lesson 7: the derivative (slides)
DESCRIPTION
Many functions in nature are described as the rate of change of another function. The concept is called the derivative. Algebraically, the process of finding the derivative involves a limit of difference quotients.TRANSCRIPT
.... NYUMathematics
Sec on 2.1–2.2The Deriva ve
V63.0121.011: Calculus IProfessor Ma hew Leingang
New York University
February 14, 2011
Announcements
I Quiz this week onSec ons 1.1–1.4
I No class Monday,February 21
ObjectivesThe Derivative
I Understand and state the defini on ofthe deriva ve of a func on at a point.
I Given a func on and a point in itsdomain, decide if the func on isdifferen able at the point and find thevalue of the deriva ve at that point.
I Understand and give several examplesof deriva ves modeling rates of changein science.
ObjectivesThe Derivative as a Function
I Given a func on f, use the defini on ofthe deriva ve to find the deriva vefunc on f’.
I Given a func on, find its secondderiva ve.
I Given the graph of a func on, sketchthe graph of its deriva ve.
OutlineRates of Change
Tangent LinesVelocityPopula on growthMarginal costs
The deriva ve, definedDeriva ves of (some) power func onsWhat does f tell you about f′?
How can a func on fail to be differen able?Other nota onsThe second deriva ve
The tangent problemA geometric rate of change
ProblemGiven a curve and a point on the curve, find the slope of the linetangent to the curve at that point.
Solu onIf the curve is given by y = f(x), and the point on the curve is(a, f(a)), then the slope of the tangent line is given by
mtangent = limx→a
f(x)− f(a)x− a
A tangent problem
Example
Find the slope of the line tangent to the curve y = x2 at the point(2, 4).
Graphically and numerically
.. x.
y
..2
..
4
.
x m =x2 − 22
x− 2
3 52.5 4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....3
..
9
x m =x2 − 22
x− 23
52.5 4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....3
..
9
x m =x2 − 22
x− 23 5
2.5 4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.5
..
6.25
x m =x2 − 22
x− 23 52.5
4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.5
..
6.25
x m =x2 − 22
x− 23 52.5 4.5
2.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.1
..
4.41
x m =x2 − 22
x− 23 52.5 4.52.1
4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.1
..
4.41
x m =x2 − 22
x− 23 52.5 4.52.1 4.1
2.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.01
..
4.0401
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01
4.01limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....2.01
..
4.0401
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....1
..1
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.991.9 3.91.5 3.5
1
3
Graphically and numerically
.. x.
y
..2
..
4
....1
..1
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.991.9 3.91.5 3.5
1 3
Graphically and numerically
.. x.
y
..2
..
4
....1.5
..
2.25
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.991.9 3.9
1.5
3.5
1 3
Graphically and numerically
.. x.
y
..2
..
4
....1.5
..
2.25
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.991.9 3.9
1.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....1.9
..
3.61
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.99
1.9
3.9
1.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....1.9
..
3.61
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 41.99 3.99
1.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....1.99
..
3.9601
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 4
1.99
3.99
1.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....1.99
..
3.9601
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 4
1.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
.
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01
limit 4
1.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....3
..
9
...2.5
..
6.25
...2.1
..
4.41
...2.01
..
4.0401
...1
..1 ...1.5
..
2.25
...1.9
..
3.61
...1.99
..
3.9601
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01limit
4
1.99 3.991.9 3.91.5 3.51 3
Graphically and numerically
.. x.
y
..2
..
4
....3
..
9
...2.5
..
6.25
...2.1
..
4.41
...2.01
..
4.0401
...1
..1 ...1.5
..
2.25
...1.9
..
3.61
...1.99
..
3.9601
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
The tangent problemA geometric rate of change
ProblemGiven a curve and a point on the curve, find the slope of the linetangent to the curve at that point.
Solu onIf the curve is given by y = f(x), and the point on the curve is(a, f(a)), then the slope of the tangent line is given by
mtangent = limx→a
f(x)− f(a)x− a
The velocity problemKinematics—Physical rates of change
ProblemGiven the posi on func on of a moving object, find the velocity ofthe object at a certain instant in me.
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 15
1.5 − 12.51.1 − 10.51.01 − 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5
− 12.51.1 − 10.51.01 − 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.5
1.1 − 10.51.01 − 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1
− 10.51.01 − 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1 − 10.5
1.01 − 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1 − 10.51.01
− 10.051.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1 − 10.51.01 − 10.05
1.001 − 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1 − 10.51.01 − 10.051.001
− 10.005
Numerical evidenceh(t) = 50− 5t2
Fill in the table:
t vave =h(t)− h(1)
t− 12 − 151.5 − 12.51.1 − 10.51.01 − 10.051.001 − 10.005
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
A velocity problemExample
Drop a ball off the roof of theSilver Center so that its height canbe described by
h(t) = 50− 5t2
where t is seconds a er droppingit and h is meters above theground. How fast is it falling onesecond a er we drop it?
Solu onThe answer is
v = limt→1
(50− 5t2)− 45t− 1
= limt→1
5− 5t2
t− 1
= limt→1
5(1− t)(1+ t)t− 1
= (−5) limt→1
(1+ t)
= −5 · 2 = −10
Velocity in generalUpshot
If the height func on is givenby h(t), the instantaneousvelocity at me t0 is given by
v = limt→t0
h(t)− h(t0)t− t0
= lim∆t→0
h(t0 +∆t)− h(t0)∆t ... t..
y = h(t)
....t0
..t
..
h(t0)
..
h(t0 +∆t)
. ∆t.
∆h
Population growthBiological Rates of Change
ProblemGiven the popula on func on of a group of organisms, find the rateof growth of the popula on at a par cular instant.
Solu onThe instantaneous popula on growth is given by
lim∆t→0
P(t+∆t)− P(t)∆t
Population growth exampleExample
Suppose the popula on of fish in the East River is given by thefunc on
P(t) =3et
1+ etwhere t is in years since 2000 and P is in millions of fish. Is the fishpopula on growing fastest in 1990, 2000, or 2010? (Es matenumerically)
AnswerWe es mate the rates of growth to be 0.000143229, 0.749376, and0.0001296. So the popula on is growing fastest in 2000.
DerivationSolu onLet∆t be an increment in me and∆P the corresponding change inpopula on:
∆P = P(t+∆t)− P(t)
This depends on∆t, so ideally we would want
lim∆t→0
∆P∆t
= lim∆t→0
1∆t
(3et+∆t
1+ et+∆t −3et
1+ et
)
But rather than compute a complicated limit analy cally, let usapproximate numerically. We will try a small∆t, for instance 0.1.
DerivationSolu onLet∆t be an increment in me and∆P the corresponding change inpopula on:
∆P = P(t+∆t)− P(t)
This depends on∆t, so ideally we would want
lim∆t→0
∆P∆t
= lim∆t→0
1∆t
(3et+∆t
1+ et+∆t −3et
1+ et
)But rather than compute a complicated limit analy cally, let usapproximate numerically. We will try a small∆t, for instance 0.1.
Numerical evidenceSolu on (Con nued)
To approximate the popula on change in year n, use the difference
quo entP(t+∆t)− P(t)
∆t, where∆t = 0.1 and t = n− 2000.
r1990
≈ P(−10+ 0.1)− P(−10)0.1
=10.1
(3e−9.9
1+ e−9.9 −3e−10
1+ e−10
)= 0.000143229
r2000
≈ P(0.1)− P(0)0.1
=10.1
(3e0.1
1+ e0.1− 3e0
1+ e0
)= 0.749376
Numerical evidenceSolu on (Con nued)
To approximate the popula on change in year n, use the difference
quo entP(t+∆t)− P(t)
∆t, where∆t = 0.1 and t = n− 2000.
r1990 ≈P(−10+ 0.1)− P(−10)
0.1
=10.1
(3e−9.9
1+ e−9.9 −3e−10
1+ e−10
)= 0.000143229
r2000 ≈P(0.1)− P(0)
0.1
=10.1
(3e0.1
1+ e0.1− 3e0
1+ e0
)= 0.749376
Numerical evidenceSolu on (Con nued)
To approximate the popula on change in year n, use the difference
quo entP(t+∆t)− P(t)
∆t, where∆t = 0.1 and t = n− 2000.
r1990 ≈P(−10+ 0.1)− P(−10)
0.1=
10.1
(3e−9.9
1+ e−9.9 −3e−10
1+ e−10
)
= 0.000143229
r2000 ≈P(0.1)− P(0)
0.1=
10.1
(3e0.1
1+ e0.1− 3e0
1+ e0
)
= 0.749376
Numerical evidenceSolu on (Con nued)
To approximate the popula on change in year n, use the difference
quo entP(t+∆t)− P(t)
∆t, where∆t = 0.1 and t = n− 2000.
r1990 ≈P(−10+ 0.1)− P(−10)
0.1=
10.1
(3e−9.9
1+ e−9.9 −3e−10
1+ e−10
)= 0.000143229
r2000 ≈P(0.1)− P(0)
0.1=
10.1
(3e0.1
1+ e0.1− 3e0
1+ e0
)
= 0.749376
Numerical evidenceSolu on (Con nued)
To approximate the popula on change in year n, use the difference
quo entP(t+∆t)− P(t)
∆t, where∆t = 0.1 and t = n− 2000.
r1990 ≈P(−10+ 0.1)− P(−10)
0.1=
10.1
(3e−9.9
1+ e−9.9 −3e−10
1+ e−10
)= 0.000143229
r2000 ≈P(0.1)− P(0)
0.1=
10.1
(3e0.1
1+ e0.1− 3e0
1+ e0
)= 0.749376
Solu on (Con nued)
r2010
≈ P(10+ 0.1)− P(10)0.1
=10.1
(3e10.1
1+ e10.1− 3e10
1+ e10
)= 0.0001296
Solu on (Con nued)
r2010 ≈P(10+ 0.1)− P(10)
0.1
=10.1
(3e10.1
1+ e10.1− 3e10
1+ e10
)= 0.0001296
Solu on (Con nued)
r2010 ≈P(10+ 0.1)− P(10)
0.1=
10.1
(3e10.1
1+ e10.1− 3e10
1+ e10
)
= 0.0001296
Solu on (Con nued)
r2010 ≈P(10+ 0.1)− P(10)
0.1=
10.1
(3e10.1
1+ e10.1− 3e10
1+ e10
)= 0.0001296
Population growth exampleExample
Suppose the popula on of fish in the East River is given by thefunc on
P(t) =3et
1+ etwhere t is in years since 2000 and P is in millions of fish. Is the fishpopula on growing fastest in 1990, 2000, or 2010? (Es matenumerically)
AnswerWe es mate the rates of growth to be 0.000143229, 0.749376, and0.0001296. So the popula on is growing fastest in 2000.
Population growthBiological Rates of Change
ProblemGiven the popula on func on of a group of organisms, find the rateof growth of the popula on at a par cular instant.
Solu onThe instantaneous popula on growth is given by
lim∆t→0
P(t+∆t)− P(t)∆t
Marginal costsRates of change in economics
ProblemGiven the produc on cost of a good, find the marginal cost ofproduc on a er having produced a certain quan ty.
Solu onThe marginal cost a er producing q is given by
MC = lim∆q→0
C(q+∆q)− C(q)∆q
Marginal Cost ExampleExample
Suppose the cost of producing q tons of rice on our paddy in a year is
C(q) = q3 − 12q2 + 60q
We are currently producing 5 tons a year. Should we change that?
AnswerIf q = 5, then C = 125,∆C = 19, while AC = 25. So we shouldproduce more to lower average costs.
Marginal Cost ExampleExample
Suppose the cost of producing q tons of rice on our paddy in a year is
C(q) = q3 − 12q2 + 60q
We are currently producing 5 tons a year. Should we change that?
AnswerIf q = 5, then C = 125,∆C = 19, while AC = 25. So we shouldproduce more to lower average costs.
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q)
AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)
4
112 28 13
5
125 25 19
6
144 24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q)
AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)
4 112
28 13
5
125 25 19
6
144 24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q)
AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)
4 112
28 13
5 125
25 19
6
144 24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q)
AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)
4 112
28 13
5 125
25 19
6 144
24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q
∆C = C(q+ 1)− C(q)
4 112
28 13
5 125
25 19
6 144
24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q
∆C = C(q+ 1)− C(q)
4 112 28
13
5 125
25 19
6 144
24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q
∆C = C(q+ 1)− C(q)
4 112 28
13
5 125 25
19
6 144
24 31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q
∆C = C(q+ 1)− C(q)
4 112 28
13
5 125 25
19
6 144 24
31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)4 112 28
13
5 125 25
19
6 144 24
31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)4 112 28 135 125 25
19
6 144 24
31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)4 112 28 135 125 25 196 144 24
31
ComparisonsSolu on
C(q) = q3 − 12q2 + 60q
Fill in the table:
q C(q) AC(q) = C(q)/q ∆C = C(q+ 1)− C(q)4 112 28 135 125 25 196 144 24 31
Marginal Cost ExampleExample
Suppose the cost of producing q tons of rice on our paddy in a year is
C(q) = q3 − 12q2 + 60q
We are currently producing 5 tons a year. Should we change that?
AnswerIf q = 5, then C = 125,∆C = 19, while AC = 25. So we shouldproduce more to lower average costs.
Marginal costsRates of change in economics
ProblemGiven the produc on cost of a good, find the marginal cost ofproduc on a er having produced a certain quan ty.
Solu onThe marginal cost a er producing q is given by
MC = lim∆q→0
C(q+∆q)− C(q)∆q
OutlineRates of Change
Tangent LinesVelocityPopula on growthMarginal costs
The deriva ve, definedDeriva ves of (some) power func onsWhat does f tell you about f′?
How can a func on fail to be differen able?Other nota onsThe second deriva ve
The definitionAll of these rates of change are found the same way!
Defini onLet f be a func on and a a point in the domain of f. If the limit
f′(a) = limh→0
f(a+ h)− f(a)h
= limx→a
f(x)− f(a)x− a
exists, the func on is said to be differen able at a and f′(a) is thederiva ve of f at a.
The definitionAll of these rates of change are found the same way!
Defini onLet f be a func on and a a point in the domain of f. If the limit
f′(a) = limh→0
f(a+ h)− f(a)h
= limx→a
f(x)− f(a)x− a
exists, the func on is said to be differen able at a and f′(a) is thederiva ve of f at a.
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h= lim
h→0(2a+ h) = 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h= lim
h→0(2a+ h) = 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h= lim
h→0(2a+ h) = 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h
= limh→0
2ah+ h2
h= lim
h→0(2a+ h) = 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h
= limh→0
(2a+ h) = 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h= lim
h→0(2a+ h)
= 2a
Derivative of the squaring functionExample
Suppose f(x) = x2. Use the defini on of deriva ve to find f′(a).
Solu on
f′(a) = limh→0
f(a+ h)− f(a)h
= limh→0
(a+ h)2 − a2
h
= limh→0
(a2 + 2ah+ h2)− a2
h= lim
h→0
2ah+ h2
h= lim
h→0(2a+ h) = 2a
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
“Can you do it the other way?”Same limit, different form
Solu on
f′(2) = limh→0
f(2+ h)− f(2)h
= limh→0
12+h −
12
h
= limh→0
2− (2+ h)2h(2+ h)
= limh→0
−h2h(2+ h)
= limh→0
−12(2+ h)
= −14
“Can you do it the other way?”Same limit, different form
Solu on
f′(2) = limh→0
f(2+ h)− f(2)h
= limh→0
12+h −
12
h
= limh→0
2− (2+ h)2h(2+ h)
= limh→0
−h2h(2+ h)
= limh→0
−12(2+ h)
= −14
“Can you do it the other way?”Same limit, different form
Solu on
f′(2) = limh→0
f(2+ h)− f(2)h
= limh→0
12+h −
12
h
= limh→0
2− (2+ h)2h(2+ h)
= limh→0
−h2h(2+ h)
= limh→0
−12(2+ h)
= −14
“Can you do it the other way?”Same limit, different form
Solu on
f′(2) = limh→0
f(2+ h)− f(2)h
= limh→0
12+h −
12
h
= limh→0
2− (2+ h)2h(2+ h)
= limh→0
−h2h(2+ h)
= limh→0
−12(2+ h)
= −14
“Can you do it the other way?”Same limit, different form
Solu on
f′(2) = limh→0
f(2+ h)− f(2)h
= limh→0
12+h −
12
h
= limh→0
2− (2+ h)2h(2+ h)
= limh→0
−h2h(2+ h)
= limh→0
−12(2+ h)
= −14
“How did you get that?”The Sure-Fire Sally Rule (SFSR) for adding fractions
Fact
ab± c
d=
ad± bcbd
1x− 1
2x− 2
=
2− x2x
x− 2=
2− x2x(x− 2)
Paul Sally
“How did you get that?”The Sure-Fire Sally Rule (SFSR) for adding fractions
Fact
ab± c
d=
ad± bcbd
1x− 1
2x− 2
=
2− x2x
x− 2=
2− x2x(x− 2) Paul Sally
What does f tell you about f′?
I If f is a func on, we can compute the deriva ve f′(x) at eachpoint x where f is differen able, and come up with anotherfunc on, the deriva ve func on.
I What can we say about this func on f′?
I If f is decreasing on an interval, f′ is nega ve (technically, nonposi ve)on that interval
I If f is increasing on an interval, f′ is posi ve (technically, nonnega ve)on that interval
What does f tell you about f′?
I If f is a func on, we can compute the deriva ve f′(x) at eachpoint x where f is differen able, and come up with anotherfunc on, the deriva ve func on.
I What can we say about this func on f′?I If f is decreasing on an interval, f′ is nega ve (technically, nonposi ve)
on that interval
I If f is increasing on an interval, f′ is posi ve (technically, nonnega ve)on that interval
Derivative of the reciprocalExample
Suppose f(x) =1x. Use the defini on of the deriva ve to find f′(2).
Solu on
f′(2) = limx→2
1/x− 1/2x− 2
= limx→2
2− x2x(x− 2)
= limx→2
−12x
= −14
..x
.
y
.
What does f tell you about f′?
I If f is a func on, we can compute the deriva ve f′(x) at eachpoint x where f is differen able, and come up with anotherfunc on, the deriva ve func on.
I What can we say about this func on f′?I If f is decreasing on an interval, f′ is nega ve (technically, nonposi ve)
on that intervalI If f is increasing on an interval, f′ is posi ve (technically, nonnega ve)
on that interval
Graphically and numerically
.. x.
y
..2
..
4
....3
..
9
...2.5
..
6.25
...2.1
..
4.41
...2.01
..
4.0401
...1
..1 ...1.5
..
2.25
...1.9
..
3.61
...1.99
..
3.9601
x m =x2 − 22
x− 23 52.5 4.52.1 4.12.01 4.01limit 41.99 3.991.9 3.91.5 3.51 3
What does f tell you about f′?FactIf f is decreasing on the open interval (a, b), then f′ ≤ 0 on (a, b).
Picture Proof.
If f is decreasing, then all secant linespoint downward, hence havenega ve slope. The deriva ve is alimit of slopes of secant lines, whichare all nega ve, so the limit must be≤ 0. ..
x.
y
......
What does f tell you about f′?FactIf f is decreasing on the open interval (a, b), then f′ ≤ 0 on (a, b).
Picture Proof.
If f is decreasing, then all secant linespoint downward, hence havenega ve slope. The deriva ve is alimit of slopes of secant lines, whichare all nega ve, so the limit must be≤ 0. ..
x.
y
......
What does f tell you about f′?FactIf f is decreasing on on the open interval (a, b), then f′ ≤ 0 on (a, b).
The Real Proof.
I If∆x > 0, then
f(x+∆x) < f(x) =⇒ f(x+∆x)− f(x)∆x
< 0
What does f tell you about f′?FactIf f is decreasing on on the open interval (a, b), then f′ ≤ 0 on (a, b).
The Real Proof.
I If∆x > 0, then
f(x+∆x) < f(x) =⇒ f(x+∆x)− f(x)∆x
< 0
I If∆x < 0, then x+∆x < x, and
f(x+∆x) > f(x) =⇒ f(x+∆x)− f(x)∆x
< 0
What does f tell you about f′?FactIf f is decreasing on on the open interval (a, b), then f′ ≤ 0 on (a, b).
The Real Proof.
I Either way,f(x+∆x)− f(x)
∆x< 0, so
f′(x) = lim∆x→0
f(x+∆x)− f(x)∆x
≤ 0
Going the Other Way?
Ques on
If a func on has a nega ve deriva ve on an interval, must it bedecreasing on that interval?
AnswerMaybe.
Going the Other Way?
Ques on
If a func on has a nega ve deriva ve on an interval, must it bedecreasing on that interval?
AnswerMaybe.
OutlineRates of Change
Tangent LinesVelocityPopula on growthMarginal costs
The deriva ve, definedDeriva ves of (some) power func onsWhat does f tell you about f′?
How can a func on fail to be differen able?Other nota onsThe second deriva ve
Differentiability is super-continuityTheoremIf f is differen able at a, then f is con nuous at a.
Proof.We have
limx→a
(f(x)− f(a)) = limx→a
f(x)− f(a)x− a
· (x− a)
= limx→a
f(x)− f(a)x− a
· limx→a
(x− a)
= f′(a) · 0 = 0
Differentiability is super-continuityTheoremIf f is differen able at a, then f is con nuous at a.
Proof.We have
limx→a
(f(x)− f(a)) = limx→a
f(x)− f(a)x− a
· (x− a)
= limx→a
f(x)− f(a)x− a
· limx→a
(x− a)
= f′(a) · 0 = 0
Differentiability is super-continuityTheoremIf f is differen able at a, then f is con nuous at a.
Proof.We have
limx→a
(f(x)− f(a)) = limx→a
f(x)− f(a)x− a
· (x− a)
= limx→a
f(x)− f(a)x− a
· limx→a
(x− a)
= f′(a) · 0
= 0
Differentiability is super-continuityTheoremIf f is differen able at a, then f is con nuous at a.
Proof.We have
limx→a
(f(x)− f(a)) = limx→a
f(x)− f(a)x− a
· (x− a)
= limx→a
f(x)− f(a)x− a
· limx→a
(x− a)
= f′(a) · 0 = 0
Differentiability FAILKinks
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
.
.
Differentiability FAILKinks
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
.
.
Differentiability FAILKinks
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
..
Differentiability FAILCusps
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILCusps
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILCusps
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILCusps
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILVertical Tangents
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILVertical Tangents
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILVertical Tangents
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILVertical Tangents
Example
Let f have the graph on the le -hand side below. Sketch the graph ofthe deriva ve f′.
.. x.
f(x)
.. x.
f′(x)
Differentiability FAILWeird, Wild, Stuff
Example
.. x.
f(x)
This func on is differen ableat 0.
.. x.
f′(x)
.
But the deriva ve is notcon nuous at 0!
Differentiability FAILWeird, Wild, Stuff
Example
.. x.
f(x)
This func on is differen ableat 0.
.. x.
f′(x)
.
But the deriva ve is notcon nuous at 0!
Differentiability FAILWeird, Wild, Stuff
Example
.. x.
f(x)
This func on is differen ableat 0.
.. x.
f′(x)
.
But the deriva ve is notcon nuous at 0!
Differentiability FAILWeird, Wild, Stuff
Example
.. x.
f(x)
This func on is differen ableat 0.
.. x.
f′(x)
.
But the deriva ve is notcon nuous at 0!
OutlineRates of Change
Tangent LinesVelocityPopula on growthMarginal costs
The deriva ve, definedDeriva ves of (some) power func onsWhat does f tell you about f′?
How can a func on fail to be differen able?Other nota onsThe second deriva ve
Notation
I Newtonian nota on
f′(x) y′(x) y′
I Leibnizian nota on
dydx
ddx
f(x)dfdx
These all mean the same thing.
Meet the MathematicianIsaac Newton
I English, 1643–1727I Professor at Cambridge(England)
I Philosophiae NaturalisPrincipia Mathema capublished 1687
Meet the MathematicianGottfried Leibniz
I German, 1646–1716I Eminent philosopher aswell as mathema cian
I Contemporarily disgracedby the calculus prioritydispute
OutlineRates of Change
Tangent LinesVelocityPopula on growthMarginal costs
The deriva ve, definedDeriva ves of (some) power func onsWhat does f tell you about f′?
How can a func on fail to be differen able?Other nota onsThe second deriva ve
The second derivative
If f is a func on, so is f′, and we can seek its deriva ve.
f′′ = (f′)′
It measures the rate of change of the rate of change!
Leibniziannota on:
d2ydx2
d2
dx2f(x)
d2fdx2
The second derivative
If f is a func on, so is f′, and we can seek its deriva ve.
f′′ = (f′)′
It measures the rate of change of the rate of change! Leibniziannota on:
d2ydx2
d2
dx2f(x)
d2fdx2
Function, derivative, second derivative
.. x.
y
.
f(x) = x2
.
f′(x) = 2x
.f′′(x) = 2
SummaryWhat have we learned today?
I The deriva ve measures instantaneous rate of changeI The deriva ve has many interpreta ons: slope of the tangentline, velocity, marginal quan es, etc.
I The deriva ve reflects the monotonicity (increasing-ness ordecreasing-ness) of the graph