from becoming too complicated one might
oversimplify to the point where the results it produces do not agree with reality. We are re-
minded of this by Einstein’s admonition: “Everything must be as simple as possible, but not
simpler.” A good mathematical model is one that produces results that are consistent with
the physical world. If a time comes when the mathematical results produced by the model
do not agree with real-world observations, then the model must be abandoned or modified
in favor of a new model that does. This is the nature of the scientific method—old models
constantly being replaced by new models that more accurately describe the real world.
••••••••••••••••••••••••••••••••••••••
In this section we will consider some simple models that involve only two variables. In our
FUNCTIONS AS MODELS general discussion we will refer to these variables as x and y, but in specific examples other
letters will be more appropriate. We will assume that the data for the phenomenon being
modeled consists of a collection of ordered pairs of measurements
(x1 , y1 ), (x2 , y2 ), (x3 , y3 ), . . . , (xn , yn )
that relate corresponding values of the variables x and y. We distinguish between two types
of phenomena—deterministic phenomena in which each value of x determines one value
of y and probabilistic phenomena in which the value of y associated with a specific x is
not uniquely determined, but rather depends on probabilities in some way. For example,
if y is the amount that a spring is stretched by a force x, then for a given spring the value
of y is uniquely determined by the value of x, so this is a deterministic phenomenon. In
contrast, if y is the weight of person whose height is x, then the value of y is not uniquely
January 12, 2001 11:10 g65-ch1 Sheet number 74 Page number 80 cyan magenta yellow black
80 Functions
y determined by x, but there is a correlation between weight and height that makes it more
50
likely for a taller person to weigh more, so this is a probabilistic phenomenon.
In a deterministic model the variable y is a function of x, so the goal is to use the data in
40 some reasonable way to find a formula y = f(x) for this function. For example, Figure 1.7.1
30 strongly suggests that the relationship between x and y is linear, so in absence of additional
information it would be natural to look for a linear function y = mx + b as a model. In a
20
probabilistic model the variable y need not be a function of x, so the goal is to find an equation
10 y = f(x) that specifies the average value of y that can be expected to occur for a given x.
x
A more precise explanation of what is meant by “average value” and “expected to occur”
1 2 3 4 5
requires ideas from probability and statistics, so we will depend on your intuition here.
Figure 1.7.1
••••••••••••••••••••••••••••••••••••••
Let us suppose that we have decided to model a particular phenomenon with a (yet to
LINEAR FUNCTIONS AS MODELS be determined) linear function y = mx + b. Ideally, we would then like to choose the
parameters m and b such that the line y = mx + b passes through all our data points. In
y practice, this may be impossible, either because of errors in our measurements or because
there is not a strictly linear relationship between the variables x and y. We are then faced
y = mx + b
with the problem of finding the line y = mx + b that “best fits” our set of data points.
The key to determining this line is based on the following idea: For any proposed linear
yi – (mxi + b) function y = mx + b, we draw a vertical connector from each data point (xi , yi ) to the point
(xi, yi) (xi , mxi + b) on the line and consider the differences yi − (mxi + b) (Figure 1.7.2). These
differences, which are called residuals, may be viewed as “errors” that result when the line
is used to model the data. Data points above the line have positive errors, those below the
x line have negative errors, and those on the line have no error. One of the most common
xi procedures is to look for a line such that the sum of the squares of the residuals is as small
Figure 1.7.2 as possible. This line, known as a least squares line or linear regression line, is one choice
for a line that “best fits” a given set of data. Most graphing calculators, spreadsheets, and
CAS programs provide methods for finding regression lines. We will assume that you have
access to some such method in this section.
It is possible to compute a regression line, even in cases where the data have no appar-
ent linear pattern. Thus, it is important to have some quantitative method of determining
whether a linear model is appropriate for the data. The most common measure of linearity
in data is called the correlation coefficient. Following convention, we denote the correla-
tion coefficient by the letter r. Although a detailed discussion of correlation coefficients is
beyond the scope of this text, here are some of the basic facts:
• The values of r are in the interval −1 ≤ r ≤ 1, where r has the same sign as the slope
of the regression line.
• If r is equal to 1 or −1, then the data points all lie on a line, so a linear model is a perfect
fit for the data.
• If r = 0, then the data points exhibit no linear tendency, so a linear model is inappropriate
for the data.
The closer r is to 1 or −1, the more tightly the data points hug the regression line and the
more appropriate the regression line is as a model; the closer r is to 0, the more scattered
the points and the less appropriate the regression line is as a model (Figure 1.7.3).
y y y y
x x x x
r = –1 r=1 r = 0.7 r=0
Figure 1.7.3
January 12, 2001 11:10 g65-ch1 Sheet number 75 Page number 81 cyan magenta yellow black
Roughly stated, the value of r 2 is a measure of the percentage of data points that fall in
a “tight linear band.” Thus r = 0.5 means that 25% of the points fall in a tight linear band,
and r = 0.9 means that 81% of the points fall in a tight linear band. (A precise explanation
of what is meant by a “tight linear band” requires ideas from statistics.)
Table 1.7.1 Example 1 Table 1.7.1 gives a set of data points relating the pressure p in atmospheres
(atm) and the temperature T (in ◦ C) of a fixed quantity of carbon dioxide in a closed cylinder.
temperature pressure
T (°C) p (atm) The associated scatter plot in Figure 1.7.4a suggests that there is a linear relationship between
the pressure and the temperature.
0 2.54
50 3.06
4.5 4.5
100 3.46
150 4.00 4.0 4.0
Pressure p (atm)
Pressure p (atm)
200 4.41
3.5 3.5
3.0 3.0
2.5 2.5
(a) Use your calculating utility to find the least squares line for the data. If your utility
can produce the correlation coefficient, then find it.
(b) Use the model obtained in part (a) to predict the pressure when the temperature is
250 ◦ C.
(c) Use the model obtained in part (a) to predict a temperature at which the pressure of
the gas will be zero.
Solution (a). The least squares line is given by p = 0.00936T + 2.558 (Figure 1.7.4b)
with correlation coefficient r = 0.999475.
Example 2 Figure 1.7.5a shows a graph of temperature versus altitude that was trans-
mitted by the Magellan spacecraft when it entered the atmosphere of Venus in October
1991. The graph strongly suggests that there is a linear relationship between temperature
and altitude for altitudes between 35 km and 60 km.
(a) Use the graph transmitted by the Magellan spacecraft to find a linear model of tem-
perature versus altitude in the Venusian atmosphere that is valid for altitudes between
35 km and 60 km.
(b) Use the model obtained in part (a) to estimate the temperature at the surface of Venus,
and discuss the assumptions you are making in obtaining the estimate.
Solution (a). Let T be the temperature in kelvins and h the altitude in kilometers. We will
first estimate the slope m of the linear portion of the graph, then estimate the coordinates
January 12, 2001 11:10 g65-ch1 Sheet number 76 Page number 82 cyan magenta yellow black
82 Functions
of a data point (h1 , T1 ) on that portion of the graph, and then use the point-slope form of a
line
T − T1 = m(h − h1 ) (1)
The graph nearly passes through the point (60, 250), so we will take h1 ≈ 60 and T1 ≈ 250.
In Figure 1.7.5b we have sketched a line that closely approximates the linear portion of the
data. Using the intersections of that line with the edges of the grid box, we estimate the
slope to be
100 − 490 390
m≈ =− = −8.125 K/km
78 − 30 48
Substituting our estimates of h1 , T1 , and m into (1) yields the equation
T − 250 = −8.125(h − 60)
or equivalently,
T = −8.125h + 737.5 (2)
Solution (b). The Magellan spacecraft stopped transmitting data at an altitude of approxi-
mately 35 km, so we cannot be certain that the linear model applies at lower altitudes. How-
ever, if we assume that the model is valid at all lower altitudes, then we can approximate the
temperature at the surface of Venus by setting h = 0 in (2). We obtain T ≈ 737.5 K.
• REMARK.
• The method of the preceding example is crude, at best, since it relies on
•
•
•
•
•
•
•
•
extracting rough estimates of numerical data from a graph. Nevertheless, the final result is
•
•
•
•
•
quite good, since the most recent information from NASA places the surface temperature
•
•
•
• of Venus at about 740 K (hot enough to melt lead).
500 500
Temperature of Venusian Atmosphere Temperature of Venusian Atmosphere
450 Magellan orbit 3213 450 Magellan orbit 3213
Date: 5 October 1991 Date: 5 October 1991
Latitude: 67 N Latitude: 67 N
400 LTST: 22:05 400 LTST: 22:05
Temperature T (K)
Temperature T (K)
350 350
300 300
250 250
200 200
150 150
100 100
30 40 50 60 70 80 90 100 30 40 50 60 70 80 90 100
Altitude h (km) Altitude h (km)
Source: NASA Source: NASA
(a) (b)
Figure 1.7.5
••••••••••••••••••••••••••••••••••••••
Although models based on linear functions y = mx +b are simple, the relationship between
QUADRATIC AND TRIGONOMETRIC the variables x and y associated with a particular phenomenon may be nonlinear, in which
FUNCTIONS AS MODELS
case replacing the function y = mx + b by the quadratic function y = ax 2 + bx + c may
provide a better model. Most calculators, spreadsheets, and CAS programs will perform a
least squares quadratic regression in a manner that is similar to linear regression.
January 12, 2001 11:10 g65-ch1 Sheet number 77 Page number 83 cyan magenta yellow black
Table 1.7.2 Example 3 A student in a physics lab is studying the equations of motion of a falling
time t (s) height h (cm) body. She collects the data displayed in Table 1.7.2, which gives the height of the object at
a number of times over a 0.15-s period of time. She knows that if air resistance is negligible
0.008333 98.4 and if the acceleration of the object due to gravity is assumed to be constant, then the height
0.025 96.9 h of the object should be a quadratic function of time t. A scatter plot of the data is provided
0.04167 95.1
in Figure 1.7.6a, which suggests a portion of an inverted parabola.
0.05833 92.9
0.075 90.8
0.09167 88.1 100 100
0.10833 85.3 95 95
0.125 82.1
(cm)
Height h (cm)
0.14167 78.6 90 90
0.15833 74.9
Height h
85 85
80 80
75 75
(a) Determine the quadratic regression curve for the data in Table 1.7.2.
(b) According to the model obtained in part (a), when will the object strike the ground?
Solution (a). Using the quadratic regression routine on a calculator, we find that the
quadratic curve that best fits the data in Table 1.7.2 has equation
h = 99.02 − 73.21t − 499.13t 2
Figure 1.7.6b shows the data points and the graph of this quadratic function on the same set
of axes. It appears that we have excellent agreement between our curve and the data.
Solution (b). Solving the equation 0 = h = 99.02 − 73.21t − 499.13t 2 , we find that the
object will strike the ground at t ≈ 0.38 s.
The trigonometric functions y = A sin(Bx −C) and y = A cos(Bx −C) are particularly
useful for modeling periodic phenomena.
Example 4 Figure 1.7.7a shows a table and scatter plot of temperature data recorded
∗
over a 24-hour period in the city of Philadelphia. Find a function that models the data, and
graph your function and data together.
Solution. The pattern of the data suggests that the relationship between the temperature
T and the time t can be modeled by a sinusoidal function that has been translated both
horizontally and vertically, so we will look for an equation of the form
C
T = D + A sin[Bt − C] = D + A sin B t − (3)
B
Since the highest temperature is 95 ◦ F and the lowest temperature is 75 ◦ F, we take 2A = 20
or A = 10. The midpoint between the high and low is 85 ◦ F, so we have a vertical shift of
D = 85. The period seems to be about 24, so 2π/B = 24 or B = π/12. The horizontal
shift appears to be about 10 (verify), so C /B = 10. Substituting these values in (3) yields
∗
This example is based on the article “Everybody Talks About It!—Weather Investigations,” by Gloria S. Dion
and Iris Brann Fetta, The Mathematics Teacher, Vol. 89, No. 2, February 1996, pp. 160–165.
January 12, 2001 11:10 g65-ch1 Sheet number 78 Page number 84 cyan magenta yellow black
84 Functions
philadelphia temperatures
from 1:00 a.m. to 12:00 midnight on 27 august 1993
(t = hours after midnight and T = degrees fahrenheit)
A.M. P.M.
t T t T
the equation
π
T = 85 + 10 sin (t − 10)
12
(Figure 1.7.7b).
Note that in Example 4 we did not use a regression routine to fit the curve to the data.
Some calculators may not be equipped to compute regression for trigonometric functions.
In this case, we can use the calculator’s graphing capability to see that a proposed model
gives a reasonable fit to the data points, though it may not be the best fit.
1. One of the lines in the accompanying figure is the regression 2. Conjecture an appropriate model (linear, quadratic, or trig-
line. Which one is it? onometric), if any, for each of the data sets shown in the
scatter plots.
y
I
7
y
6 II 14
5 12
10
4 8
3 6
4
2
2
x
1
x 2 4 6 8 10 12 14
1 2 3 4 5 6 7 Figure Ex-1 (a)
January 12, 2001 11:10 g65-ch1 Sheet number 79 Page number 85 cyan magenta yellow black
y
5. A 20-liter container holds 100 g of N2 . The pressure p of
40
this gas is measured at various temperatures T (see the ac-
companying table).
30
(a) Find the least squares line for this collection of data
20 points. If your calculating utility can produce the cor-
relation coefficient, then find it.
10 (b) Use the model obtained in part (a) to predict the pressure
x of the gas at a temperature of −50 ◦ C.
–4 –2 2 4 6 8 10 (c) Use the model obtained in part (a) to predict a temper-
(b) ature at which the pressure of the gas will be zero.
y
8
temperature T (°C) pressure p (atm)
6
4 0 3.99
25 4.34
2
x 50 4.70
–5 5 10 15 75 5.08
–2 100 5.45
–4 Table Ex-5
(c)
y 6. A 40-liter container holds 20 g of H2 . The pressure p of this
15 gas is measured at various temperatures T (see the accom-
12.5 panying table).
10 (a) Find the least squares line for this collection of data
7.5
points. If your calculating utility can produce the cor-
relation coefficient, then find it.
5
(b) Use the model obtained in part (a) to predict a temper-
2.5
x ature at which the pressure of the gas will be zero.
–4 –2 2 4 6 8 10
(c) At approximately what temperature of the gas will a
10 ◦ C increase in temperature result in a 5% increase in
(d )
pressure?
Figure Ex-2
3. Table 1.1.1 provides data for the top qualifying speeds at temperature T (°C) pressure p (atm)
the Indianapolis 500 from 1980 to 1999. Find the least
0 5.55
squares line for these data. What is the correlation coef-
30 6.13
ficient? Sketch the least squares line on a scatter plot of the
60 6.75
data points.
90 7.35
4. A 25-liter container holds 150 g of O2 . The pressure p of 120 7.98
the gas is measured at various temperatures T (see the ac- Table Ex-6
companying table).
(a) Determine the least squares line for the data given in
the table. 7. The resistivity of a metal is a measure of the extent to which
(b) Use the model obtained in part (a) to estimate the pres- a wire made from the metal will resist the flow of electri-
sure of the gas at a temperature of −50 ◦ C. cal current. (The actual resistance of the wire will depend
on both the resistivity of the metal and the dimensions of
temperature T (°C) pressure p (atm) the wire.) A common unit for resistivity is the ohm-meter
(1·m). Experiments show that lowering the temperature of
0 4.18 a metal also lowers its resistivity. The accompanying table
50 4.96 gives the resistivity of copper at various temperatures.
100 5.74 (a) Find the least squares line for this collection of data
150 6.49
points.
200 7.26
(b) Using the model obtained in part (a), at what tempera-
Table Ex-4 ture will copper have a resistivity of zero?
January 12, 2001 11:10 g65-ch1 Sheet number 80 Page number 86 cyan magenta yellow black
86 Functions
H (the constant of proportionality) is called Hubble’s con- 15. A student is studying the equations of motion for an ob-
stant. When applying this formula it is usual to express v ject moving along a number line with constant acceleration.
in kilometers per second (km/s) and d in millions of light- The accompanying table gives the position in meters of the
years (Mly), in which case H has units of km/s/Mly. The object at various times.
accompanying figure shows an original plot and trend line of (a) Use quadratic regression to model the position of the
the velocity-distance relationship obtained by Hubble and a object as a function of time.
collaborator Milton L. Humason (1891–1972). (b) Based on the model obtained in part (a), what will be
(a) Use the trend line in the figure to estimate Hubble’s the position of the object after 2 s?
constant.
(b) An estimate of the age of the universe can be obtained
by assuming that the galaxies move with constant ve- time (s) position (m)
locity v, in which case v and d are related by d = vt. 0.2537 0.045
Assuming that the Universe began with a “big bang” 0.4064 0.09
that initiated its expansion, show that the Universe is 0.5981 0.165
roughly 1.5×1010 years old. [Use the conversion 1 Mly 0.75 0.24
≈ 9.048 × 1018 and take H = 20 km/s/Mly, which is 0.8781 0.315
in keeping with current estimates that place H between 1.032 0.42
15 and 27 km/s/Mly. (Note that the current estimates 1.1846 0.54
are significantly less than that resulting from Hubble’s 1.3208 0.66
data.)] Table Ex-15
(c) In a more realistic model of the Universe, the velocity v
would decrease with time. What effect would that have 16. Table 1.1.3 gives data for the U.S. population at 10-year
on your estimate in part (b)? intervals from 1790 to 1850. Use quadratic regression to
model the U.S. population as a function of time since 1790.
20,000 What does your model predict as the population of the
United States in the year 2000? How accurate is this predic-
Radial velocity (km/s)
15,000 tion?
17. The accompanying table gives the minutes of daylight pre-
10,000
dicted for Davidson, North Carolina, in 10-day increments
during the year 2000. Find a function that models the data
5,000
in this table, and graph your function on a scatter plot of the
data.
0 20 40 60 80 100 120
Distance (millions of light-years)
day daylight (min) day daylight (min)
Figure Ex-13
10 716 190 986
20 727 200 975
14. A professor wishes to use midterm grades as a predictor of 30 744 210 961
final grades in a small seminar that he teaches once a year. 40 762 220 944
The midterm grades and final grades for last year’s seminar 50 783 230 926
are listed in the accompanying table. 60 804 240 905
(a) Find the linear regression model that expresses the final 70 826 250 883
grade in terms of the midterm grade. 80 848 260 861
(b) Suppose that a student in this year’s seminar earned a 90 872 270 839
100 894 280 817
midterm grade of 88. Use the model obtained in part (a)
110 915 290 795
to predict the student’s final grade in the seminar.
120 935 300 774
130 954 310 755
midterm grade final grade 140 969 320 738
150 982 330 723
78 78 160 990 340 712
94 91 170 993 350 706
78 76 180 992 360 706
84 82
95 92 Table Ex-17
96 93
77 75 18. The accompanying table gives the fraction of the Moon that
is illuminated at midnight and visible from Earth (eastern
Table Ex-14
January 12, 2001 11:10 g65-ch1 Sheet number 82 Page number 88 cyan magenta yellow black
88 Functions
2 1 32 1
20. (a) The accompanying table below provides data on five
4 0.94 34 0.93
6 0.81 36 0.79 moons of the planet Saturn. In this table r is the or-
8 0.63 38 0.62 bital radius (the average distance between the moon
10 0.44 40 0.43 and Saturn) and t is the time in days required for the
12 0.26 42 0.25 moon to complete one orbit around Saturn. For each
14 0.12 44 0.10 data pair calculate tr −3/2 , and use your results to find a
16 0.02 46 0.01 reasonable model for r as a function of t.
18 0 48 0.01 (b) Use the model obtained in part (a) to estimate the orbital
20 0.07 50 0.11 radius of the moon Enceladus, given that its orbit time
22 0.22 52 0.29 is t ≈ 1.370 days.
24 0.43 54 0.51 (c) Use the model obtained in part (a) to estimate the orbit
26 0.66 56 0.73 time of the moon Tethys, given that its orbital radius is
28 0.85 58 0.90 r ≈ 2.9467 × 105 km.
30 0.97 60 0.99
Table Ex-18 moon radius orbit time
(100,000 km) (days)
19. The accompanying table provides data about the relation- 1980S28 1.3767 0.602
ship between distance d traveled in meters and elapsed time 1980S27 1.3935 0.613
1980S26 1.4170 0.629
t in seconds for an object dropped near the Earth’s surface.
1980S3 1.5142 0.694
Plot time versus distance and make a guess at a “square-root
1980S1 1.5147 0.695
function” that provides a reasonable model for t in terms of
d. Use a graphing utility to confirm the reasonableness of Table Ex-20
your guess.
Thus far, our study of graphs has focused on graphs of functions. However, because
such graphs must pass the vertical line test, this limitation precludes curves with self-
intersections or even such basic curves as circles. In this section we will study an
alternative method for describing curves algebraically that is not subject to the severe
restriction of the vertical line test.
This material is placed here to provide an early parametric option. However, it can be
deferred until Chapter 11, if preferred.
••••••••••••••••••••••••••••••••••••••
Suppose that a particle moves along a curve C in the xy-plane in such a way that its x- and
PARAMETRIC EQUATIONS y-coordinates, as functions of time, are
x = f (t), y = g(t)
We call these the parametric equations of motion for the particle and refer to C as the
trajectory of the particle or the graph of the equations (Figure 1.8.1). The variable t is
called the parameter for the equations.
Example 1 Sketch the trajectory over the time interval 0 ≤ t ≤ 10 of the particle whose
parametric equations of motion are
x = t − 3 sin t, y = 4 − 3 cos t (1)
January 12, 2001 11:10 g65-ch1 Sheet number 83 Page number 89 cyan magenta yellow black
y
Solution. One way to sketch the trajectory is to choose a representative succession of
C times, plot the (x, y) coordinates of points on the trajectory at those times, and connect the
points with a smooth curve. The trajectory in Figure 1.8.2 was obtained in this way from
Table 1.8.1 in which the approximate coordinates of the particle are given at time increments
(x, y) of 1 unit. Observe that there is no t-axis in the picture; the values of t appear only as labels
on the plotted points, and even these are usually omitted unless it is important to emphasize
x the location of the particle at specific times.
Figure 1.8.1 t x y
y 0 0.0 1.0
1 –1.5 2.4
8 t=3 2 –0.7 5.2
t=9
3 2.6 7.0
6 t=4 t = 10
t=2 4 6.3 6.0
t=8 5 7.9 3.1
4
t=5 6 6.8 1.1
t=1 2 7 5.0 1.7
t=7
t=0 t=6 x 8 5.0 4.4
-2 2 4 6 8 10 12 9 7.8 6.7
10 11.6 6.5
Figure 1.8.2
Although parametric equations commonly arise in problems of motion with time as the
parameter, they arise in other contexts as well. Thus, unless the problem dictates that the
parameter t in the equations
x = f (t), y = g(t)
represents time, it should be viewed simply as an independent variable that varies over some
interval of real numbers. (In fact, there is no need to use the letter t for the parameter; any
letter not reserved for another purpose can be used.) If no restrictions on the parameter are
stated explicitly or implied by the equations, then it is understood that it varies from −⬁ to
+⬁. To indicate that a parameter t is restricted to an interval [a, b], we will write
(x, y)
x = f (t), y = g(t) (a ≤ t ≤ b)
1
y
t (1, 0)
Example 2 Find the graph of the parametric equations
x x = cos t, y = sin t (0 ≤ t ≤ 2π) (2)
Solution. One way to find the graph is to eliminate the parameter t by noting that
x 2 + y 2 = sin2 t + cos2 t = 1
Thus, the graph is contained in the unit circle x 2 +y 2 = 1. Geometrically, t can be interpreted
x = cos t, y = sin t as the angle swept out by the radial line from the origin to the point (x, y) = (cos t, sin t)
(0 ≤ t ≤ 2p) on the unit circle (Figure 1.8.3). As t increases from 0 to 2π, the point traces the circle
Figure 1.8.3 counterclockwise, starting at (1, 0) when t = 0 and completing one full revolution when
t = 2π. One can obtain different portions of the circle by varying the interval over which
the parameter varies. For example,
x = cos t, y = sin t (0 ≤ t ≤ π) (3)
represents just the upper semicircle in Figure 1.8.3.
January 12, 2001 11:10 g65-ch1 Sheet number 84 Page number 90 cyan magenta yellow black
90 Functions
••••••••••••••••••••••••••••••••••••••
The direction in which the graph of a pair of parametric equations is traced as the parameter
ORIENTATION increases is called the direction of increasing parameter or sometimes the orientation
imposed on the curve by the equations. Thus, we make a distinction between a curve,
which is a set of points, and a parametric curve, which is a curve with an orientation
imposed on it by a set of parametric equations. For example, we saw in Example 2 that the
circle represented parametrically by (2) is traced counterclockwise as t increases and hence
has counterclockwise orientation. As shown in Figures 1.8.2 and 1.8.3, the orientation of a
parametric curve can be indicated by arrowheads.
x (1, 0) To obtain parametric equations for the unit circle with clockwise orientation, we can
–t replace t by −t in (2), and use the identities cos(−t) = cos t and sin(−t) = − sin t. This
y
1 yields
(x, y) x = cos t, y = − sin t (0 ≤ t ≤ 2π)
Here, the circle is traced clockwise by a point that starts at (1, 0) when t = 0 and completes
one full revolution when t = 2π (Figure 1.8.4).
x = cos (–t), y = sin (– t)
(0 ≤ t ≤ 2p)
• FOR THE READER.
• When parametric equations are graphed using a calculator, the orien-
•
•
•
Figure 1.8.4 •
•
• tation can often be determined by watching the direction in which the graph is traced on the
•
•
•
•
•
•
•
screen. However, many computers graph so fast that it is often hard to discern the orien-
•
y
•
•
•
•
tation. See if you can use your graphing utility to confirm that (3) has a counterclockwise
•
•
•
• orientation.
8
Example 3 Graph the parametric curve
x = 2t − 3, y = 6t − 7
by eliminating the parameter, and indicate the orientation on the graph.
Solution. To eliminate the parameter we will solve the first equation for t as a function
of x, and then substitute this expression for t into the second equation:
x
t = 12 (x + 3)
y = 6 12 (x + 3) − 7
y = 3x + 2
x = 2t – 3, y = 6t – 7
Thus, the graph is a line of slope 3 and y-intercept 2. To find the orientation we must look
Figure 1.8.5 to the original equations; the direction of increasing t can be deduced by observing that
x increases as t increases or by observing that y increases as t increases. Either piece of
information tells us that the line is traced left to right as shown in Figure 1.8.5.
y
(–1, 1) (1, 1) • REMARK.
• Not all parametric equations produce curves with definite orientations; if the
•
•
•
•
•
•
•
equations are badly behaved, then the point tracing the curve may leap around sporadically
•
•
•
•
•
or move back and forth, failing to determine a definite direction. For example, if
•
•
x = sin t, y = sin2 t
•
•
•
•
•
•
x •
then the point (x, y) moves along the parabola y = x 2 . However, the value of x varies
•
•
•
•
•
periodically between −1 and 1, so the point (x, y) moves periodically back and forth along
•
•
•
•
•
•
Figure 1.8.6 •
•
• the parabola between the points (−1, 1) and (1, 1) (as shown in Figure 1.8.6). Later in the
•
•
•
•
•
•
text we will discuss restrictions that eliminate such erratic behavior, but for now we will
•
•
•
• just avoid such complications.
••••••••••••••••••••••••••••••••••••••
An equation y = f(x) can be expressed in parametric form by introducing the parameter
EXPRESSING ORDINARY t = x; this yields the parametric equations x = t, y = f (t). For example, the portion of
FUNCTIONS PARAMETRICALLY
the curve y = cos x over the interval [−2π, 2π] can be expressed parametrically as
x = t, y = cos t (−2π ≤ t ≤ 2π)
(Figure 1.8.7).
January 12, 2001 11:10 g65-ch1 Sheet number 85 Page number 91 cyan magenta yellow black
y
t = –2p 1 t=0 t = 2p
x
-7 7
-1
t = –p t=p
Figure 1.8.7
••••••••••••••••••••••••••••••••••••••
Many graphing utilities allow you to graph equations of the form y = f(x) but not equations
GENERATING PARAMETRIC of the form x = g(y). Sometimes you will be able to rewrite x = g(y) in the form y = f(x);
CURVES WITH GRAPHING UTILITIES
however, if this is inconvenient or impossible, then you can graph x = g(y) by introducing
a parameter t = y and expressing the equation in the parametric form x = g(t), y = t.
(You may have to experiment with various intervals for t to produce a complete graph.)
x x x
x = 31 cos t – 7 cos (31/ 7)t x = 17 cos t + 7 cos (17/ 7)t x = cos t + (1/2) cos 7t + (1/3)sin 17t
y = 31 sin t – 7 sin (31/ 7)t y = 17 sin t – 7 sin (17/ 7)t y = sin t + (1/2) sin 7t + (1/3)cos 17t
(0 ≤ t ≤ 14p) (0 ≤ t ≤ 14p) (0 ≤ t ≤ 2p)
Figure 1.8.9
••••••••••••••••••••••••••••••••••••••
If a parametric curve C is given by the equations x = f (t), y = g(t), then adding a constant
TRANSLATION to f (t) translates the curve C in the x-direction, and adding a constant to g(t) translates
it in the y-direction. Thus, a circle of radius r, centered at (x0 , y0 ) can be represented
parametrically as
x = x0 + r cos t, y = y0 + r sin t (0 ≤ t ≤ 2π) (4)
(Figure 1.8.10). If desired, we can eliminate the parameter from these equations by noting
January 12, 2001 11:10 g65-ch1 Sheet number 86 Page number 92 cyan magenta yellow black
92 Functions
(x0 , y0)
x = x0 + r cos t
y = y0 + r sin t
(0 ≤ t ≤ 2p)
Figure 1.8.10
that
(x − x0 )2 + (y − y0 )2 = (r cos t)2 + (r sin t)2 = r 2
Thus, we have obtained the familiar equation in rectangular coordinates for a circle of radius
r, centered at (x0 , y0 ):
(x − x0 )2 + (y − y0 )2 = r 2 (5)
••••••••••••••••••••••••••••••••••••••
If a parametric curve C is given by the equations x = f (t), y = g(t), then multiplying
SCALING f (t) by a constant stretches or compresses C in the x-direction, and multiplying g(t) by a
constant stretches or compresses C in the y-direction. For example, we would expect the
parametric equations
x = 3 cos t, y = 2 sin t (0 ≤ t ≤ 2π)
to represent an ellipse, centered at the origin, since the graph of these equations results from
stretching the unit circle
y
x = cos t, y = sin t (0 ≤ t ≤ 2π)
b
by a factor of 3 in the x-direction and a factor of 2 in the y-direction. In general, if a and b
x
are positive constants, then the parametric equations
–a a x = a cos t, y = b sin t (0 ≤ t ≤ 2π) (6)
represent an ellipse, centered at the origin, and extending between −a and a on the x-axis
–b
and between −b and b on the y-axis (Figure 1.8.11). The numbers a and b are called the
semiaxes of the ellipse. If desired, we can eliminate the parameter t in (6) and rewrite the
x = a cos t, y = b sin t equations in rectangular coordinates as
(0 ≤ t ≤ 2p)
x2 y2
+ =1 (7)
Figure 1.8.11 a2 b2
ellipse that is centered at the origin and that extends between −4 and 4 in the x-direction
•
•
•
•
•
and between −3 and 3 in the y-direction. Generate an ellipse with the same dimensions,
•
•
•
•
•
•
•
•
• but translated so that its center is at (2, 3).
••••••••••••••••••••••••••••••••••••••
In the mid-1850s the French physicist Jules Antoine Lissajous (1822–1880) became inter-
LISSAJOUS CURVES ested in parametric equations of the form
x = sin at, y = sin bt (8)
in the course of studying vibrations that combine two perpendicular sinusoidal motions.
January 12, 2001 11:10 g65-ch1 Sheet number 87 Page number 93 cyan magenta yellow black
The first equation in (8) describes a sinusoidal oscillation in the x-direction with frequency
a /2π, and the second describes a sinusoidal oscillation in the y-direction with frequency
b/2π. If a /b is a rational number, then the combined effect of the oscillations is a periodic
motion along a path called a Lissajous curve. Figure 1.8.12 shows some typical Lissajous
curves.
y y y y
1 1 1
1
x x x x
-1 1 -1 1 -1 1 -1 1
-1 -1 -1
-1
a = 1, b = 2 a = 2, b = 3 a = 3, b = 4 a = 4, b = 5
Figure 1.8.12
••••••••••••••••••••••••••••••••••••••
If a wheel rolls in a straight line along a flat road, then a point on the rim of the wheel will
CYCLOIDS trace a curve called a cycloid (Figure 1.8.13). This curve has a fascinating history, which
we will discuss shortly; but first we will show how to obtain parametric equations for it. For
this purpose, let us assume that the wheel has radius a and rolls along the positive x-axis of
a rectangular coordinate system. Let P (x, y) be the point on the rim that traces the cycloid,
and assume that P is initially at the origin. We will take as our parameter the angle θ that
is swept out by the radial line to P as the wheel rolls (Figure 1.8.13). It is standard here to
regard θ to be positive, even though it is generated by a clockwise rotation.
The motion of P is a combination of the movement of the wheel’s center parallel to the
x-axis and the rotation of P around the center. As the radial line sweeps out an angle θ , the
point P traverses an arc of length aθ , and the wheel moves a distance aθ along the x-axis
(why?). Thus, as suggested by Figure 1.8.14, the center moves to the point (aθ, a), and the
coordinates of P (x, y) are
x = aθ − a sin θ, y = a − a cos θ (9)
These are the equations of the cycloid in terms of the parameter θ .
y
P(x, y)
a u a cos u
a sin u a
a y = a – a cos u
x
x
P ca oa åa
x = au – a sin u
A cycloid
au
Figure 1.8.13 Figure 1.8.14
94 Functions
••••••••••••••••••••••••••••••••••••••
The cycloid is of interest because it provides the solution to two famous mathematical
THE ROLE OF THE CYCLOID IN problems—the brachistochrone problem (from Greek words meaning “shortest time”) and
MATHEMATICS HISTORY
the tautochrone problem (from Greek words meaning “equal time”). The brachistochrone
problem is to determine the shape of a wire along which a bead might slide from a point P
to another point Q, not directly below, in the shortest time. The tautochrone problem is to
P find the shape of a wire from P to Q such that two beads started at any points on the wire
between P and Q reach Q in the same amount of time (Figure 1.8.15). The solution to both
problems turns out to be an inverted cycloid.
∗
In June of 1696, Johann Bernoulli posed the brachistochrone problem in the form of a
Q
challenge to other mathematicians. At first, one might conjecture that the wire should form
Figure 1.8.15
a straight line, since that shape results in the shortest distance from P to Q. However, the
inverted cycloid allows the bead to fall more rapidly at first, building up sufficient initial
speed to reach Q in the shortest time, even though it travels a longer distance. The problem
was solved by Newton and Leibniz as well as by Johann Bernoulli and his older brother
∗
BERNOULLI. An amazing Swiss family that included several generations of outstanding mathematicians and
scientists. Nikolaus Bernoulli (1623–1708), a druggist, fled from Antwerp to escape religious persecution and
ultimately settled in Basel, Switzerland. There he had three sons, Jakob I (also called Jacques or James), Nikolaus,
and Johann I (also called Jean or John). The Roman numerals are used to distinguish family members with identical
names (see the family tree below). Following Newton and Leibniz, the Bernoulli brothers, Jakob I and Johann I,
are considered by some to be the two most important founders of calculus. Jakob I was self-taught in mathematics.
His father wanted him to study for the ministry, but he turned to mathematics and in 1686 became a professor
at the University of Basel. When he started working in mathematics, he knew nothing of Newton’s and Leibniz’
work. He eventually became familiar with Newton’s results, but because so little of Leibniz’ work was published,
Jakob duplicated many of Leibniz’ results.
Jakob’s younger brother Johann I was urged to enter into business by his father. Instead, he turned to medicine
and studied mathematics under the guidance of his older brother. He eventually became a mathematics professor
at Groningen in Holland, and then, when Jakob died in 1705, Johann succeeded him as mathematics professor at
Basel. Throughout their lives, Jakob I and Johann I had a mutual passion for criticizing each other’s work, which
frequently erupted into ugly confrontations. Leibniz tried to mediate the disputes, but Jakob, who resented Leibniz’
superior intellect, accused him of siding with Johann, and thus Leibniz became entangled in the arguments. The
brothers often worked on common problems that they posed as challenges to one another. Johann, interested in
gaining fame, often used unscrupulous means to make himself appear the originator of his brother’s results; Jakob
occasionally retaliated. Thus, it is often difficult to determine who deserves credit for many results. However, both
men made major contributions to the development of calculus. In addition to his work on calculus, Jakob helped
establish fundamental principles in probability, including the Law of Large Numbers, which is a cornerstone of
modern probability theory.
Among the other members of the Bernoulli family, Daniel, son of Johann I, is the most famous. He was a
professor of mathematics at St. Petersburg Academy in Russia and subsequently a professor of anatomy and then
physics at Basel. He did work in calculus and probability, but is best known for his work in physics. A basic law
of fluid flow, called Bernoulli’s principle, is named in his honor. He won the annual prize of the French Academy
10 times for work on vibrating strings, tides of the sea, and kinetic theory of gases.
Johann II succeeded his father as professor of mathematics at Basel. His research was on the theory of heat
and sound. Nikolaus I was a mathematician and law scholar who worked on probability and series. On the
recommendation of Leibniz, he was appointed professor of mathematics at Padua and then went to Basel as a
professor of logic and then law. Nikolaus II was professor of jurisprudence in Switzerland and then professor of
mathematics at St. Petersburg Academy. Johann III was a professor of mathematics and astronomy in Berlin and
Jakob II succeeded his uncle Daniel as professor of mathematics at St. Petersburg Academy in Russia. Truly an
incredible family!
Nikolaus Bernoulli
(1623–1708)
Jakob; it was formulated and solved incorrectly years earlier by Galileo, who thought the
answer was a circular arc.
metric equations of motion are 17. The portion of the parabola x = y 2 joining (1, −1) and
(1, 1), oriented down to up.
x = cos(πt), y = sin(πt) 18. The circle of radius 4, centered at (1, −3), oriented coun-
(b) Indicate the direction of motion on your sketch. terclockwise.
(c) Make a table of x- and y-coordinates of the particle at 19. In each part, match the parametric equation with one of the
times t = 0, 0.25, 0.5, 0.75, 1. curves labeled
√ (I)–(VI), and explain your reasoning.
(a) x = t, y = sin 3t (b) x = 2 cos t, y = 3 sin t
(d) Mark the position of the particle on the curve at the times
3t 3t 2
in part (c), and label those positions with the values of t. (c) x = t cos t, y = t sin t (d) x = , y=
1+t 3 1 + t3
3 2
In Exercises 3–12, sketch the curve by eliminating the pa- t 2t
(e) x = , y= (f ) x = 2 cos t, y = sin 2t
rameter, and indicate the direction of increasing t. 1 + t2 1 + t2
3. x = 3t − 4, y = 6t + 2 y y
y
4. x = t − 3, y = 3t − 7 (0 ≤ t ≤ 3) x x
5. x = 2 cos t, y = 5 sin t (0 ≤ t ≤ 2π)
√ x
6. x = t, y = 2t + 4
7. x = 3 + 2 cos t, y = 2 + 4 sin t(0 ≤ t ≤ 2π) I II III
96 Functions
20. Use a graphing utility to generate the curves in Exercise 19, and that R(x, y) is the point on the line segment corre-
and in each case identify the orientation. sponding to a specified value of t (see the accompanying
21. (a) Use a graphing utility to generate the trajectory of a par- figure). Show that t = r /q, where r is the distance from
ticle whose equations of motion over the time interval P to R and q is the distance from P to Q.
0 ≤ t ≤ 5 are (b) What value of t produces the midpoint between points
x = 6t − 12 t 3 , y = 1 + 12 t 2 P and Q?
(b) Make a table of x- and y-coordinates of the particle at (c) What value of t produces the point that is three-fourths
times t = 0, 1, 2, 3, 4, 5. of the way from P to Q?
(c) At what times is the particle on the y-axis?
(d) During what time interval is y < 5?
t=1
(e) At what time is the x-coordinate of the particle maxi-
mum? Q(x 1, y1)
t
22. (a) Use a graphing utility to generate the trajectory of a
paper airplane whose equations of motion for t ≥ 0 are R(x, y)
t=0
x = t − 2 sin t, y = 3 − 2 cos t
P(x 0, y0)
(b) Assuming that the plane flies in a room in which the
Figure Ex-27
floor is at y = 0, explain why the plane will not crash
into the floor. [For simplicity, ignore the physical size
of the plane by treating it as a particle.] 28. Find parametric equations for the line segment joining
(c) How high must the ceiling be to ensure that the plane P (2, −1) and Q(3, 1), and use the result in Exercise 27
does not touch or crash into it? to find
(a) the midpoint between P and Q
In Exercises 23 and 24, graph the equation using a graphing
utility. (b) the point that is one-fourth of the way from P to Q
(c) the point that is three-fourths of the way from P to Q.
23. (a) x = y + 2y + 1
2
29. Explain why the parametric curve
(b) x = sin y, −2π ≤ y ≤ 2π
x = t 2, y = t4 (−1 ≤ t ≤ 1)
24. (a) x = y + 2y 3 − y 5
(b) x = tan y, −π/2 < y < π/2 does not have a definite orientation.
25. (a) By eliminating the parameter, show that the equations 30. (a) In parts (a) and (b) of Exercise 25 we obtained paramet-
x = x0 + (x1 − x0 )t, y = y0 + (y1 − y0 )t ric equations for a line segment in which the parameter
varied from t = 0 to t = 1. Sometimes it is desir-
represent the line passing through the points (x0 , y0 ) able to have parametric equations for a line segment in
and (x1 , y1 ). which the parameter varies over some other interval, say
(b) Show that if 0 ≤ t ≤ 1, then the equations in part (a)
t0 ≤ t ≤ t1 . Use the ideas in Exercise 25 to show that
represent the line segment joining (x0 , y0 ) and (x1 , y1 ),
the line segment joining the points (x0 , y0 ) and (x1 , y1 )
oriented in the direction from (x0 , y0 ) to (x1 , y1 ).
can be represented parametrically as
(c) Use the result in part (b) to find parametric equations for
the line segment joining the points (1, −2) and (2, 4), t − t0
x = x0 + (x1 − x0 ) ,
oriented in the direction from (1, −2) to (2, 4). t1 − t0
(d) Use the result in part (b) to find parametric equations for (t0 ≤ t ≤ t1 )
the line segment in part (c), but oriented in the direction t − t0
y = y0 + (y1 − y0 )
from (2, 4) to (1, −2). t1 − t0
26. Use the result in Exercise 25 to find
(b) Which way is the line segment oriented?
(a) parametric equations for the line segment joining the
points (−3, −4) and (−5, 1), oriented from (−3, −4) (c) Find parametric equations for the line segment traced
to (−5, 1) from (3, −1) to (1, 4) as t varies from 1 to 2, and check
(b) parametric equations for the line segment traced from your result with a graphing utility.
(0, b) to (a, 0), oriented from (0, b) to (a, 0).
31. (a) By eliminating the parameter, show that if a and c are
27. (a) Suppose that the line segment from the point P (x0 , y0 ) not both zero, then the graph of the parametric equations
to Q(x1 , y1 ) is represented parametrically by
x = x0 + (x1 − x0 )t, x = at + b, y = ct + d (t0 ≤ t ≤ t1 )
(0 ≤ t ≤ 1)
y = y0 + (y1 − y0 )t is a line segment.
January 12, 2001 11:10 g65-ch1 Sheet number 91 Page number 97 cyan magenta yellow black
x = 2t − 1, y =t +1 (1 ≤ t ≤ 2)
y
Buffing
pad
(– 12 , – 12 ) ( 1
2,
– 12 )
Figure Ex-34 x
Robot
arms
35. (a) Find parametric equations for the ellipse that is centered
at the origin and has intercepts (4, 0), (−4, 0), (0, 3),
and (0, −3).
(b) Find parametric equations for the ellipse that results by Figure Ex-38
translating the ellipse in part (a) so that its center is at
(−1, 2). 39. Describe the family of curves described by the parametric
(c) Confirm your results in parts (a) and (b) using a graph- equations
ing utility. x = a cos t + h, y = b sin t + k (0 ≤ t ≤ 2π)
36. We will show later in the text that if a projectile is fired from if
ground level with an initial speed of v0 meters per second (a) h and k are fixed but a and b can vary
at an angle α with the horizontal, and if air resistance is (b) a and b are fixed but h and k can vary
neglected, then its position after t seconds, relative to the
(c) a = 1 and b = 1, but h and k vary so that h = k + 1.
coordinate system in the accompanying figure is
40. A hypocycloid is a curve traced by a point P on the cir-
x = (v0 cos α)t, y = (v0 sin α)t − 12 gt 2
cumference of a circle that rolls inside a larger fixed circle.
where g ≈ 9.8 m/s2 . Suppose that the fixed circle has radius a, the rolling circle
(a) By eliminating the parameter, show that the trajectory has radius b, and the fixed circle is centered at the origin.
is a parabola. Let φ be the angle shown in the following figure, and as-
(b) Sketch the trajectory if α = 30 ◦ and v0 = 1000 m/s. sume that the point P is at (a, 0) when φ = 0. Show that the
January 12, 2001 11:10 g65-ch1 Sheet number 92 Page number 98 cyan magenta yellow black
98 Functions
hypocycloid generated is given by the parametric equations 41. If b = 14 a in Exercise 40, then the resulting curve is called
a four-cusped hypocycloid.
a−b
x = (a − b) cos φ + b cos φ (a) Sketch this curve.
b
(b) Show that the curve is given by the parametric equations
a−b x = a cos3 φ, y = a sin3 φ.
y = (a − b) sin φ − b sin φ
b
(c) Show that the curve is given by the equation
y
x 2 /3 + y 2 /3 = a 2 /3
in rectangular coordinates.
42. (a) Use a graphing utility to study how the curves in the
f x
family
(a, 0) x = 2a cos2 t, y = 2a cos t sin t (−2π < t < 2π)
SUPPLEMENTARY EXERCISES
1. Referring to the cigarette consumption graph in Figure 4. A student begins driving toward school but 5 minutes into
1.1.2b, during what 5-year period was the annual cigarette the trip remembers that he forgot his homework. He drives
consumption per adult increasing most rapidly on average? home hurriedly, retrieves his notes, and then drives at great
Explain your reasoning. speed toward school, hitting a tree 5 minutes after leaving
2. Use the graphs of the functions f and g in the accompanying home. Sketch a rough graph that reasonably describes the
figure to solve the following problems. student’s distance from home as a function of the elapsed
(a) Find the values of f (−1) and g(3). time.
(b) For what values of x is f(x) = g(x)? 5. A rectangular storage container with an open top and a
(c) For what values of x is f(x) < 2? square base has a volume of 8 cubic meters. Material for the
(d) What are the domain and range of f ? base costs $5 per square meter, and material for the sides $2
(e) What are the domain and range of g? per square meter. Express the total cost of the materials as
(f ) Find the zeros of f and g. a function of the length of a side of the base.
6. You want to paint the top of a circular table. Find a formula
y
that expresses the amount of paint required as a function
5 of the radius, and discuss all of the assumptions you have
made in finding the formula.
f 7. Sketch the graph of the function
x
−1, x ≤ −5
-5 5 √
g f(x) = 25 − x 2 , −5 < x < 5
x − 5, x≥5
Supplementary Exercises 99
(b) Find an inequality that specifies the domain of the 19. (a) A surveyor measures the angle of elevation α of a tower
function in part (a). from a point A due south of the tower and also measures
(c) Estimate the dimensions of the box of largest volume. the angle of elevation β from a point B that is d feet
due east of the point A (see the accompanying figure).
10 ft Show that the height h of the tower in feet is given by
d tan α tan β
h=
6 ft tan2 α − tan2 β
g(x) 3 2 1 –3 –1 –4 4 –2 0 (c) Based on this model, how many days during the year
( f ° g)(x)
would you expect the temperature to be below 0 ◦ F?
21. The accompanying figure shows the graph of the equa-
(g ° f )(x)
tions y = 1 + 2 sin x and y = 2 sin(x /2) + 2 cos(x /2) for
−2π ≤ x ≤ 2π. Without the aid of a calculator, label each
16. (a) Write an equation for the graph that is obtained by re- curve by its equation, and find the coordinates of the points
flecting the graph of y = |x − 1| about the y-axis, then A, B, C, and D. Explain your reasoning.
stretching that graph vertically by a factor of 2, then
y
translating that graph down 3 units, and then reflecting
B C
that graph about the x-axis.
(b) Sketch the original graph and the final graph.
17. In each part, classify the function as even, odd, or nei- x
ther. o
O
(a) x 2 sin x (b) sin2 x D
A
(c) x + x 2
(d) sin x tan x
18. (a) Find exact values for all x-intercepts of
Figure Ex-21
y = cos x − sin 2x
in the interval −2π ≤ x ≤ 2π. 22. The accompanying figure shows a model for the tide varia-
(b) Find the coordinates of all intersections of the graphs tion in an inlet to San Francisco Bay during a 24-hour period.
of y = cos x and y = sin 2x if −2π ≤ x ≤ 2π, and use Find an equation of the form y = y0 + y1 sin(at + b) for
a graphing utility to check your answer. the model, assuming that t = 0 corresponds to midnight.
January 12, 2001 11:10 g65-ch1 Sheet number 94 Page number 100 cyan magenta yellow black
100 Functions
102 Functions
(d) Using a graphing utility with a graphing window con- hour water level (m) hour water level (m)
taining (0, f (0)), graph y = f(x) and the polynomial
you found in part (c). How do the graphs compare as 0 0.526 36 0.534
you zoom closer to the point (0, f (0))? 2 0.157 38 0.192
[Note: The best quadratic approximation to y = cos x near 4 0.161 40 0.141
x = 0 is given by y ≈ −0.5x 2 + 1.] 6 0.486 42 0.426
8 0.779 44 0.849
43. The accompanying table gives the water level (in meters 10 0.740 46 1.032
above the mean low-water mark) at a Cape Hatteras, North 12 0.412 48 0.765
Carolina, fishing pier, recorded in 2-hour increments start- 14 0.141 50 0.281
ing from midnight, July 1, 1999. Why should we expect 16 0.260 52 0.042
that a trigonometric function should fit these data? Find a 18 0.633 54 0.157
function that models the data, and graph your function on 20 1.015 56 0.587
a scatter plot. 22 1.021 58 0.777
24 0.670 60 0.620
26 0.231 62 0.241
28 0.128 64 0.045
30 0.345 66 0.195
32 0.697 68 0.613
34 0.821 70 0.945
Table Ex-43
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Iterative Processes
Recall that in the notation y = f(x), the variable x is called an input of the function f , and
the variable y is called the corresponding output. Suppose that we start with some input, say
x = c, and each time we compute an output we feed it back into f as an input. This generates the
following sequence of numbers:
f(c), f(f(c)), f(f(f(c))), f(f(f(f(c)))), . . .
This is called an iterated function sequence for f (from the Latin word iteratus, meaning “to re-
peat”). The number c is called the seed value for the sequence, the terms in the sequence are called
iterates, and each time f is applied we say that we have performed an iteration. Iterated function
sequences arise in a wide variety of physical processes that are collectively called dynamical
systems.
January 12, 2001 11:10 g65-ch1 Sheet number 97 Page number 103 cyan magenta yellow black
•••••••••••
Exercise 1 Let f(x) = x 2 .
(a) Calculate the first 10 iterates in the iterated function sequence for f , starting with seed values
of c = 0.5, 1, and 2. In each case make a conjecture about the long-term behavior of the
iterates, that is, the behavior of the iterates as more and more iterations are performed.
(b) Try your own seed values, and make a conjecture about the effect of a seed value on the
long-term behavior of the iterates.
Recursion Formulas
The proliferation of parentheses in an iterated function sequence can become confusing, so for
simplicity let us introduce the following notation for the successive iterates
y0 = c, y1 = f(c), y2 = f(f(c)), y3 = f(f(f(c))), y4 = f(f(f(f(c)))), . . .
••••••••••• √
Exercise 2 Use (1) to approximate 5 by generating successive iterates on a calculator√until
you encounter two successive iterates that are the same. Compare this approximation of 5 to
that produced directly by your calculator.
•••••••••••
Exercise 3
(a) Find iterates y1 up to y6 of the sequence that is generated by the recursion formula
y0 = 1, yn+1 = 12 yn
(b) By examining the terms generated in part (a), find a formula that expresses yn as a function of n.
•••••••••••
Exercise 4 Suppose that you deposit $1000 in a bank at 5% interest per year and allow it to
accumulate value without making withdrawals.
(a) If yn denotes the value of the account at the end of the nth year, how could you find the value
of yn+1 if you knew the value of yn ?
(b) Starting with y0 = 1000 (dollars), use the result in part (a) to calculate y1 , y2 , y3 , y4 , and y5 .
(c) Find a recursion formula for the sequence of yearly account values assuming that y0 = 1000.
(d) Find a formula that expresses yn as a function of n, and use that formula to calculate the value
of the account at the end of the 15th year.
January 12, 2001 11:10 g65-ch1 Sheet number 98 Page number 104 cyan magenta yellow black
104 Functions
Applications to Ecology
There are numerous models for predicting the growth and decline of populations (flowers, plants,
people, animals, etc.). One way to model populations is to give a recursion formula that describes
how the number of individuals in each generation relates to the number of individuals in the
January 12, 2001 11:10 g65-ch1 Sheet number 99 Page number 105 cyan magenta yellow black
preceding generation. One of the simplest such models, called the exponential model, assumes
that the number of individuals in each generation is a fixed percentage of the number of individ-
uals in the preceding generation. Thus, if there are c individuals initially and if the number of
individuals in any generation is r times the number of individuals in the preceding generation,
then the growth through successive generations is given by the recursion formula
y0 = c, yn+1 = ryn (n = 0, 1, 2, 3, . . .)
•••••••••••
Exercise 7 Suppose that a population with an exponential growth model has c individuals ini-
tially.
(a) Express the iterates y1 , y2 , y3 , and y4 in terms of c and r.
(b) Find a formula for yn+1 in terms of c and r.
(c) Describe the eventual fate of the population if r = 1, r < 1, and r > 1.
There is a more sophisticated model of population growth, called the logistic model, that
takes environmental constraints into account. In this model, it is assumed that there is some max-
imum population that can be supported by the environment, and the population is expressed as
a fraction of the maximum. Thus, in each generation the population is represented as a number
in the interval 0 ≤ yn ≤ 1. When yn is near 0 the population has lots of room to grow, but when
yn is near 1 the population is close to the maximum and the environmental factors tend to inhibit
further growth. Models of this type are given by recursion formulas of the form
y0 = c, yn+1 = kyn (1 − yn ) (2)
in which k is a positive constant that depends on the ecological conditions.
Figure 1 illustrates a graphical method for tracking the growth of a population described by
(2). That figure, which is called a cobweb diagram, shows graphs of the line y = x and the curve
y = kx(1 − x).
•••••••••••
Exercise 8 Explain why the values y1 , y2 , and y3 are the populations for the first three gener-
ations of the logistic growth model given by (2).
•••••••••••
Exercise 9 The cobweb diagram in Figure 2 tracks the growth of a population with a logistical
growth model given by the recursion formula
y0 = 0.1, yn+1 = 2.9yn (1 − yn )
(a) Find the populations y1 , y2 , . . . , y5 of the first five generations.
(b) What happens to the population over the long term?
y y
y=x
y2 0.8
0.7
y3 0.6
y1 0.5
0.4
0.3
0.2
0.1
x x
c 1 0.1 1
Figure 1 Figure 2
January 12, 2001 11:10 g65-ch1 Sheet number 100 Page number 106 cyan magenta yellow black
106 Functions