lesson03 the concept of limit 027 slides
DESCRIPTION
The limit is where algebra ends and caclulus begins. We describe the definition of limit as a game to find the acceptable tolerance for each error.TRANSCRIPT
. . . . . .
Section1.2–1.3A CatalogofEssentialFunctions
TheLimitofaFunction
V63.0121.027, CalculusI
September10, 2009
Announcements
I SyllabusisonthecommonBlackboardI OfficeHoursMTWR 3–4pmI ReadSections1.1–1.3ofthetextbookthisweek.
. . . . . .
Outline
ClassesofFunctionsLinearfunctionsQuadraticfunctionsCubicfunctionsOtherpowerfunctionsRationalfunctionsTrigonometricFunctionsExponentialandLogarithmicfunctions
TransformationsofFunctions
CompositionsofFunctions
LimitsHeuristicsErrorsandtolerancesExamplesPathologies
. . . . . .
ClassesofFunctions
I linearfunctions, definedbyslopeanintercept, pointandpoint, orpointandslope.
I quadraticfunctions, cubicfunctions, powerfunctions,polynomials
I rationalfunctionsI trigonometricfunctionsI exponential/logarithmicfunctions
. . . . . .
Linearfunctions
Linearfunctionshaveaconstantrateofgrowthandareoftheform
f(x) = mx + b.
ExampleInNewYorkCitytaxiscost$2.50togetinand$0.40per 1/5mile. Writethefare f(x) asafunctionofdistance x traveled.
AnswerIf x isinmilesand f(x) indollars,
f(x) = 2.5 + 2x
. . . . . .
Linearfunctions
Linearfunctionshaveaconstantrateofgrowthandareoftheform
f(x) = mx + b.
ExampleInNewYorkCitytaxiscost$2.50togetinand$0.40per 1/5mile. Writethefare f(x) asafunctionofdistance x traveled.
AnswerIf x isinmilesand f(x) indollars,
f(x) = 2.5 + 2x
. . . . . .
Linearfunctions
Linearfunctionshaveaconstantrateofgrowthandareoftheform
f(x) = mx + b.
ExampleInNewYorkCitytaxiscost$2.50togetinand$0.40per 1/5mile. Writethefare f(x) asafunctionofdistance x traveled.
AnswerIf x isinmilesand f(x) indollars,
f(x) = 2.5 + 2x
. . . . . .
Quadraticfunctions
Thesetaketheform
f(x) = ax2 + bx + c
Thegraphisaparabolawhichopensupwardif a > 0, downwardif a < 0.
. . . . . .
Quadraticfunctions
Thesetaketheform
f(x) = ax2 + bx + c
Thegraphisaparabolawhichopensupwardif a > 0, downwardif a < 0.
. . . . . .
Otherpowerfunctions
I Wholenumberpowers: f(x) = xn.
I negativepowersarereciprocals: x−3 =1x3.
I fractionalpowersareroots: x1/3 = 3√x.
. . . . . .
Rationalfunctions
DefinitionA rationalfunction isaquotientofpolynomials.
Example
Thefunction f(x) =x3(x + 3)
(x + 2)(x− 1)isrational.
. . . . . .
ExponentialandLogarithmicfunctions
I exponentialfunctions(forexample f(x) = 2x)I logarithmicfunctionsaretheirinverses(forexample
f(x) = log2(x))
. . . . . .
Outline
ClassesofFunctionsLinearfunctionsQuadraticfunctionsCubicfunctionsOtherpowerfunctionsRationalfunctionsTrigonometricFunctionsExponentialandLogarithmicfunctions
TransformationsofFunctions
CompositionsofFunctions
LimitsHeuristicsErrorsandtolerancesExamplesPathologies
. . . . . .
TransformationsofFunctions
Takethe sinefunction andgraphthesetransformations:
I sin(x +
π
2
)I sin
(x− π
2
)I sin (x) +
π
2I sin (x) − π
2
Observethatifthefiddlingoccurswithinthefunction, atransformationisappliedonthe x-axis. Afterthefunction, tothey-axis.
. . . . . .
TransformationsofFunctions
Takethe sinefunction andgraphthesetransformations:
I sin(x +
π
2
)I sin
(x− π
2
)I sin (x) +
π
2I sin (x) − π
2Observethatifthefiddlingoccurswithinthefunction, atransformationisappliedonthe x-axis. Afterthefunction, tothey-axis.
. . . . . .
VerticalandHorizontalShifts
Suppose c > 0. ToobtainthegraphofI y = f(x) + c, shiftthegraphof y = f(x) adistance c units
upward
I y = f(x) − c, shiftthegraphof y = f(x) adistance c units
downward
I y = f(x− c), shiftthegraphof y = f(x) adistance c units
totheright
I y = f(x + c), shiftthegraphof y = f(x) adistance c units
totheleft
. . . . . .
VerticalandHorizontalShifts
Suppose c > 0. ToobtainthegraphofI y = f(x) + c, shiftthegraphof y = f(x) adistance c unitsupward
I y = f(x) − c, shiftthegraphof y = f(x) adistance c units
downward
I y = f(x− c), shiftthegraphof y = f(x) adistance c units
totheright
I y = f(x + c), shiftthegraphof y = f(x) adistance c units
totheleft
. . . . . .
VerticalandHorizontalShifts
Suppose c > 0. ToobtainthegraphofI y = f(x) + c, shiftthegraphof y = f(x) adistance c unitsupward
I y = f(x) − c, shiftthegraphof y = f(x) adistance c unitsdownward
I y = f(x− c), shiftthegraphof y = f(x) adistance c units
totheright
I y = f(x + c), shiftthegraphof y = f(x) adistance c units
totheleft
. . . . . .
VerticalandHorizontalShifts
Suppose c > 0. ToobtainthegraphofI y = f(x) + c, shiftthegraphof y = f(x) adistance c unitsupward
I y = f(x) − c, shiftthegraphof y = f(x) adistance c unitsdownward
I y = f(x− c), shiftthegraphof y = f(x) adistance c units totheright
I y = f(x + c), shiftthegraphof y = f(x) adistance c units
totheleft
. . . . . .
VerticalandHorizontalShifts
Suppose c > 0. ToobtainthegraphofI y = f(x) + c, shiftthegraphof y = f(x) adistance c unitsupward
I y = f(x) − c, shiftthegraphof y = f(x) adistance c unitsdownward
I y = f(x− c), shiftthegraphof y = f(x) adistance c units totheright
I y = f(x + c), shiftthegraphof y = f(x) adistance c units totheleft
. . . . . .
Outline
ClassesofFunctionsLinearfunctionsQuadraticfunctionsCubicfunctionsOtherpowerfunctionsRationalfunctionsTrigonometricFunctionsExponentialandLogarithmicfunctions
TransformationsofFunctions
CompositionsofFunctions
LimitsHeuristicsErrorsandtolerancesExamplesPathologies
. . . . . .
Composing
ExampleLet f(x) = x2 and g(x) = sin x. Compute f ◦ g and g ◦ f.
Solutionf ◦ g(x) = sin2 x while g ◦ f(x) = sin(x2). Notetheyare not thesame.
. . . . . .
Composing
ExampleLet f(x) = x2 and g(x) = sin x. Compute f ◦ g and g ◦ f.
Solutionf ◦ g(x) = sin2 x while g ◦ f(x) = sin(x2). Notetheyare not thesame.
. . . . . .
Decomposing
ExampleExpress
√x2 − 4 asacompositionoftwofunctions. Whatisits
domain?
SolutionWecanwritetheexpressionas f ◦ g, where f(u) =
√u and
g(x) = x2 − 4. Therangeof g needstobewithinthedomainof f.Toinsurethat x2 − 4 ≥ 0, wemusthave x ≤ −2 or x ≥ 2.
. . . . . .
Outline
ClassesofFunctionsLinearfunctionsQuadraticfunctionsCubicfunctionsOtherpowerfunctionsRationalfunctionsTrigonometricFunctionsExponentialandLogarithmicfunctions
TransformationsofFunctions
CompositionsofFunctions
LimitsHeuristicsErrorsandtolerancesExamplesPathologies
. . . . . .
Zeno’sParadox
Thatwhichisinlocomotionmustarriveatthehalf-waystagebeforeitarrivesatthegoal.
(Aristotle Physics VI:9,239b10)
. . . . . .
HeuristicDefinitionofaLimit
DefinitionWewrite
limx→a
f(x) = L
andsay
“thelimitof f(x), as x approaches a, equals L”
ifwecanmakethevaluesof f(x) arbitrarilycloseto L (asclosetoL aswelike)bytaking x tobesufficientlycloseto a (oneithersideof a)butnotequalto a.
. . . . . .
Theerror-tolerancegame
A gamebetweentwoplayerstodecideifalimit limx→a
f(x) exists.
I Player1: Choose L tobethelimit.I Player2: Proposean“error”levelaround L.I Player1: Choosea“tolerance”levelaround a sothat
x-pointswithinthattolerancelevelaretakento y-valueswithintheerrorlevel.
IfPlayer1canalwayswin, limx→a
f(x) = L.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig
.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig
.Stilltoobig
.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig
.Thislooksgood
.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood
.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
Theerror-tolerancegame
.
.Thistoleranceistoobig.Stilltoobig.Thislooksgood.Sodoesthis
.a
.L
I Tobelegit, thepartofthegraphinsidetheblue(vertical)stripmustalsobeinsidethegreen(horizontal)strip.
I IfPlayer2shrinkstheerror, Player1canstillwin.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.I Iftheerrorlevelis 0.01, I needtoguaranteethat
−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.
I Iftheerrorlevelis 0.01, I needtoguaranteethat−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.
I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.I Iftheerrorlevelis 0.01, I needtoguaranteethat
−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.
I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.I Iftheerrorlevelis 0.01, I needtoguaranteethat
−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.
I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.I Iftheerrorlevelis 0.01, I needtoguaranteethat
−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
ExampleFind lim
x→0x2 ifitexists.
Solution
I I claimthelimitiszero.I Iftheerrorlevelis 0.01, I needtoguaranteethat
−0.01 < x2 < 0.01 forall x sufficientlyclosetozero.I If −0.1 < x < 0.1, then 0 ≤ x2 ≤ 0.01, soI winthatround.I Whatshouldthetolerancebeiftheerroris 0.0001?
Bysettingtoleranceequaltothesquarerootoftheerror, wecanguaranteetobewithinanyerror.
. . . . . .
Example
Find limx→0
|x|x
ifitexists.
Solution
Thefunctioncanalsobewrittenas
|x|x
=
{1 if x > 0;
−1 if x < 0
Whatwouldbethelimit?Theerror-tolerancegamefails, but
limx→0+
f(x) = 1 limx→0−
f(x) = −1
. . . . . .
Example
Find limx→0
|x|x
ifitexists.
SolutionThefunctioncanalsobewrittenas
|x|x
=
{1 if x > 0;
−1 if x < 0
Whatwouldbethelimit?
Theerror-tolerancegamefails, but
limx→0+
f(x) = 1 limx→0−
f(x) = −1
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
Theerror-tolerancegame
. .x
.y
..−1
..1 .
.
.Part of graph in-side blue is notinside green
.Part of graph in-side blue is notinside green
I Thesearetheonlygoodchoices; thelimitdoesnotexist.
. . . . . .
One-sidedlimits
DefinitionWewrite
limx→a+
f(x) = L
andsay
“thelimitof f(x), as x approaches a fromthe right, equals L”
ifwecanmakethevaluesof f(x) arbitrarilycloseto L (asclosetoL aswelike)bytaking x tobesufficientlycloseto a (oneithersideof a)and greater than a.
. . . . . .
One-sidedlimits
DefinitionWewrite
limx→a−
f(x) = L
andsay
“thelimitof f(x), as x approaches a fromthe left, equals L”
ifwecanmakethevaluesof f(x) arbitrarilycloseto L (asclosetoL aswelike)bytaking x tobesufficientlycloseto a (oneithersideof a)and less than a.
. . . . . .
Example
Find limx→0
|x|x
ifitexists.
SolutionThefunctioncanalsobewrittenas
|x|x
=
{1 if x > 0;
−1 if x < 0
Whatwouldbethelimit?Theerror-tolerancegamefails, but
limx→0+
f(x) = 1 limx→0−
f(x) = −1
. . . . . .
Example
Find limx→0+
1xifitexists.
SolutionThelimitdoesnotexistbecausethefunctionisunboundednear0. Nextweekwewillunderstandthestatementthat
limx→0+
1x
= +∞
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good
.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good
.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Theerror-tolerancegame
. .x
.y
.0
..L?
.The graph escapes thegreen, so no good.Evenworse!
.The limit does not existbecause the function isunbounded near 0
. . . . . .
Example
Find limx→0+
1xifitexists.
SolutionThelimitdoesnotexistbecausethefunctionisunboundednear0. Nextweekwewillunderstandthestatementthat
limx→0+
1x
= +∞
. . . . . .
Weird, wildstuff
ExampleFind lim
x→0sin
(π
x
)ifitexists.
I f(x) = 0 when x =
1kforanyinteger k
I f(x) = 1 when x =
12k + 1/2
foranyinteger k
I f(x) = −1 when x =
12k− 1/2
foranyinteger k
. . . . . .
Weird, wildstuff
ExampleFind lim
x→0sin
(π
x
)ifitexists.
I f(x) = 0 when x =
1kforanyinteger k
I f(x) = 1 when x =
12k + 1/2
foranyinteger k
I f(x) = −1 when x =
12k− 1/2
foranyinteger k
. . . . . .
Weird, wildstuff
ExampleFind lim
x→0sin
(π
x
)ifitexists.
I f(x) = 0 when x =1kforanyinteger k
I f(x) = 1 when x =
12k + 1/2
foranyinteger k
I f(x) = −1 when x =
12k− 1/2
foranyinteger k
. . . . . .
Weird, wildstuff
ExampleFind lim
x→0sin
(π
x
)ifitexists.
I f(x) = 0 when x =1kforanyinteger k
I f(x) = 1 when x =1
2k + 1/2foranyinteger k
I f(x) = −1 when x =
12k− 1/2
foranyinteger k
. . . . . .
Weird, wildstuff
ExampleFind lim
x→0sin
(π
x
)ifitexists.
I f(x) = 0 when x =1kforanyinteger k
I f(x) = 1 when x =1
2k + 1/2foranyinteger k
I f(x) = −1 when x =1
2k− 1/2foranyinteger k
. . . . . .
Weird, wildstuffcontinued
Hereisagraphofthefunction:
. .x
.y
..−1
..1
Thereareinfinitelymanypointsarbitrarilyclosetozerowheref(x) is 0, or 1, or −1. Sothelimitcannotexist.
. . . . . .
Whatcouldgowrong?SummaryofLimitPathologies
Howcouldafunctionfailtohavealimit? Somepossibilities:I left-andright-handlimitsexistbutarenotequalI Thefunctionisunboundednear aI Oscillationwithincreasinglyhighfrequencynear a
. . . . . .
MeettheMathematician: AugustinLouisCauchy
I French, 1789–1857I RoyalistandCatholicI madecontributionsingeometry, calculus,complexanalysis,numbertheory
I createdthedefinitionoflimitweusetodaybutdidn’tunderstandit
. . . . . .
PreciseDefinitionofaLimitNo, thisisnotgoingtobeonthetest
Let f beafunctiondefinedonansomeopenintervalthatcontainsthenumber a, exceptpossiblyat a itself. Thenwesaythatthe limitof f(x) as x approaches a is L, andwewrite
limx→a
f(x) = L,
ifforevery ε > 0 thereisacorresponding δ > 0 suchthat
if 0 < |x− a| < δ, then |f(x) − L| < ε.
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L
. . . . . .
Theerror-tolerancegame= ε, δ
.
.L + ε
.L− ε
.a− δ .a + δ
.This δ istoobig
.a− δ .a + δ
.This δ looksgood
.a− δ .a + δ
.Sodoesthis δ
.a
.L