You are on page 1of 7

Limits - Unit 1

Starred problems/topics are for BC students only.
Definition & interpretation of a limit:
L x f
a x
·

) ( lim
means that the function f maps x values near a to y values near L. The notation x → a means to let x
approach a from the right (values greater than a) and from the left (values less than a). The y values get infinitely
close to L as the x values get infinitely close to a, either from the right or the left. It does not matter if f (a) = L or if
f (a) is even defined. As an example,
4
1
lim
!
!
·


x
x
x
since
!

x
x −
can "e made ar"itrarily close to # "y evaluating it at
x values very close to !. $y %ar"itrarily close& I mean that no matter ho' close you(d li)e a y value to "e to #, it is
possi"le to find an x value near ! such that f (x) is as close or closer to #.
1 *nter the function a"ove into a graphing calculator and +oom in near the x value !. a. Approaching from the left, ho'
close to ! do you need to "e in order to ensure that the functional values are 'ithin ,.,1 of ,.!- .
". Approaching from the rigt, ho' close to ! do you need to "e in order to ensure that the functional values are 'ithin
,.,1 of ,.!- .
c. $ased on your ans'ers in parts a and ", ho' close to ! do you need to "e in order to ensure that the y values are 'ithin
,.,1 of ,.!- .
d. /hat interval centered at x = ! 'ill assure that the y values are in the (,.!4, ,.!0) .
To evaluate the limit a"ove 'e are a"le simply to plug in !, "ut it is not al'ays that simple. 1onsider the function
¹
¹
¹
'
¹
· −
≠ −
·
0 , 22
0 , 4 -
) (
x
x x
x g
Then,
!0 ) ( lim
0
·

x g
x
even though g(0) = 322. As far as the limit goes as x approaches 0, 'e are not
interested in 0 itself. /e(re only interested in values very close to 0. In fact, if g(0) 'ere left undefined, the limit 'ould
still "e !0. 4o'ever, "ecause
), 0 ( ) ( lim
0
g x g
x


'e can conclude that g is not continuous at x = 0. (5ore a"out
continuity later.)
! a. 6efine a piece'ise function 'hose limit as x → 1, is , "ut f (1,) = 11.7.
". 1ome up 'ith a function 'hose limit as x → - is "ut is undefined at x = -.
c. Thin) of a function 'hose limit as x → 34 is 1 and is continuous at x = 34.
5ainly due to tidal forces of the moon, the earth(s day length is increasing at a rate of 18-,, of a second per century.
9ur current period of rotation is ! h, -0 min, 4.1 s. As the duration of time that passes from the present approaches !
centuries, 'hat does the earth(s period approach. :tate your ans'er in the form of a limit.
! Sample problem ": If
1; 1!
1! - !
) (
!

− +
·
x
x x
x f , find
) ( lim
! 8
x f
x→
.
:olution< =eplacing x 'ith 8! yields a , 8 , situation. (/e call this an indeterminant form.) >ote that the numerator
approaching +ero tends to ma)e the 'hole fraction +ero, "ut the denominator approaching +ero tends to ma)e the
numerator "lo' up to t ∞. :o, 'e have competing effects, and as far as 'e )no' right no', the limit could +ero, it
could "e t ∞, it could "e any constant, or might not even exist. To ma)e a determination 'e "egin "y factoring<
) ( lim
! 8
x f
x→
) ! ( 0
) 4 )( ! (
lim
! 8

+ −
·

x
x x
x
. 1learly the function is undefined at x = 8!, "ut this is "eside the point 'hen it
comes to evaluating a limit? 'e are only concerned 'ith x values very near 8!. =educing,
) ( lim
! 8
x f
x→
0
4
lim
! 8
+
·

x
x

0
4 ! 8 +
·
# 1181!.
4 @actor to find
12!
4 !
lim

4



x
x
x
.
- 1onsider the function
7
14
7
!
) (



·
x x
x
x f and the limit
) ( lim
7
x f
x→
. If 'e 'ere to ta)e the limit term "y term, 'e(d
find that each the limit does not exist for either term. This is "ecause each term approaches t ∞ depending on the
direction in 'hich x approaches 7. @or either term, since the one3sided limits are different, the overall limit 6.>.*. :o it
'ould appear that the limit of the sum of the terms 'ould not exist either. 4o'ever, the theorem that the limit of a sum is
the sum of the limits only applies 'hen limits exist. @ind this limit "y re'riting f (x) as a single fraction.
0 @ind
; -
;
- 1
lim
-
;



x
x
x
"y simplifying the fraction.
(
! Sample problem "": @ind

2
lim
2



x
x
x
.
:olution< 4ere again 'e have a , 8 , indeterminant situation. 9ne 'ay to do this is to multiply numerator and
denominator "y the conAugate of the denominator.

2
lim
2



x
x
x
=
) )( (
) )( 2 (
lim
2
+ −
+ −

x x
x x
x
2
) )( 2 (
lim
2

+ −
·

x
x x
x
0 2 ) ( lim
2
· + · + ·

x
x
.
7 The last pro"lem can also "e done "y factoring and reducing. 6o this "y forcing the numerator to factor as the
difference of sBuares.
; Crove that
+∞ ·


+

-
!-
lim
!
-
x
x
x
"y factoring and
simplifying. (The 5athematica syntax to find this limit33"ut
not the proof33is sho'n at the right.)
! Sample problem """: Diven
· ) (x f

x x
e
x x
! ) 4 ( sin
!
tan 4
!
+ −

, find
) ( lim
1,
x f
x→
.
:olution< 9stensi"ly this pro"lem is extremely intimidating. 4o'ever, as x → 1,, "oth the numerator and denominator
approach a non+ero constant. Thus 'e simply evaluate f (x) at x = 1,<
) ( lim
1,
x f
x→
1, ! ) 4 1, ( sin
!
) 1, tan( 4 ) 1, (
!
⋅ + − ⋅
·

e

≈ 7.1 × 1,
1!7
, an enormous num"er. (=adian mode is reBuired here.)
2 To get an in)ling of Aust ho' large this num"er is, convert this many millimeters into light years. :ho' your 'or) as a
product of fractions (dimensional analysis). Ese the fact that light travels a"out 1;0,,,, miles in a single second.
1, @or f (x) a"ove, find
) ( lim
11
x f
x→
, ma)ing sure your calculator is in radian mode.
! Sample problem "$ ("y =aBuel =oney, >ayeon Fang, G Ayush 6ulguun)< *valuate the limit<
n
n
n
n
!
!
lim
+
∞ →
.
:olution<
n
n
n
n
!
!
lim
+
∞ →
=

,
_

¸
¸
+

,
_

¸
¸
∞ →
n
n
n
n
n
!
!
!

lim
=

,
_

¸
¸
+

,
_

¸
¸
∞ →
1
!

lim
n
n
= ) 1 ( lim
!

lim
∞ → ∞ →
+

,
_

¸
¸
n
n
n
1 1
!

lim · +

,
_

¸
¸
·
∞ →
n
n
.
:ince ! 8 ≈ ,.;7 H 1, y = ( ! 8 )
x
is an exponential function that decays to +ero. (If you ta)e ;7I of a num"er
over and over, you get closer and closer to +ero. Another 'ay to thin) a"out this limit is to recall geometric seBuences
from last year. 4ere the common ratio "et'een successive terms is r = ! 8 . :ince
, 1 < r
the limit of the
seBuence is +ero. This means our limit is , J 1 = 1 and that the function
x
x
x
x f
!
!
) (
+
·
has a hori+ontal asymptote of
y = 1.
11 @ind
( )
x
x
x
!-
1, -
lim

π +
∞ →
.
! Sample problem $ ("y =aBuel =oney, >ayeon Fang, G Ayush 6ulguun)< *valuate
x
x
x
tan
4
lim
, →
.
:olution< /e 'ill ma)e use of the fact that 1
sin
lim
,
·

x
x
x
.
x
x
x
tan
4
lim
, →
= x
x
x
x
x x
x x
x
x x x x
cos lim
sin
4
lim
sin
cos 4
lim
cos 8 sin
4
lim
, , , , → → → →
× · · 1
sin
lim 4
,
× ·

x
x
x
= 4(1) × 1 = 4. >ote that
1
1
1
sin
lim
) 1 ( lim
sin
1
lim
sin
lim
,
,
, ,
· · · ·


→ →
x
x
x
x
x
x
x
x
x x
.
1! @ind
x
x x
x
1
tan sec -
lim
,
π π

%umerical appro&imation of a limit:
:ome limits are difficult, if not impossi"le, to evaluate analytically. :o, 'e often evaluate the function very close to the
value of interest to approximate the limit.
! Sample problem $"< Ket f (x) =
-
) ! (
cos


x
e x
x
and determine the limit<
) ( lim
-
x f
x


.
.
:olution< >ote that f (-) is not defined, and there is no 'ay to reduce the fraction, so 'e suspect the function has a
vertical asymptote at x = -. If there is a L.A. at x = -, then our limit 'ould come out to "e tM. To test this let(s
evaluate f (x) at values of x slightly less than -. I used 5athematica to do this, "ut you could do it on a calculator 'ith
the ta"le feature. @irst I defined the function<
fx_:3 2 x E^Cos3 xx 5
5athematica is very particular a"out syntax. /hen defining a function use an underscore after the varia"le on the left
side of the eBuation and use a colon "efore the eBual sign. All "uilt3in functions, li)e cosine, as 'ell as user3made
functions enclose the arguments in sBuare "rac)ets, not parentheses. >ext I 'ill evaluate f (x) at five different values of
x very close to -, 'hich are enclosed in "races. (The output is right "elo' this input).
f4.9, 4.95, 4.9993, 4.999999, 4.9999999999
39.8819, 71.7562, 4683.59, 3.2746910
6
, 3.2746910
10

The output sho's the corresponding functional values. :ince these values appear to "e gro'ing 'ithout "ound, 'e
conclude that our limit is J M. 4o'ever, this analysis does not represent a proof.
'symptotes:
Kimits underlie the existence or nonexistence of hori+ontal, vertical, and o"liBue (slant) asymptotes of graphs of functions.
! Sample problem $"": Ket
1 7 -
!- 1-, 7-
) (
!
! 4
− +
+ −
·
x x
x x
x h . a. Ese a limit to prove that h(x) has no hori+ontal
asymptote.
:olution< A function has a 4.A. if its graph approaches some constant as x "ecomes very positive or very negative.
5ore precisely, if the limit as x → t ∞ is a constant, the function has a 4.A. (It is possi"le for the limits to "e different
constants as x → ∞ and x → 3 ∞, such as 'ith y = arctan x, "ut 'ith rational functions, such as h(x), if there is a
4.A., there is Aust one.) Nou should recall from precalculus that, for a rational function (a polynomial over a
polynomial), 'henever the degree of the numerator is larger than that of the denominator, there is no 4.A. and the graph
+ooms off to t ∞ on each side. This is our situation, "ut let(s use a limit to prove this is indeed the case. 9ur techniBue
'ill "e to divide each term, top and "ottom, "y the highest po'er of x in the denominator<
) ( lim x h
x ∞ →

8 1 8 7 -
8 !- 8 1-, 7-
lim
x x
x x x
x
− +
+ −
·
∞ →
. >ext 'e use the follo'ing facts< the limit of a Buotient is the Buotient of the limits? the
limit of a sum is the sum of the limits? and the limit of a constant is the constant. Thus,
) ( lim x h
x ∞ →
( )
( )

8 1 8 7 - lim
8 !- 8 1-, 7- lim
x x
x x x
x
x
− +
+ −
·
∞ →
∞ →

( ) ( )
( ) ( )

8 1 lim 8 7 lim - lim
8 !- lim 8 1-, lim 7- lim
x x
x x x
x x x
x x x
∞ → ∞ → ∞ →
∞ → ∞ → ∞ →
− +
+ −
·
, , -
, , 7- lim
− +
+ −
·
∞ →
x
x
= ∞. /e could use the
exact same process to sho' that
−∞ ·
−∞ →
) ( lim x h
x
. Therefore, h(x) has no 4.A.
". Ese a limit to find the o"liBue asymptote for h(x).
:olution< Nou should recall from precalculus that, for a
rational function, 'henever the degree of the numerator
exceeds that of the denominator "y exactly one, the graph
has an o"liBue asymptote. The techniBue for finding it is to
employ long division. 9f course, you are expected to "e a"le
to do this "y hand, "ut let(s see ho' it(s done 'ith
5athematica. @or convenience 'e define t'o functions, one
for the numerator and one for the denominator. Then 'e use
t'o "uilt3in functions to find the Buotient and remainder.
(The Buotient has "een assigned to q for later reference.)
:o, 'e no' )no' that
1 7 -
4 1-
!1 1- ) (
!
!
− +
+ + −
+ − ·
x x
x x
x x h , from 'hich it is easy to see that the fractional portion of
h(x) shrin)s to +ero as x "ecomes infinitely large. That is, ,
1 7 -
4 1-
lim
!
!
·
− +
+ + −
∞ →
x x
x x
x
, since the degree of the numerator
is less than that of the denominator. This means that as x gets larger and larger in magnitude, the graph of h(x) "egins
to appear more and more li)e the graph of
!1 1- − · x y
. Thus this is the eBuation of our o"liBue asymptote.
1 Crove that the a"ove limit is indeed +ero 'ith the techniBue used in part a.
c. @ind the vertical asymptotes for h(x) and express them in terms of limits.
:olution< In a reduced rational function, L.A.(s
occur at x values in 'hich the denominator is
+ero "ut the numerator isn(t. It is appropriate to
use technology to find the roots of the cu"ic
denominator. The >:olve function is for
solving an eBuation numerically. This means
the output is approximate rather than in
complex radical form. >ote the use of the
dou"le eBual sign 'hen solving an
eBuation. There are three roots of the
denominator, and to rule out the possi"ility
of holes in the graph, 'e evaluate each in
the numerator (in the form of a list
enclosed in "races). :ince the numerator is
not +ero at any of these values, they are
indeed locations of L.A.(s. This means that
h(x) +ooms off to t ∞ at each of these
three x values. Ket(s loo) at a graph to
determine 'hether the graph is heading up
or do'n at these locations. The Clot
function 'as used to graph "oth h(x) and
the o"liBue asymptote, 'hich 'as defined
as q a"ove. The graph ma)es the
follo'ing limits apparent<
+∞ ·

− →
) ( lim
!77 . 1
x h
x
,
−∞ ·
+
− →
) ( lim
!77 . 1
x h
x
,
−∞ ·

− →
) ( lim
40! . ,
x h
x
, and
+∞ ·
+
− →
) ( lim
40! . ,
x h
x
.
14 /rite the remaining t'o one3sided limits for L.A.(s
1- *ach student should come up 'ith an example of his8her o'n of a function 'ith the follo'ing properties< t'o
nonremova"le discontinuities (one Aump, one infinite)? one remova"le discontinuity? a domain of all real num"ers? a
hori+ontal at y = ? x3intercepts at t!? a y3intercept at J7. /rite the one3sided limits at your remova"le and Aump
discontinuities. /rite limits as x → t M. /rite one3sided limits as x approaches the point 'here you have an infinite
discontinuity.
( )ormal definition of a limit:
, , , ) ( lim > ∃ > ∀ ⇔ ·

δ ε L x f
a x
such that O f (x) 3 L O H ε 'henever , H O x - a O H δ.
ε and δ are small, positive real num"ers? ∀means %for each& or %for every&? ∃means %there exists&?
⇔ means %if and only if & or %is logically eBuivalent to&? O x - a O is the distance "et'een the varia"le x and the constant
a? O f ( x ) 3 L O is the distance "et'een the y value corresponding to x and the limit as x approaches a.
This is a formal 'ay of saying that no matter ho' close you(d li)e the y values to "e to L, this can "e al'ays "e
accomplished using x values close to a.
P ! Sample problem $""": 6o a formal ε-δ proof to sho' that
. 11 ) 1 ( lim
4
− · + −

x
x

:olution< Ket ε Q , "e given. >o matter ho' small this ε is, 'e must sho' there is a δ such that 'henever x is 'ithin
δ units of 4, y 'ill "e 'ithin ε units of 311. /e "egin 'ith some preliminary 'or)< The distance "et'een the y
values and the proposed limit is given "y O (3x J 1) 3 (311) O = O 3x J 1! O = O 3(x 3 4) O = O3O ×O x 3 4O = O x 3 4O.
This is 'hat must "e less than ε, no matter ho' small ε is. :o 'e 'rite O x 3 4O H ε. If the limit really is 311, then
even if ε = ,.,,,,,,,,,,,1, then there should still "e a δ such that 'hen the distance "et'een x and 4 is less than δ
("ut not +ero), the corresponding y values 'ill "e 'ithin ,.,,,,,,,,,,,1 of 311. To sho' that such a δ exists for any
conceiva"le ε, 'e have to find a relationship "et'een ε and δ. O x 3 4O H ε implies that O x 3 4O H ε 8. This means
that the distance "et'een x and 4 must "e less than ε 8. Thus 'e choose δ = ε 8. To complete the proof 'e essentially
'or) "ac)'ards and sho' that this is the appropriate choice for δ < If O x 3 4O H ε 8, then O x 3 4O H ε, implying
O x 3 1! O H ε and O 3(x 3 1!) O H ε. :o, O 3x J 1! O H ε, and O (3x J 1) 3 (311) O H ε. /e have used the definition of
a limit to sho' that the as x gets infinitely close to 4, y gets infinitely close to 311.
P 10 6o a formal ε-δ proof to sho' that
. 7 ) - ( lim
!
− · + −

x
x
P 17 6o a formal ε-δ proof to sho' that
. ; ) 1, ! ( lim
1
· +
− →
x
x
P 1; /hat choice of δ 'ould you ma)e (in terms of ε) to sho' that
L b mx
c x
· +

) ( lim
*
P 12 /e(ve "een ma)ing use of the fact that for real num"ers a and b, |ab| = |a|⋅ |b|. Crove that this is indeed the case.
4int< 1onsider the follo'ing cases for a and b< "oth positive? "oth negative? one positive and one negative? and at least
one of them +ero. Also, use the fact that OxO = x 'hen x ≥ , and OxO = 3x 'hen x ≤ ,.
P ! Sample problem "+< Ese the definition of a limit to argue that
. ! . !, ) 1 ( lim
7
≠ −

x
x

:olution< To do this 'e must sho' that there is some ε such that it is impossi"le for the y values to "e 'ithin ε of
!,.! for all x values 'ithin δ of 7, no matter ho' small δ is. :uppose 'e try ε = ,.-. This choice 'ill not suffice
since the interval (!,.! 3 ε, !,.! J ε) includes the real limit of !,.,. That is, for a small δ li)e ,.,1, all the x values in
the interval (7 3 δ, 7 J δ) are mapped to y values contained in (!,.! 3 ε, !,.! J ε). :o, let(s try a smaller ε, one that 'ill
not allo' any %overlap& "et'een the true limit and the claimed limit. If ε = ,.1, then no matter ho' small δ is, at least
some of the values in (7 3 δ, 7 J δ) 'ill "e not "e mapped to (!,.! 3 ε, !,.! J ε ). /e )no' this is the case since x
values very close to 7 are mapped to y values very close to !,, closer to !, than ε can %reach& from !,.!.
P !, Ese the definition of a limit to argue that
. 22 . 1- ) 4 ( lim
1!
≠ −
− →
x
x
P ! Sample problem +< Ese the definition of a limit to prove that
. 1- ) 1 4 ( lim
!
!
· −

x
x
:olution< Ket ε Q , "e given. O f (x) 3 L O = O 4x
!
3 1 3 1- O = O 4x
!
3 10 O = 4 O x J ! O ×O x 3 ! O. :ince O x 3 ! O = O x 3 a O, the
Buantity 4 O x J ! O is analogous to the slope in the linear proofs, and 'e need to get a handle on Aust ho' "ig this can "e.
If δ H 1, then our x values are in the interval (1, ). The "iggest 4 O x J ! O can "e in this interval is !, ('hen x = ). :o
'e choose δ = min(1, ε /!,), 'hich means let δ "e the smaller of 1 and ε /!, (remem"er that δ depends on ε ). @or
values of ε over !,, 'e(d go 'ith 1 for δ ? for small values of ε, 'e(d go 'ith a value of δ t'enty times smaller. >o'
'e sho' our choice for δ 'or)s< , H Ox 3 ! O H δ ⇒ Ox 3 ! O H min(1, ε /!,) ⇒ O x 3 ! O H ε /!, and O x 3 ! O H 1.
O x 3 ! O H ε /!, ⇒ !, O x 3 ! O H ε , and since O x 3 ! O H 1, 'e )no' x H and 4 O x J ! O H !,. Together these ineBualities
imply 4 O x J ! O ×O x 3 ! O H ε, from 'hich it follo' that O 4x
!
3 1 3 1- O H ε. Thus,
. 1- ) 1 4 ( lim
!
!
· −

x
x
>ote< 'or)ing in
an interval of radius 1 around x = ! 'as completely ar"itrary (the radius, that is)? the proof 'ould have "een very
similar for any choice of interval.
P !1 6o a formal ε-δ proof to sho' that
. 72 ) 4 ( lim
!
-
· +

x
x
Definition of Continuity:
.
If
), ( ) ( lim a f x f
a x
·

then 'e say f (x) is continuous at a.
/e say f (x) has a removable discontinuity at x = a if f (x) is not continuous at a "ut can "e made continuous 'ith an
appropriate definition for f (a). Draphically, they correspond to holes. *xample< f (x) = (x
!
3 x) 8 x is discontinuous at
x = ,. $y reducing the fraction, the function can "e 'ritten as f (x) = x 3 1, for x ≠ ,. Thus the graph is simply a line 'ith
a hole at (,, 31). In this case the discontinuity is remova"le since it is possi"le to %plug the hole.& /e do so adding to the
definition of f (x). 1urrently f (,) is undefined since it yields division "y +ero, "ut if 'e define f (,) to "e 31, then the
discontinuity vanishes.
/e say f (x) has a nonremovable discontinuity at x = a if f (x) is not continuous at a and cannot "e made continuous
'ith any definition for f (a). Draphically, they correspond to vertical asymptotes, Aumps, or infinite oscillations (see pg.
70). *xample< f (x) = (x

J 7) 8 (x J -) is discontinuous at x = 3-, at 'hich point exists a vertical asymptote. /hile it is
possi"le to define f (3-), no definition 'ill ma)e f (x) continuous at x = 3-.
! Sample problem +" ("y =aBuel =oney, >ayeon Fang, G Ayush 6ulguun)< Is the function
{
! H , 1 !
! , 0

) (
x x
x x
x f

≥ −
· continuous at x = ! .
:olution< /e must see if f (!) =
). ( lim
!
x f
x→
f (!) = 0(!) 3 = 2, "ut
) ( lim
!
x f
x→
6.>.*. since the one3sided limits are
not eBual. Approaching from the right, 'e use the top rule<
·
+

) ( lim
!
x f
x
!(!)

3 1 = 1-. Approaching from the left 'e
use the "ottom rule<
·


) ( lim
1
!
x f
x
0(!) 3 = 2. Thus f (x) is not continuous at x = !.
! Sample problem +"": 6etermine 'hether the function
1!
2
) (
!
!
− +

·
x x
x
x f is continuous for all values of x, and if
not, ma)e a continuous extension "y removing any remova"le discontinuities.
:olution< @irst factor<
) )( 4 (
) )( (
) (
− +
− +
·
x x
x x
x f
. /e see that the function is continuous except 'hen x = (a hole) and
'hen x = 34 (a L.A.). /e can re'rite the function as< . ), 4 8( ) ( ) ( ≠ + + · x x x x f /e no' ta)e a limit<
. 0
4

lim

− ·

+

x
x
x
>ote that it doesn(t matter that f () is undefined 'hen ta)ing this limit. Thus 'e must redefine f (x)
so that f () 'ill "e 30<
¹
'
¹
· −
≠ + +
·
, 0
), 4 8( ) (
) (
x
x x x
x f
/ith this definition 'e have removed the remova"le
discontinuity, "ut there is still a nonremova"le discontinuity (in the form of a vertical asymptote) at x = 34.
!! Esing the correspondence aR1, bR!, cR,S, find the num"er associated 'ith a"solute value of the difference of
your o'n first and last initial. 1all this num"er p. @or example, for me, 'ith initials K.$., I get p = O 1!3! O = 1,. @or
you, 'hat is the limit of f (x) = x
!
3 as x→ p . P Crove this using the definition of a limit. Crove that f (x) is
continuous at x = p. P Ese the definition of a limit to prove that the limit as x→7 of f (x) is not 14. /hat is the limit of
f (x) = x
!
3 as x→ t M . 6oes f (x) have any hori+ontal asymptotes. /hy or 'hy not. /hat a"out vertical
asymptotes.
! Is it possi"le for a function 'ith the follo'ing properties to exist. (*xplain 'hy not or give an example.) The limit as
x → M does not exist "ut the function is "ounded. The function is not periodic. ($ounded means there exists some
num"er M such that O f (x) O H M, and periodic means that f (x) has a period, T, 'here f (x J T ) = f (x) for any value of x
such that x and x + T are in the domain of f. )
!4 Cro"lem T0 on page 04.
!-. If you haven(t already, do'nload or open the po'er program that I 'rote in *xcel< 9ffice UC version 9ffice !,,,
version (Nou(ll need to "e on a machine on 'hich 5icrosoft *xcel has "een installed.) This program allo's you to
explore limits of po'er functions. /hen you open it, choose %*na"le 5acros.&
a. /hat types of values of and n 'ill generate the follo'ing types of curves< concave up, decreasing? concave up,
increasing? concave do'n, increasing? and concave do'n, decreasing. (There are t'o sets of ans'ers for some cases.)
". /hich of the a"ove types of graphs 'ill yield a %green length& greater than a %red length&. /hen are the lengths
al'ays eBual. 1ompose a clearly 'ritten, detailed paragraph explaining the follo'ing< 'hy the light "lue lines are al'ays
eBually spaced relative to L "ut not to x,? and P 'hy delta must "e chosen to "e smaller than the minimum of the red and
green lengths.
P c. /hen y = sBrt(x) and x, = !, use the program to find appropriate delta values for the follo'ing values of epsilon<
,.-, ,.4, ,., ,.!, and ,.1.
P d. :uppose you thought (incorrectly) that the limit in part c 'as 1.7. Ese the definition of the derivative to prove that
this can(t "e true. That is, use the program to find an epsilon such that no delta 'ill assure that the y3values are 'ithin
epsilon of 1.7. (Clacing the pointer 'here the "lue lines connect 'ith the axes 'ill display their x or y positions.) *xplain
your ans'er.