0% found this document useful (0 votes)
43 views33 pagesChapter 2 Textbook
Uploaded by
Aseel mujahedCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
43 views33 pagesChapter 2 Textbook
Uploaded by
Aseel mujahedCopyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
/ 33
Chapter 2
Functions
2.1 Domain, Range, One-to-One, Onto
Next in importance to the primitive notion of set is the idea of function. This is
a term which nowadays is introduced informally quite early in the mathematics
curriculum. It’s often defined as a rule which associates to each element of a set
A (often the set R of real numbers) an element of another set B (often R as well).
The function keys on a calculator are so named because with each number entered
into the calculator, another number is (usually) associated, namely, that number
which appears when the function key is pressed.! The essential characteristic of
a function is that the value which it associates with a given element is uniquely
determined by that element. If one enters 4 into a calculator and then presses
some function key, one is presented with exactly one number, not a choice of
several numbers: 4 | / |= 2, for instance, not +2.
Here is the formal definition of function.
2.1 DEFINITION. A function from a set A to a set B is a binary relation f from A
to B with the property that, for every a € A, there is exactly one 6 € B such
that (a,b) € f. (In some areas of mathematics, map is a commonly employed
synonym for function.)
In this definition, the idea that a function should associate with each element
a Aa unique element b € B is captured by viewing a function as a subset of
Ax B which, for each a € A, contains just one pair (a,b) having first coordinate
a. If A= {1,2,3}, B= {x,y} and f is
f={(,2), (2,9), @,2)},
then f is just the rule which associates x with 1, y with 2 and x with 3.
‘Here we assume that one does not enter a number like ~1 and then press the square root
key!
63 64 Chapter 2. Functions
There are two key points about the definition of function. First, every ain
A must be the first coordinate of an ordered pair in the function. Again with
A= {1,2,3} and B= {x,y}, the set
9={,2), (3,y)}
is not a function from A to B because g contains no ordered pair with first
coordinate 2. Second, each element of A must be the first coordinate of exactly
one ordered pair. With the same A and B,
h= {(1,2), (2,2), (3,y), (2,y)}
is not a function because 2 is the first coordinate of two ordered pairs.
Problem 1. Suppose A is the set of surnames of people listed in the Salt Lake
City telephone directory, Is it likely that
F ={(an) | ais on page n}
is a function from A to the set of natural numbers? Comment.
Solution. By definition of f, each element of A is the first coordinate of a pair
in f, 0 the key question here is whether or not each a € A determines a unique
7 €N. Is each surname on a unique page in the telephone directory? This is not
very likely since some surnames (Smith?) are undoubtedly listed on a number
of different pages. It is unlikely that f is a function, ao
When f isa function from A to B, the element 6 € B, which is uniquely
determined by the element a € A, is denoted f(a) and called the image of a.
Thus,
(1) [ aurer =
2.2 FUNCTION NOTATION. It is customary to write f: A+ B to mean that fis
a function from A to B and to write f: a++ 6 to mean that f(a) = b. (Note the
differences between the symbols —> between sets and +) between elements.)
f(a
Thus, the function f = {(1,2),(2,y),(3,2)} could also be described by
i,
wme
Pat
aes
although the description as ordered pairs seems simpler,
Sometimes (often in calculus), a function is sufficiently nice that it is possi-
ble to write down a precise formula showing how f(x) is determined by 2; for
example, f(x) = 2°, f(z) = 3¢—7, f(e) = log(x). When one talks about “the 2.1. Domain, Range, One-to-One, Onto 65
function f(x) = 22,” one is really talking about that function f: R — R which
associates with any x € R, its square x”; that is to say, x +> 2, As a binary
relation, f = {(c,0?) | z € R}. When this function is graphed as usual in the
‘ry-plane, one is, in fact, making a picture of the ordered pairs which comprise
it,
2.3. DEFINITIONS. Let f: A> B be a function from A to B.
© The domain of f, written dom f, is the set A.
© The range or image of f, written mg f, is
ing f = {be B| (a,b) € f for some a € A}
= {be B| b= f(a) for some a€ A}.
e It is onto or surjective if mg f = B; that is, if and only if
beB = b=f(a) forsomeae A.
# It is one-to-one (1-1) or injective if and only if
f(a) = faa) => 1 = 2
# It is a bijection or bijective function if it is both one-to-one and onto.
‘The definition of one-to-one is sometimes stated via the contrapositive of the
definition given above. A function f is one-to-one if and only if
a Fan => f(a) # f(a2)-
Different inputs to f yield different outputs.
Some Discrete Examples
3,4}, B= {x,y,z} and
f = {(1,2), (2,9), (8,2), (4,y)}
‘Then f is a fimetion A> B with domain A and range B. Since mg f = B, f is
onto. Since f(2) = f(4) (= y) but 2 #4, f is not one-to-one. [In fact, there can
exist no one-to-one function A+ B. Why not? See Exercise 26(a),]
Example 1. Suppose A= {1 66 Chapter 2. Functions
Example 2. Suppose A= {1,2,3}, B= {x,y,z
f= {(1,w), (2,9); (3,0)}-
‘Then f: A> Bis a function with domain A and range {w,y,x}. Since mg f #
B, f is not onto. [No function A ++ B can be onto. Why not? See Exercise 26(b),]
This function is one-to-one because f(1), f(2), and f(3) are all different: f(a,
F(a2) => a = a9. 1
} and
Example 3. Suppose A = {1,2,3}, B= {v,y,2},
L={G,2),2),B,w)} and g = {(1,2), (2,4), (3,2)}.
Then f and g are functions from A to B. The domain of f is A and domg = A
too. The range of f is {z, y}, which is a proper subset of B, so f is not onto. On
the other hand, g is onto because mg g = {z,y,} = B. This function is also one-
to-one because (1), 9(2), and g(3) are all different: g(a) = g(a) => a; = ap.
Notice that f is not one-to-one: f(2) = f(3) (=y), yet 243.
v Example 4. Let f: Z -+Z be defined by f(x) = 2 ~3. Then dom f = Z. To
find mg f, note that
bemgf <= b=2a—3 for some integer a
<> b=2%(a—2)+1 for some integer a
and this occurs if and only if b is odd. Thus, the range of f is the set of odd
integers. Since mg f # Z, f is not onto. It is one-to-one, however:
(2) F(e1) = f@2) => 2x1 -
tg -3 => 21 = 22.
- Example 5. Let f: NN be defined by f(x) = 2x—3. This might look like a
perfectly good function, as in the last example, but actually there is a difficulty.
If we try to calculate f(1), we obtain f(1) = 2(1) ~3 = —1 and —1 ¢ N. Hence,
no function has been defined. aa
Problem 2. Define f: ZZ by f(x) =
not f is one-to-one and/or onto.
— 5x +5. Determine whether or
Solution. To determine whether or not f is one-to-one, we consider the pos-
sibility that (#1) = f(e2). In this case, x} — 5a, +5 = 23 — 522 +5, so
vj — £3 = 5a — Say and (x ~ x2)(ay + 22) = 5(a%, — 22). This equation in-
deed has solutions with x # 22: Any 1,22 satisfying 21 + 2 =5 will do, for Pause 1.
2.1. Domain, Range, One-to-One, Onto 67
instance, 2; = 2, m2 = 3. Since f(2) = f(3) = -1, we conclude that f is not
one-to-one.
Is f onto? Recalling that the graph of f(x) = 2? —5e+5,2€R,isa
parabola with vertex (3, —3), then clearly any integer less than —1 is not in the
range of f. Alternatively, it is easy to see that 0 is not in the range of f because
2? — 5¢ +5 = 0 has no integer solutions (by the quadratic formula). Either
argument shows that f is not onto. oO
Problem 3. Define f: Z+ Z by f(c) =
f is one-to-one and/or onto.
a — x. Determine whether or not
Solution. Suppose f(e1) = /(w2) for 1,22 € Z. Then
30? — ay = 38 — a => 3(29 — 23) = 2,
=> 3(a1 — a2)(0? + eyx9 + 23) = 2-29.
If ay # ©2, we must have x? + x22 + 23 = 1, which is impossible since 21 and
rq are integers. Thus, x1 = a2 and f is one-to-one.
Is f onto? If yes, then every integer b could be written in the form b =
f(a) = 3a — a for some integer a. This seems unlikely and, after a moment’s
thought, it occurs to us that the integer b = 1, for example, cannot be written
this way: 1 = 3a5 — a for some integer a implies a(3a? — 1) = 1. But the only
pairs of integers whose product is 1 are the pairs 1,1 and —1,—1. So here, we
would require @ = a? — 1 = 1 or a = 3a? — 1 = ~1, neither of which is possible.
The integer ) = 1 is a counterexample to the assertion that f is onto, so f is not
onto. o
Define g: Z — Z by g(x) = 227 +72. Is g onto? Is g one-to-one?
Functions of a Real Variable}
Example 6. Let f: R + R be defined by f(«) = 2c —3. The domain of f
is R and mg f = R since any real number y can be expressed y = 2x — 3 (for
x = 3(y+3)). Graphically, this means that any horizontal line intersects the
graph of y = 2x — 3, (See Fig. 2.1.) Since rng f = R, f is onto. It is also one-to-
one (the argument in (2) can be used again), so-being onto and one-to-one, it is
a bijection from R to R. ey 68 Chapter 2. Functions
FIGURE 2.1: The graph of y = 2r — 3.
Example 7. Let g: R + R be defined by g(x) = 2. The domain of g is R;
the range of g is the set of nonnegative real numbers. Since this is a proper
subset of R, g is not onto. Neither is g is one-to-one since 9(3) = g(~3), but
32-3, peace
Example 8. Define h: [0,00) + R by h(x) = 2. This function is identical
to the function g of the preceding example except for its domain. By restricting
the domain of g to the nonnegative reals we have produced a function h which
is one-to-one: A(#1) = h(w2) => 2? = 23 => 2 = +29. Then, since 2 > 0
and a > 0, we must have a = rp. ae,
Example 9. Let, f: R— R be defined by f(x) = 323 — x. Students of calculus
should be able to plot the graph of f and see immediately that f is onto, but
not one-to-one, (See Fig. 2.2.) ey
FIGURE 2.2: The graph of y = 32° — 2, 21. Domain, Range, One-to-One, Onto 69
2.4 REMARK. Contrast this last example with Problem 3 on page 67. There we saw
that the function with the same rule as the one here but with Z as domain was
one-to-one but not onto. A function is more than a rule; domain and range are
critical too.
2.5 THE ABSOLUTE VALUE FUNCTION. The absolute value of a number 2, de-
noted |x|, is defined by
j<{ = fez0
“\-c ife R defined by f(z) =
Lx]. (At one time, the floor function was commonly called the greatest integer
function.) Similarly, the ceiling function is the function g: R > R defined by
J(2) = [x]. The domain of these functions is R and both have range Z. The
graph of the floor function is shown in Fig. 2.3. 70 Chapter 2. Functions
Fumetions whose graphs resemble the floor function are encountered fre-
quently. In October 1996, the Canadian post office set postal rates as shown
in Table 24. The graph of this postal function would certainly resemble the
floor function, with jumps at 30, 50, 100, and 200 gms.
Weight w (gms) Cost
w<30 $0.45,
30 A
defined by ¢4(a) =a for all a € A. In terms of ordered pairs,
ta = {(a,a) | ae A}.
When there is no possibility of confusion about A, we will often write v, rather
than ta.
Notice that the graph of the identity function on R is the familiar line with
equation y = 2.
The identity function on a set A is indeed a function A A since, for any
a6 A, there is precisely one pair of the form (a, y) 4, namely, the pair (a, a),
Pause 2. Prove that the identity function on a set A is one-to-one and onto.
Answers to Pauses
1. Since the graph of g(x) = 20? + 72, x € R, is a parabola, g is not onto.
To determine whether or not g is one-to-one, we must see if g(r1) = g(:r2)
implies that 2; must equal x2. Suppose 9(«1) = 9(xo). Then 22? + 7x, =
203 + Tra and 80 20; — a2)(0, +22) = 72g — ti). If x1 # a, then 2.1, Domain, Range, One-to-One, Onto 71
1+ 22 = —$, which is impossible for 21,09 € Z. ‘Thus B= 22; 9 is
one-to-one,
2 Tf v(ar) = e(aa), then ar = a9 (since o(a;) = ay and e(a2)
one-to-one. To show it is onto, we have to show that given
Some x such that a = 1(z). It is evident that 2 = a works.
= a2), soe is
a A, there is
EXERCISES
1. Determine whether or not eacl of the following relations is a function with
~ domain {1,2, 3,4}. For any relation that is not a function, explain why it
isn’t.
(a) [BBP f= {(1, 1),(2,1),(3,1), (4, 1), (3,3)}
(b) £ = {(1,2), 2,3), (4,2)}
(©) £={(1,1), (2,1), (3,1), (4,1)}
(4) f= {(1,1), (1,2), (1,3), (1,4)}
(©) F = {(1,4), (2,3), (8,2), (4,1)}
»
Suppose A is the set of students currently registered at the University of
Alberta, B is the set of professors at the University of Alberta, cut Cis the
set of courses currently being offered at the University of Alberca Under
what conditions is each of the following a function?
(@) [BB] {(a,0) | a is taking a course from D}
(b) {(a,0) | a's first class each week is in c}
(©) {(a,0) | a has a class in ¢ on Friday afternoon}
Explain your answers.
5. Suppose A and B are nonempty sets. Can Ax B ever bea function A + B?
Explain.
4, Give an example of a function N + N which is
(a) [BB] one-to-one but not onto;
(b) onto but not one-to-one;
(c) neither one-to-one nor onto;
(d) both one-to-one and onto.
Remember that [BB]
‘means that an answer or solution can be found in the Back of the
Book, 72. Chapter 2. Functions
5. Let X be the set of all countries in the British Commonwealth and ¥ be
the set of all people who live in these countries,
(a) [BB] Show that Prime Minister: X —s ¥ and Domicile: ¥ + ¥ are
functions.
(b) Show that Prime Minister is one-to-one but not onto.
(©) Show that Domicile is onto but not one-to-one,
8. Give cxamples of some functions which People commonly encounter whose
Braphs resemble the graph of the floor function,
7. Let S = {1,2,3,4,5} and define f:S-4Z by
f(a) =f P+1 ifwiseven
~ | 2-5 ifzis odd
Express f as a subset of $x Z. Is f one-to-one?
8. The familiar operations of addition and multiplication of real numbers are
functions add, mult: R x RR, where
add(z,y)=2+y; mult(e, y) = zy
(a) [BB] Is add one-to-one? Is it onto?
(b) Is mult one-to-one? Is it onto?
Explain your answers,
8. For any x > 0, let Vz denote the nonnegative square root of . ‘Thus,
vu > 0 and (\/z)? = 2,
(a) Prove that Vz? = |x| for any 2 ER.
(b) Prove that fa + y| < le] + [yl for any x,y € R, (This important
inequality is called the triangle inequality.)
10. [BB] Define g: Z-+ B hy 9(2) = |n|+1, Determine (with reasons) whether
OF not g is one-to-one and whether or not it is onto in each of the following
cases.
(a) B=z
(b) B=N
11. Define f: A+ A by f(z) = 32 4.5, Determine (with reasons) whether or
not f is one-to-one and whether or not f is onto in each of the following
cases.
(a) [BB] A=Q
(b) A=N 2d
12.
13.
14.
15,
16.
17.
Domain, Range, One-to-One, Onto 73
Define h: A+ A by h(x) = 2? +2. Determine (with reasons) whether or
not h is one-to-one and whether or not h is onto in each of the following
cases.
(a) A=Z
(b) A=N
Define g: A + A by g(x) = 3c? + 14x ~ 51. Determine (with reasons)
whether or not g is one-to-one and whether or not g is onto in each of the
following cases.
(a) A=Z
(b) A=R
Define f: A + B by f(x) = 2 + 14x — 51. Determine (with reasons)
whether or not f is one-to-one and whether or not it is onto in each of the
following cases.
(a) [BB] A=N, B= {b€Z|b>-100}
(b) A=Z, Bas in (a)
(c) A=R, B= {bER|b > -100})
Find the domain and range of each of the functions of a real variable given
below. In each case, determine whether the function is one-to-one and
whether it is onto.
(a) [BB] f: RR defined by f(x)
(b) g: RR defined by g(x) = 2|
(c) 8: (4,00) + R defined by A(z) = loga(32 — 4)
(d) f: RR defined by f(x) = 2°-1 +3.
(a) [BB] Define f: R + R by f(z) = 323 +2. Graph f in order to
determine whether or not f is one-to-one and/or onto.
(b) [BB] Define f: Z + Z by f(x) = 32% +x. Determine (with reasons)
whether or not f is one-to-one and/or onto.
(a) Define g: R + R by g(x) = 23 —x+1. Graph g in order to determine
whether or not g is one-to-one and/or onto.
(b) Define g: Z + Z by g(x) = 28 +1. Determine (with reasons)
whether or not g is one-to-one and/or onto.
(c) Repeat (b) for the function g: N+ N defined by g(x) 4
18.
19.
20.
21.
22.
Chapter 2. Functions
Determine whether or not each of the following defines a one-to-one and/or
an onto function. Hither give a proof or exhibit a counterexample to justify
every answer.
(a) [BB] f(n,m) = 2n+3m; f:NxN—+N
(b) [BB] f(n,m) = 2n 43m; f:ZxZ3Z
(c) f(n,m) = 14n + 22m; f: Nx NON
(A) f(r, m) = 89n +246m; f: ZxZ + Z (Hint: 1 = (-17)(246) +47(89),]
(e) fam) =n? +m? 41; f:ZxZ5N
(f) f(n,m) = [2] +1; 2 NxNON
For each of the following, find the largest subset A of R such that the given
formula for f(¢) defines a function f with domain A. Give the range of f
in each case.
(@) (BB) s(«) =
1
(b) f(x) =
In each of the following cases, explain why the given function is not one-to-
one. Then restrict the domain of the function so as to make a one-to-one
function.
(@) f= {(a.4), (6,8), (c,7),(d,a)}
(b) [BB] f: RR defined by f(x) = —da? +122 9
(©) f: RR defined by f(r) = sin
Let 5 be a set containing the number 5. Let A = {f: § + $} be the set
of all functions $+ $. For f,g € A, define f ~ g if f(5) = 9(5).
(a) Prove that ~ defines an equivalence relation on A.
(b) Suppose $ is the set {5,a,8}. Find the, equivalence class of f =
{(5,a), (a, 8), (6,6)}.
Let A be a set and, let f: A+ A bea function. For z,y € A, define 2 ~ y
if f(x) = f(y).
(a) Prove that ~ is an equivalence relation on A,
(b) For A =R and f(«) = [2], find the equivalence classes of 0, 2, and
3
3
(c) Suppose A = {1,2, 3,4,5,6} and
f= (01,2), QA), 8,0) (4,5), 6,6). (6,0)
Find all equivalence Apsses. Bi.
23.
24.
25.
26.
2r.
28.
29.
Domain, Range, One-to-One, Onto 75
[BB] Let X = {a,b} and Y = {1,2,3}.
(a) List all the functions from X to Y and from ¥ to X
(b) List all the one-to-one functions from X to ¥ and from Y to X.
(c) List all the onto funetions from X to ¥ and ¥ to X.
Let X = {a,b,c} and Y = {1,2,3,4}.
(a) How many one-to-one functions are there from X + Y and Y > X?
In each case, list all the functions.
(b) How many onto functions are there X > Y and Y + X? In each
case, list all the functions.
Given sets X and Y with |X| = m and [Y| =n, guess a general formula
for the number of functions X + Y. (A table showing the possible values
for each possible pair m,n € {1,2,3,4} may be helpful.)
Suppose A and B are sets, A containing n elements and B containing m
elements.
(a) Prove that ifn > m, then no function A> B can be one-to-one
(b) Prove that if m > n, then no function A+ B can be onto.
(a) [BB (=> )] Suppose that A and B are sets each containing the same
(finite) number of elements and that f: A > B is a function, Prove
that f is one-to-one if and only if f is onto.
(b) Give an example of a one-to-one function N > N which is not onto.
Does this contradict (a)? Explain.
(c) Give an example of an onto function N + N which is not one-to-one.
Does this contradict (a)? Explain.
[BB] Let f: R > R be defined by f(z) = 2 + [x].
(a) Graph this function.
(b) Find the domain and range of f.
Define s: RR by s(x) = |. Is s one-to-one? Is it onto? Explain. 76 Chapter 2. Functions
2.2 Inverses and Composition
The Inverse of a Function
Suppose that f is a one-to-one onto function from A to B. Given any b € B,
there exists a € A such that f(a) = b (because f is onto) and only one such a
(because f is one-to-one). Thus, for each b € B, there is precisely one pair of the
form (a,b) € f. It follows that the set {(b, a) | (a,b) € f}, obtained by reversing
the ordered pairs of f, is a function from B to A (since each element of B occurs
precisely once as the first coordinate of an ordered pair).
2.8 DEFINITION. A function f: A - B is said to have an inverse if and only if
the set obtained by reversing the ordered pairs of f is a function B + A. If
f: A— B has an inverse, the function
(b,a) | (a,b) € f}
is called the inverse of f.
Example 10. If A= {1,2,3,4} and B = {e,y,z,¢}, then
f ={(1,2), (2,y), (3,2), (4,0)}
is a one-to-one onto function from A to B and, reversing its pairs, we obtain a
function B— A:
£7 = {(e1),(y.2), (3), (64)}-
ts
Note that f~1, which is pronounced “f inverse,” doesn’t mean 4! It is simply
the name of a certain function, the inverse of f.
We have just seen that any one-to-one onto function has an inverse. On the
other hand, suppose that a function f: A + B has an inverse f~!: B > A.
Then f must be onto for, given b € B, there is some pair (b,a) € f~' (since
dom f~! = B), and so the pair (a,b) is in f. Moreover, f must be one-to-one.
Pause 1. Why?
So we have the following proposition. 2.2. Inverses and Composition 77
2.9 PROPOSITION. A function f:A— B has an inverse B > A if and only if f is
one-to-one and onto.
For any function g, remember that (x,y) € g if and only if g(x) = y; in
particular, (b,a) € f-! <=> a= f-"(b). Thus,
a= f(b) => bale sf! <> (abe Ff <> fla)=o.
The equivalence of the first and last equations here is very important:
ee ee ee
(3) a=f-(b) <=> f(a) =o.
For example, if for some function f, x = f~!(—7), then we can conclude that
J (n) = ~7. If f(4) = 2, then 4 = f-i(2)
The solution to the equation 27 = 5 is x = § = 2-1.5, Generally, to solve
the equation ax = b, we ask if a 0, and if this is the case, we multiply each
side of the equation by a~!, obtaining z = a~! = ©. Since all real numbers
except 0 have a multiplicative inverse, checking that a # 0 is just checking that
@ has an inverse.
Look again at statement (3). One solves the equation f() = y for « in the
same way one solves au = } for 2. One first asks if f has an inverse, and if it
does, applies /~” to each side of the equation obtaining x = f(y),
The “application” of ft to each side of the equation y = f(z) is very much
like multiplying each side by ~1. “Multiplying by f-!” may sound foolish but
there is a context (called group theory) in which it makes good sense. Our intent
here is just to provide a natural way to remember the fandamental relationship
expressed in (3).
Example 11, If /: R > R is defined by f(x) = 2x ~ 3, then f is one-to-one
and onto, so an inverse function exists. According to (3), if y = f-'(e), then
@= f(y) = 2y-3. Thus, y = 3(e +3) = f-1(2), =
Example 12. Let A= {x eR |@<0},B={teR|[x> O} and define
J: A > B by f(z) = 27. This is just the squaring function with domain re-
stricted so that it is one-to-one as well as onto. Since f is one-to-one and onto,
it has an inverse. To obtain f-"(x), let y = f-1(2), deduce (by 3) that f(y) =2
and so y? =. Solving for y, we get y = +. Since x = fly), ye A, 804 <0.
Thus, y = —/z; f-! —ve. ——
Example 13. Denoting the positive real numbers by R*, let f:R Rt be
defined by f(z) = 3°. Since f is one-to-one and onto, it has an inverse. To
find f~"(¢), let y = f-'(z) and write down its equivalent form, + = f(y) = 3%,
Solving for y, we get y = log, x. So f-'(z) = log, x. aa) 78 Chapter 2. Functions
4x
we
1s f one-to-one? Find mg f. Explain why f: A — mg f has an inverse. Find
dom f-", mg f-1, and a formula for f-1(2).
, 4a 4b
Solution. f(a) = f(b) — iat = aT
= > 8ab—4a = 8ba — 4b
=> a=b, so f is one-to-one.
Problem 4. Let
= {a |x 3} and define f: A R by f(a) =
Next, yemgf y= f(x) for some x € A
for some z € A
for some « € A
= 2(2y-4)=y for some z € A,
iy # 2, then 2y ~4 #0 and so, dividing by 2y — 4, we can solve for x. Let
B = {y €R| y #2}. We have just shown that Bc mg f. Conversely, if
HE meh, then there exists « € A such that 2(2y—4) =y. If'y = 2, we obtain
0 = y, which is not true. ‘Thus y #2 and ings ¢ B. AN this shows that
mg f = B.
Since f: A + B is one-to-one and onto, it has an inverse FB A.
Also, dom f-* = mg f = B and mg f-! = domf = A. To fied f-'(@), set
y=f-(2). Then
2= fly) =—
r= ia) = ae
and, solving for y, we get y = ea = f(a). a
Composition of Functions
2.10 DEFINITION. If f: A+ B and g: B+ are functions, then the composition
of g and f is the function go f: A+ C defined by (9 f)(@) = 9(f(a)) for all
aed.
Example 14. If A = {a,, ch, B= {z,y} and C= {u,v, w}, and if f: A+ B
and g: B + C are the functions
f= {(a,2),(,u),(c2)}, 9
(z,u); (y,w)}, 2.2. Inverses and Composition 79
then
(90 f)(a) = o(F(a)) = 9(z) =u,
(9° ®) = 9(f) = 9y) = wv,
(9° A)(6) = 9(F() = g(x) =u
and 80 go f = {(a,u),(b,w), (¢,u)}. moll
Tt is seldom the case that 90 f = fog. In this last example, for instance,
£09 is not even defined. One must be careful then to distinguish between “the
composition of g and f” and “the composition of f and 9.”
Example 15. If f and g are the functions RR defined by
S(z)= 22-8, 92) = 2? 41,
then both go f and fo are defined and we have
(90 f)(2) = 9(F(@)) = g(20 — 3) = (22-3)? 41
and
(Fog)(z) = f(o(a)) = f(a? +1) = 22? +1) -3,
Example 16. In the definition of go f, it is required that mg f © B = domg.
Tf: R-+ Rand g: R {1} +R are the functions define by
f(z) =22-3 and ga)=
then go f is not defined because mg f = R ¢ dom .g- On the other hand, f og
is defined and
«
(Fea\(z) =2/
Peel
Example 17. If f: A+ Ais any function and 1 denotes the identity function
on A, then fot is a function A+ A. Also for any a ¢ A, (Fow(a) = f(e(a)) =
f(a). Similarly, (v0 f)(a) = f(a) for any ae A. Thus, for=f=vof. It is
for just this reason that « is called the identity function on A; it behaves with 80 Chapter 2. Functions
Tespect to o the way the mumber 1 behaves with tespect to multiplication,
———I
Under what conditions are two functions f and g equal? Since a function is a
set and two sets are equal if and only if they contain the same elements, it must
be that f = 9 if and only if it is the case that
DEF > (beg,
Suppose f = g and a € dom f. Then (4,0) € f for some 6 = f(a), so (a,b) Eg.
Thus, a € domg and 5 = 9(a). ‘This shows that dom fC dom and, for every
@€domf, f(a) = 9(2). Reversing the roles of f and g, we see similarly that
dom g € dom f and g(a) = f(a) for every a € domg. We are led to the following
Recessary and sufficient condition for the equality of functions,
2.11 EQUALITY OF FUNCTIONS Functions f and g are equal if and only if they
have the same domain and f(a) = 92) for every a in this common domain
Notice how we proved that fio. = f in Example 17. We noted that each of
the functions fo and f has domain A and proved that (f 01)(a) = f(a) for all
aed,
Proof. We must prove that (f 0 9) °h= f(g oh) whenever each of the two
functions—(fog)oh and fo (90h) —is defined. Thus, we assume that for certain
pits A: B, C, and D, h is a fimetion A > B, gis a function BC, and f isa
function C+ D. A direct proof is Suggested and this we provide,
Since the domain of (f 0 9) o h is the domain of f 0 (g 0 h) (namely, the set
A), we have just to Prove that ((f 0g) 0h)(a) = (f° (g0h))(a) for any ae A.
For this, we have
(Leg) oh)(a) = (fo IH(h(a)) = F(g(A(a)))
and
(Fo (gon)\(a) = f(g °h)(a)) = F(9(A(a)})
as desired. o
If: A— B has an inverse (1; By A, then, recalling (3),
IW =a = b= Fa), 2.2. Inverses and Composition 81
we see that for any a € A,
a= f) =F F@) = fo f(a);
in other words, the composition f-? o f = v4, the identity function on A. Simi-
larly, for any element 6 € B,
=f@ =fF7®) = fof;
thus, the composition fo f—! = tp is the identity function on B. We summarize.
2.13 PROPOSITION. Functions f: A + B and g: B > A are inverses if and only if
go f =ta and f og = ip; that is, if and only if,
9(f(a)) =a and f(g(b)) =b for alla € A and all be B.
Problem 5, Show that the functions f: R + (1,00) and g: (1,00) + R defined
by
S(@)=9*+1, gle) =f logs(z- 1)
are inverses.
Proof. For any 2 €R,
(9° f)(@) = 9(f(2)) = 9(3* +1)
= 3(logs[(3"* + 1) — 1)
= }(log, 3?*
and for any x € (1,00),
(fo 9)(x) = f(9(«)) = f(5 logs(« — 1))
= 32CF lows (2—-1)) 4
= gless(e—1) 4 4
-1)4+1
By Proposition 2.13, f and g are inverses. a
We have noted the similarity in the solutions to the equations az = 6 and
J(x) = y. Ifa (or f) has an inverse, we multiply by this inverse and obtain
x =a~'b (or « = f(y). It is interesting also to observe that the connec-
tion between a and its inverse a~+ (their product is 1) is strikingly similar to
the connection between f and its inverse f~} (their composition is the identity
function), 82 Chapter 2. Functions
Answers to Pauses
1. Suppose f(a1) = (a2) for some an a2 € A. Let b= fas) (= f(a3)),
Then (a1,5) and (a2,d) are both in F, s0 (0,41) and (6,a2) are both in
f"". But f~' is a function and so exch clement of B occurs just once as
one-to-one.
2. Since f: A + B and ip: B+ %, the composition tp 0 f is a function
A B. Thus, domip of = A= dom f. Also, for any a €'A, 1g 0 f(a) =
ta(F(a)) = F(a), because F(a) € B and 1y(b) = b for all b € B. Thus, the
functions 15 0 f and f are equal,
EXERCISES
1. Find the inverses of each of the following functions, being sure to specify
the domain and range of each inverse
(a) [BB] f: QQ given by f(z)
(b) fi RR given by f(e) = 299
(©) 9: RR given by g(z) = 2a.
(d) B: (§,00) + R given by (2) = log, (3x — 4),
ar + 5.
©
| Define f:Z + N by f(2) { 2c] ifx9
(a) Show that f has an inverse,
(b) Find ~1(2586)
3. Suppose is the set of all married People, mother: A + A is the func-
tion which assigns to each married person his/her mother, and father and
phouse have similar meanings. Give sensible interpretations of each of the
following:
(a) [BB] mother o mother
(b) mother father
(C) fathero mother
(4) mother o spouse
(e) spouse o mother
(f) father o spouse
(8) spouse o spouse
(h) (spouse o father) o mother
@) spouse (father o mother) 2.2. Inverses and Composition 83
4, Let $= {1,2,3,4,5} and T = {1,2,8,8,9} and define functions f: $ +7
and g: S$ by
+ (3,9), (4,3), (2,1), (5, 2)},
+2),(3, 1), (2,2), (4,3), (5, 2)}
(a) Find fog or explain why f og is not defined. Repeat for go f, fo f,
and gog.
(b) Which of f, g are one-to-one? Which are onto? Explain.
(c) Find f~* if it exists. If it doesn’t, explain why not.
(d) Find g~ if it exists. If it doesn’t, explain why not.
e
Define functions f,9,h: Z—+Z by
F(n)=(-1)", g(n) =2n, h(n) =2n 41.
Find fog(n), goh(n), hog(n), foh(n), fogoh(n), hofog(n), gofoh(n),
and fohog(n).
6. Let S = {1,
4} and define functions f,g: S + $ by
f= {(1,3), (2,2), (3,4), (4,1)}
9= {(1,4),(2,3),(3,1),(4,2)}.
Find
(a) [BB] 4° fog
(b) fog tog
(c) gofog™
(d) gogtof
(e) frtogtofog
7. Let S = {1,2,3,4,5} and answer all parts of Exercise 6 for the functions
fig: 5+ S defined by
f= {(1,3), (2,4), (3,2), (4,5), (5,1)},
9 = {(1,4), (2,5), (3,2), (4, 1), (5,3)}-
@
. [BB] Let f, g and h: R— R be defined by
f(e)=e+2, g(x) = war Me) =3.
Compute go f(x), fog(x), hogo F(z), goho f(z), go ft ° f(x), and
{tego f(z). 84 Chapter 2. Fy unctions
9. Let A De these A= (@ER|2>0} and deine 9,8: AR by
& 1
fe= SFT a)=4, ne) ett,
Find 9 f(z), fo 9(z), hogo f(a), and fogoh(a).
10. Let A denote the set R\ {0,1}. Let 1 denote the identity function on 4
and define the functions f,9, A, sr: A A by
(x)
(a) Show that fog = 4 and 9° = 9, as shown in the table below.
Complete the table, thereby Zevealing that the composition of any
‘wo of the given functions is one of the given five, or the identity, 3
fohrs
(b) Which of the given six functions have inverses? Find (and identify)
any inverses which exist,
11. Let A= {1,2,3} and define Sis fas fs, fas fis: A~ Aas follows:
Ar ={(1,1), 2,8), (3, 2)}
f= {(1,38), (2,2), (3,1)}
fs ={(,2), (2,1), (3,3)}
fo={(1,2), (2,3), (3,1)
45 = {(1,3), (2,1), (3,2)}
(a) Show that each composite function f,o Sf; is one of the given functions,
“The table you will construct in this exercise is the multiplication table for an important
ct known as a
mathematical obje: group. This particular group is the smallest one which is not
commutative, 2.2. Inverses and Composition 85
or the identity, by completing a table like the one below.*
fo_fs_fa_fo
(b) Find the inverses of those of the six functions in (a) which have in-
verses.
12/ Let S = {1,2,3,4,5} and let f,g,h: S — S be the functions defined by
f= {(1,2); (2,1); (3,4), (4,5), (5,3)}
9 = {(1,3), (2,5), 3,1), (4,2), 6,4)}
{(1,2); (2,2), (3,4); (4,3), (5, D}
(a) [BB] Find fog and go f. Are these functions equal?
(b) Explain why f and g have inverses but h does not. Find f~1 and g~
(c) Show that (f og)" sgiofif¢fiogt. ,
13. Let A= {cr €R| #2} and B={xER| x1}. Define f: A> Band
g: B Aby
F(a) =
(a) Find (f o9)(z).
(b) Are f and g inverses? Explain.
14. Suppose f: A> B and g: B + C are functions.
(a) [BB] If go f is one-to-one and f is onto, show that g is one-to-one.
(b) If go f is onto and g is one-to-one, show that f is onto.
15. (a) [BB] Prove that the composition of one-to-one functions is a one-to-
one function.
(b) Show, by an example, that the converse of (a) is not true.
(c) Show that if go f is one-to-one, then f must be one-to-one.
As with Exercise 10, this table also describes the multiplication for the smallest group
which is not commutative. 86 Chapter 2. Functions
16. (a) Prove that the composition of onto functions is onto.
(b) [BB] Show, by an example, that the converse of (a) is not true.
(c) Show that if go f is onto, then g must be onto.
17. Is the composition of two bijective functions bijective? Explain.
n-2 n> 1000
f(i(n+4)) n< 1000.
(a) Find the values of f(1000), (999), (998), f(997), and f(996).
(b) Guess a formula for f(n).
(c) Guess the range of f.
18. Pete 122-24 0) =|
19. Let t: R—+R be the function defined by t(x) = 2[2| — x.
(a) Graph ¢ and use this to decide whether or not ¢ has an inverse.
(b) [BB] Prove that ¢ is one-to-one without using the graph of t.
(c) Prove that t is onto without using the graph of t.
(d) Find a formula for t-1(2).
2.3 One-to-One Correspondence and
the Cardinality of a Set
In this section we think about the size of a set, or, more correctly, its cardinality.
We show that our natural instincts about the relative sizes of finite sets can be
extended to allow the comparison of infinite sets too, though sometimes with
surprising results. It turns out, for example, that the natural numbers and the
set of all rational numbers have the same “size,” whereas both these sets are
“smaller” than the set of real numbers.
In discussions about cardinality, it is common to employ the term one-to-one
correspondence rather than its synonym, one-to-one onto function. Either of the
statements
“A and B are in one-to-one correspondence” or
“There is a one-to-one correspondence between A and B”
means that there is a one-to-one onto function from A to B. 2.3. One-to-One Correspondence; Cardinality 87
2.14 DEFINITIONS. A finite set is a set which is either empty or in one-to-one corre-
spondence with the set {1,2,3,... ,n} of the first n natural numbers, for some
nN. A set which is not finite is called infinite.
If Ais a finite set and A # 0, then, for some natural number n, there exists a
one-to-one onto function f: {1,2,3,...,n} + A. Letting f(é) = a;, this means
that A = {a1,@2,... ,an}.
If A is a finite set, the cardinality of A is the number of elements in A; this
is denoted |A|. Thus, || = 0 and, if A = {a,a2,... ,a,}, then [A| =n.
Examples 18.
* I{a,b,2}| =3,
© the letters of the English alphabet comprise a set of cardinality 26,
© {ER |x? +1=0}|=0.
anes
How might we determine if two finite sets contain the same number of ele-
ments? We could count the elements in each set, but if the sets were large, this
method would be slow and highly unreliable. Confronted with two enormous
pails of jelly beans and asked to determine if the number of jelly beans in each
pail is the same, the best strategy would be to pair the beans in the two pails.
Remove one jelly bean from each pail and lay the two aside; remove another jelly
bean from each pail, lay these aside, and so on. If the last jelly beans in each
pail are removed together, then certainly we would know that the numbers in
each pail were the same.
If the numbers are the same, it is important to observe that the set of pairs
of beans removed from the pails defines a one-to-one onto function from the jelly
beans in one pail to those in the other, Different beans in the first pail are paired
with different beans in the second (one-to-one) and every bean in the second pail
is paired with some bean in the first (onto),
If two finite sets A and B have the same cardinality—we write [4| = |B|—
then there is a one-to-one onto function from A to B. Conversely, if there is a
one-to-one onto function from A to B, then |A| = |B|. This idea allows us to
extend the notion of same size.
2.15 DEFINITION. Sets A and B have the same cardinality and we write |A| = |B], if
and only if there is a one-to-one correspondence between them: that is, if and
only if there exists a one-to-one onto function from A to B (or from B to A).
Example 19. a ++ 2, b+» y is a one-to-one correspondence between {a,b}
and {x,y}; hence, |{a,b}| = |{,y}| (= 2.) —1 88 Chapter 2. Functions
Example 20. The function f: N— NU {0} defined by f(n) =n~1isa
one-to-one correspondence between N and NU {0}: so IN| = [NU {o}].
=e)
Example 21. ‘The function f: Z—4 97 defined by f(n) = 2n isa one-to-one
correspondence between the set Z of integers and the set 2Z of even integers;
thus, Z and 2Z have the same cardinality, a |
Many people find the second and third examples here surprising and perhaps
mildly disconcerting; nevertheless, they must be accepted. Remember that our
definition of same cardinality coincides with what we know to be the case with
finite sets.
We can prove that {a,d} and {z, xy} have the same cardinality (without using
the word two) by pairing the elements in the two sets
with 2n.
Many sets have the same cardinality as the natural numbers,
Problem 6. Show that |N| = [2].
Solution. We must find a one-to-one correspondence between N and Z. Here
is one.
L1—- 0
25 1
34 -1
A+ 2
5 > ~2
6+ 3
Formally, this is just the fimetion f: N+ Z defined by
n/2 ——ifnis even
soy={ —(n-1)/2 ifnis odd
which is certainly both one-to-one and onto. a
PAUSE 1. Can you express this function f with a single equation?
Here's another example which, at first sight, may seem mystifying, 2.3. One-to-One Correspondence; Cardinality 89
Problem 7. Show that the set Rt of positive real numbers has the same
cardinality as the set (0,1) = {2 € R | 0 < z < 1} of real numbers between 0
and 1
Solution. Let f: R+ —+ (0,1) be defined by
z
1G) +1"
We claim that f establishes a one-to-one correspondence between R+ and (0,1).
To show that f is onto, we have to show that any y € (0, 1) is f(2) for some
positive real number x. But
z
etl
y = 9
which is positive since y > 0 and 1~y > 0 (remember that 0 < y < 1). Therefore,
y
€ (0,1) = y= f(—_
ve (1) (=>)
showing that f is onto. Also f is one-to-one because
m+l 29+]
F(t1) = f(e2) =>
=> ay +0, = 20g +29
=Sn=m Oo
Just as cardinality partitions finite sets into classes, the sets in each class
having the same number of elements, so cardinality also partitions infinite sets.
For our purposes, we consider just two classes of infinite sets.
2.16 DEFINITIONS. A set A is countably infinite if and only if |A| = |N| and countable
if and only if it is either finite or countably infinite. An infinite set which is not
countable is uncountable.
The symbol Xo (pronounced “aleph naught”) has traditionally been used to
denote the cardinality of the natural numbers. ‘Thus, a countably infinite set has
cardinality Np. For example, in light of previous discussion, [NU {0}] = |Z] = No.
Countably infinite sets, as their name suggests, are those whose elements can be
listed in a systematic and definite way, because to list them is to rank them as
first, second, third, and so on, and this ranking establishes the required one-to-
one correspondence with N.
To establish |Z] = |N] in (6) above, we merely listed the integers in a sys-
tematic way—0,1,—1,2,—2,.... This list then gave the desired one-to-one cor
respondence with N. PAUSE 2.
PAUsE 3.
90 Chapter 2. Functions
By the phrase “in a systematic and definite way” we mean that when one per-
jan yerminates the list (with three dots), another should be able to continue. ‘This
ituplies that in a good list it should be possible to determine the position of any
clement. It is not acceptable to “list” the integers as 0, 1,2, 3,.. »-1,-2,-3,...
Pecatise —1 does not have a position: 0 is the first mmiber, 1 ie the second, but
in what position is -17 Similarly, it is not acceptable to “list” the integers as
-+ 3, —-2,—1,0,1,2,3,... because no integer has a position. What's the first
clement in the list? In fact, this is not a list at all. On the other hand, listing
She integers tla way=—D, 1, <1, 2)-2) 9.9), .—easign every integer a definite
position.
In the list 0, 1, —1, 2, -2, 3, —3,..., —1is the third number and 3 is the sixth. In
what position is 1003? What number is in position 1003?
The following argument establishes the remarkable fact that the set Nx N=
{(m,n) | m,n € N} is countable; it has the same cardinality as N. We simply
The arrows indicate the order in which the elements of N x N should be listed—
(1,1), (2,1), (1,2), (1,3), (2, 2),.... Wherever the arrows terminate, there is no
difficulty in continuing, so each ordered pair acquires a definite position.
ua) (2,4) (8,4) (4,4)
i
(1,3) (2,3) (3,8) (4,3)
TN
(1,2) (2,2) (3,2) (4,2)
N
GY)(21) (31)4(4,1) (5,1)
FIGURE 2.5: The elements of N x N can be systematically listed by
starting at (1,1) and following the arrows,
Suppose at the point (m, n) in this listing, we write the Positive rational “.
Then every rational number appears in the picture, actually many times over,
since, for instance, } appears beside the points (2,3), (4,6), (6,9), (40,60), and
80 on. By ignoring the repetitions of a given fraction, the set of positive rationals
can be enumerated. We conclude that the set of positive rationels is « countable
set.
The scheme which we have just described for listing the positive rationals begins
3 2 1/2, 1/8 (ignore the number 2/2 which has already been listed), 3, 4, 3/2,
2/3, 1/4. Continue the list in order to determine the position of 3/4.
Not so long ago, there came to light another way to establish a one-to-one
correspondence between the positive rationals and N. For a very neat and direct
approach to an interesting problem, see Exercise 35 of Section 3.3, PAUSE 4.
2.3. One-to-One Correspondence; Cardinality 91
This result is astonishing when you consider how much “bigger” than the
natural numbers the set of positive rationals “seems.” It is nonetheless true
that these infinite sets have exactly the same cardinality. Actually, more is
true. Since we are able to list the positive rational numbers—a},4a2,a3,...—it
becomes easy to list all the rationals, simply by intertwining the positive and
negative rationals, like this—0,a,,—a1,a2,—a,a3,.... What do we conclude?
The entire set Q of rational numbers is countable!
In the enumeration of the rationals just described, in what position is 3/4?
At this point, one might be tempted to conjecture that every set is countable,
but such is indeed not the case! We present another old and well-known arguinent
which shows that the set of real numbers is not countable. Actually it shows
that the set (0,1) = {@ € R| 0 |Al< |B|. Formulate a theorem and prove it
5. Suppose S is a set and for A, B € P(S), we define A = B to mean |A| < |BI.
1s this relation a partial order on P(S)? Explain,
P (@) [BB] Find a one-to-one correspondence between the intervals (1, 00)
and (3,00). What do you conclude about the cardinalities of (1,00)
and (3,00)? (Recall that (4,00) is standard notation for certain set
of real numbers. See Definition 1.8.)
(b) Let a and 6 be real numbers, Find a one-to-one correspondence be-
tween (a,00) and (b, 00) 94 Chapter 2. Functions
A. Show that for any sets A and B, |A x Bl = |B x Al.
8. [BB] Let X, Y and Z be sets,
(a) True ot false: (X x ¥) x Z =X x (¥ x 2). Provide a proof or give
@ counterexample.
(b) Find a one-to-one correspondence (X x ¥) x 7-9 X x (Y x 2).
9. J” [BB] Prove that the subsets (0,1) and (1,2) of R have the same car-
dinality.
(b) Prove that (0,1) and (4,6) have the same cardinality.
(c) Prove that (0,1) and (10,00) have the same cardinality.
10. Prove that the notion of same cardinality is an equivalence relation on the
family of all sets. Explain, with reference to Definition 2.15.
11. Prove that the function defined by f(x) = 3° establishes a one-to-one
Rr eepondence between the real numbers R and the positive real numbers
RY. What can you conclude about cardinality?
12. Prove that each of the following sets is countable by listing its elements in
Hse and definite way. (Show at least the first dozen terms of your
lists.)
(a) all positive and negative powers of 2
(b) those natural numbers which leave a remainder of 1 when divided by
3
(c) [BB] those positive rational numbers m/n with n odd
(a) Nxz
(e) ZxZ
13. Determine, with justification, whether each of the following sets is finite,
countably infinite, or uncountable:
(a) [BBI {ee R|1<2<9}
(b) {ew EQl1