Comprehensive Guide to Integrals
Comprehensive Guide to Integrals
1. Integrals
(Chapter 5)
Nguyễn Anh Tú
natu@hcmiu.edu.vn
Indefinite Integrals
Antiderivatives
Let f and F be functions defined on (a, b), such that F ′ (x) = f (x),
for all x ∈ (a, b). Then F is called an antiderivative of f on (a, b).
Example
• The antiderivative of f (x) = x 2 is x 3 /3 + C .
• A particle moves on a straight line with its acceleration given by
a(t) = 6t + 4. If the initial velocity is v (0) = 6 cm/s and the
initial position is s(0) = 9 cm. Find its position function.
Proof. Take any two points x1 ̸= x2 in (a, b). Then by the Mean
Value Theorem for derivatives, there exists c ∈ (a, b) such that
F (x2 ) − F (x1 )
F ′ (c) = .
x2 − x1
Since by hypothesis, F ′ (c) = 0, we obtain F (x1 ) = F (x2 ).
Theorem
Suppose F and G are two antiderivatives of f on (a, b). Then there is
constant C so that G (x) = F (x) + C on (a, b).
Definition
We call the collection of all antiderivatives of f the indefinite integral
of f , denoted by Z
f (x)dx.
If F is an antiderivative of f then
Z
f (x)dx = F (x) + C .
x a+1
Z
x a dx = +C
a+1
a = −1:
1
Z
dx = ln |x| + C
x
Z Z
sin xdx = − cos x + C , cos xdx = sin x + C ,
Z Z
tan xdx = − ln | cos x| + C , cot xdx = ln | sin x| + C ,
Z Z
sec2 xdx = tan x + C , csc2 xdx = − cot x + C ,
Z Z
sec x tan xdx = sec x + C , csc x cot xdx = − csc x + C ,
Z Z
dx
sec xdx = ln | sec x + tan x| + C , = arctan x + C
1 + x2
Z Z
dx
csc xdx = − ln | csc x + cot x| + C , √ = arcsin x + C
1 − x2
Definite Integrals
lim R(f , P, c)
∥P∥→0
n
X
L(f , P) = ∆xi inf f (x).
[xi−1 ,xi ]
i=1
• Upper sum:
h i
2 2 2 3 2 4 2
U4 = 14 41 + 15
4
+ 4
+ 4
= 32
= 0.46875
• Lower sums:
h i
2 1 2 2 2 3 2
L4 = 14 04 + 7
4
+ 4
+ 4
= 32
= 0.21875
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 15 / 106
Approximation with 8 rectangles
U8 = 0.3984375, L8 = 0.2734375
Note that U1000 and L1000 agree to 2 decimal places and their average
L1000 +U1000
2
= 0.3333335 is correct to 6 decimal places.
Definition
A bounded function f on [a, b] is integrable if it satisfies one of the
equivalent properties above. The definite integral of f on [a, b] is
denoted Z b
f (x)dx = lim R(f , P, c).
a ∥P∥→0
Remarks:
• The definite integral ab f (x)dx is a number, whereas f (x)dx
R R
is a collections of functions.
• Any letter can be used for the variable of integration
Z b Z b Z b
f (x)dx = f (t)dt = f (r )dr
a a a
ci = 3i/n, i = 0, . . . , n
• Linearity:
Z b Z b Z b
[αf (x) + βg (x)] cdx = α f (x)dx + β g (x)dx,
a a a
• If a ≤ b then
Z b Z b
f (x)dx ≤ |f (x)|dx.
a a
Definition
If f is integrable on [a, b] then the average value or mean value of f
on [a, b] is
Z b
¯ 1
f = f (x)dx.
b−a a
The MVT states that a continuous function attains its mean value on
the interval.
Theorem
If f is continuous on [a, b] there exists c ∈ [a, b] so that
b
1
Z
f (c) = f (x)dx.
b−a a
G ′ (x) = f (x).
In other words, Z x
d
f (t)dt = f (x).
dx a
Thus, if |x − c| < δ
Z x
G (x) − G (c) 1
− f (c) = (f (t) − f (c))dt
x −c |x − c| c
1
≤ |x − c| max |f (t) − f (c)|
|x − c| t∈[c,x]
1
≤ |x − c|ϵdt = ϵ.
|x − c|
This implies that
G (x) − G (c)
lim = f (c).
x→c x −c
It follows that G ′ (c) = f (c).
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 35 / 106
Proof of FTC part 2
Rx
Let G (x) = a f (t)dt be the area function. By FTC part 1, G is an
antiderivative of f . We know that two antiderivatives of f differ by a
constant, hence
F (x) = G (x) + C .
Therefore,
Z b
F (b) − F (a) = G (b) − G (a) = f (x)dx.
a
Theorem
If u and v are differentiable and
Z v (x)
F (x) = f (t)dt,
u(x)
then
F ′ (x) = f (v (x))v ′ (x) − f (u(x))u ′ (x).
Note: The FTC part 1 is a special case where u(x) = a and v (x) = x.
Find derivative of Z x2
F (x) = ln(t 3 + 4)dt
3
and Z 5
G (x) = tan(cos(u) − 4)du
2x
Solution:
F ′ (x) = 2x ln(x 6 + 4)
and
G ′ (x) = −2 tan(cos(2x) − 4)
√
• Apply the theorem to f (u) = u and u(x) = x 2 + 1.
√
• We have f (u(x))u ′ (x) = 2x x 2 + 1.
• So Z √ Z
√ 2
2
2x x + 1dx = udu = u 3/2 + C .
3
• Substitute u by x 2 + 1 we get
Z √ 2
2x x 2 + 1dx = (x 2 + 1)3/2 + C .
3
• Write x 4 in terms of u
x 4 = (x 2 + 1 − 1)2 = (u − 1)2
• u = 3 − 5x and du = −5dx → dx = − 51 du
• x = 1 → u = −2, x = 2 → u = −7
• We obtain
Z 2 Z −7
dx 1 du 1 −7 1
2
= −5 2
=− −2
=
1 (3 − 5x)
Nguyễn Anh Tú (natu@hcmiu.edu.vn)
−2 u 5u
1. Integrals
14
Fall 2022 48 / 106
Section 5
Integration by Parts
Theorem
If u and v are differentiable then
Z Z
udv = uv − vdu.
• u = x, dv = sin xdx
• du = dx, v = − cos x
•
Z Z
x sin xdx = x(− cos x) − (− cos x)dx
Z
= −x cos x + cos xdx
= −x cos x + sin x + C
R
Find ln xdx
• u = ln x, dv = dx
• du = x1 dx, v = x
•
1
Z Z
ln xdx = x ln x − x dx
x
= x ln x − x + C
R
Evaluate I = e x sin xdx
• u = sin x, dv = e x dx
• du = cos xdx, v = e x
Z
x
I = e sin x − e x cos xdx
= e x (sin x − cos x) − I
So
2I = e x (sin x − cos x) + C
or
I = 12 e x (sin x − cos x) + C
Calculate In = x n e −x dx
R
• u = x n , dv = e −x dx
• du = nx n−1 dx, v = −e −x
Integrating by parts, we get
Z
n −x
In = −x e +n x n−1 e −x dx
= −x n e −x + nIn−1 .
I0 = −e −x + C .
Partial Fractions
Evaluate
x 3 + 3x 2
Z
dx
x2 + 1
Divide
x 3 + 3x 2 x +3
= x + 3 −
x2 + 1 x2 + 1
x 3 + 3x 2 x +3
Z Z Z
dx = (x + 3)dx − dx
x2 + 1 x2 + 1
1 1
= x 2 + 3x − ln(x 2 + 1)
2 2
− 3 arctan x + C
I.e.
k
Y l
Y
Q(x) = a (x + bi )ni
(x 2 + cj x + dj )mj .
i=1 j=1
If Q(x) has a linear factor (ax + b)n then the partial fraction
P(x)
decomposition of Q(x) contains the terms
A1 A2 An
+ + · · · + .
(ax + b) (ax + b)2 (ax + b)n
x + 5 = A(x + 2) + B(x − 1)
x 4 − 2x 2 + 4x + 1 4x
3 2
=x +1+ 3 2
x −x −x +1 x −x −x +1
• Factorzing Q: x 3 − x 2 − x + 1 = (x − 1)2 (x + 1)
• Form of partial fraction decomposition
4x A B C
2
= + 2
+
(x − 1) (x + 1) x − 1 (x − 1) x +1
for all x.
• Substituting special values of x:
• x = 1 =⇒ B = 2
• x = −1 =⇒ C = −1
• x = 0 =⇒ A = 1
• We obtain
Z
1 2 1
I = x +1+ + − dx
x − 1 (x − 1)2 x + 1
x2 2
= + x + ln |x − 1| − − ln |x + 1| + C .
2 x −1
Trigonometric Integrals
sec2 x = tan2 x + 1
sec′ x = sec x tan x
tan′ x = sec2 x = tan2 x + 1
= 13 sin3 x − 51 sin5 x + C .
• Find I = sin2 x cos2 xdx.
R
Solution:
1
Z
I = (1 − cos 2x)(1 + cos 2x)dx
4
1
Z
= (1 − cos2 (2x))dx
4
1 1
Z
= (1 − (1 + cos(4x)))dx
4 2
1 1
= x− sin(4x) + C .
8 32
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 73 / 106
R
tann x secm xdx
1 If n = 2k + 1 is odd: Let u = sec x, then
Z Z
n m
tan x sec xdx = tan2k x secm−1 x sec x tan xdx
Z
= (u 2 − 1)k u m−1 du.
Z
= u n (u 2 + 1)k−1 du.
Trigonometric Substitution
√
• Used to get rid of square roots of the form Ax 2 + Bx + C .
• Can always write Ax 2 + Bx + C = ±(αx + β)2 + γ.
• Thus, by the change of variable u = αx + β, we reduce to one
of the following three cases
√
• a2 − u 2 , let u = a sin t.
√
• a2 + u 2 , let u = a tan t.
√
• u 2 − a2 , let u = a sec t.
2 sec2 tdt 1 1
Z Z Z
sec t cos t
I = = dt = dt
2
4 tan t 2 sec t 4 2
tan t 4 sin2 t
• Let t = sin t then
1 1 1
Z
dt
I = 2
=− +C =− +C
4 t 4t 4 sin t
3 − 2x − x 2 = 4 − (x + 1)2
• Let u = x + 1 then
u−1
Z
I = √ du
4 − u2
• Let u = 2 sin t then
2 sin t − 1
Z
I = 2 cos tdt = −2 cos t − t + C
2 cos t
√ x +1
= − 3 − 2x − x 2 − sin−1 ( ) + C.
2
√
Here we have used 2 cos t = 4 − u 2 and t = sin−1 (u/2).
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 81 / 106
Section 9
Improper Integrals
∞ b
1 1
Z Z
dx = lim dx
1 x2 b→∞ 1 x
2
1
= lim 1 − =1
b→∞ b
p-integrals of Type 1
For 0 < a < ∞,
Z ∞ a1−p
−p
n
p−1
if p > 1
x dx =
a ∞ if p ≤ 1
Evaluate
Z +∞
2 arctan 2x
1 dx
1 + 4x 2
Z0 ∞
ln x
2 dx.
x3
Z1 ∞
3 xe −x dx.
Z0 ∞
dx
4
2
.
−∞ 1 + x
Z ∞
dx
5 √ .
x
R 1∞
6
0
cos xdx.
5 5
1 1
Z Z
√ dx = lim+ dx √
2 x −2 t→2 t x −2
√ 5
= lim+ 2 x − 2 t
t→2
√ √
= lim+ (2 3 − 2 t − 2)
t→2
√
= 2 3.
p-integral of Type 2
For 0 < a < ∞,
a a1−p
Z
−p
n
1−p
if p < 1
x dx =
0 ∞ if p ≥ 1
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 88 / 106
Example
Evaluate Z 1
I = ln xdx.
0
Solution:
Z 1
1
I = lim+ ln xdx = lim+ (x ln x − x) c
c→0 c c→0
ln c
= −1 − lim+ (c ln c − c) = −1 − lim+
c→0 c→0 1/c
L’Hôpital 1/c
= −1 − lim+ = −1 − 0 = −1
c→0 −(1/c 2 )
Theorem I
Let f and g be continuous on (a, b), and 0 ≤ f (x) ≤ g (x) for
x ∈ (a, b).
• If ab g (x)dx converges then ab f (x)dx also converges, and
R R
Z b Z b
f (x)dx ≤ g (x)dx.
a a
Rb Rb
• If a
f (x)dx diverges then a
g (x)dx diverges.
Theorem II
Let f and g be continuous on (a, b), and |f (x)| ≤ g (x) for
x ∈ (a, b).
• If ab g (x)dx converges then ab f (x)dx also converges, and
R R
Z b Z b
f (x)dx ≤ g (x)dx.
a a
Rb Rb
• If a
f (x)dx diverges then a
g (x)dx diverges.
Example
Show that Z ∞
2
I = e −x dx
0
converges.
Thus, I converges.
Determine
Z ∞ whether each integral is convergent or divergent.
2
sin x
1. dx.
1 x2 + 1
Z ∞
2
2. xe −x dx.
0
∞
x +1
Z
3. dx.
1 x 2 + 2x
Z ∞
ln x
4. dx.
1 x
Z ∞
dx
5. .
−∞ 4x 2 + 4x + 5
Numerical methods
Theorem
If f ′′ is continuous on [a, b] and |f ′′ (x)| ≤ K then
Z b
K (b − a)3
f (x)dx − Mn ≤ .
a 24n2
Theorem
If f ′′ is continuous on [a, b] and |f ′′ (x)| ≤ K then
b
K (b − a)3
Z
f (x)dx − Tn ≤ .
a 12n2
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 1. Integrals Fall 2022 102 / 106
Example
h
Sn = (y0 + 4y1 + 2y2 + 4y3 + ... + 2yn−2 + 4yn−1 + yn
3
h X X X
= ( yend + 4 yodd + 2 yeven )
3
Theorem
Suppose |f (4) (x)| ≤ K on [a, b], then
b
K (b − a)5
Z
| f (x)dx − Sn | ≤ .
a 180n4
Example
2
1
Z
Find n so that the error of approximating dx using the Simpson
1 x
rule is less than 10−4 .
Solution:
24
• Since f (4) (x) = , K = 24
x5
24
• Need n4 > or n > 6.866....
108 · 10−4
• Take n = 8.
Nguyễn Anh Tú
natu@hcmiu.edu.vn
1 Lengths of Curves
3 Volumes
4 Disk Method
Lengths of Curves
x = x(t), y = y (t).
Example: The circle centered at (0, 0) of radius 1 can be
parameterized by x = cos(t), y = sin(t), t ∈ [0, 2π]
can be parameterized by
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 2. Applications of Integrals Fall 2022 5 / 39
• Parametric equations describe a moving point.
• The range of this point is the curve.
• There are more than one way to parameterize a curve. E.g.
• x = cos(2t), y = sin(2t), t ∈ [0, 2π] is another
parameterization for the unit cirle.
The point now goes around the circle twice, counterclockwise,
starting from (1, 0)
• x = sin(2t), y = cos(2t), t ∈ [0, 2π] is another
parameterization for the unit cirle.
The point goes around the circle twice, clockwise, starting from
(0, 1).
• Suppose x(t) and y (t) are differentiable. Then the tangent line
to the curve at (x(t), y (t)) has direction (x ′ (t), y ′ (t)). Thus, the
formula is
x − x(t) y − y (t)
′
= .
x (t) y ′ (t)
• Example: Let C be the parametric curve
x = t 2, y = t 3 − 3t.
Show that C has two tangent lines at (3, 0) and find their
equations.
• Thus,
n
X p
L≈ (x ′ (ti ))2 + (y ′ (ti ))2 ∆t.
i=1
p
• This is a Riemann sum for the function (x ′ (t))2 + (y ′ (t))2 .
Thus, passing to the limit n → ∞, we obtain
Z b p
L= (x ′ (t))2 + (y ′ (t))2 dt.
a
• y = x 3/2 so y ′ = 23 x 1/2
• Arc length
Z 4q
L= 1 + 94 xdx
1
Subs: u = 1 + 94 x
Z 10 √
4
L= 9
udu
13/4
1
√ √
= 27
(80 10 − 13 13)
Notes: In general, we may need to find the zeros of f (x) − g (x) and
determine its signs in between these points.
xL = 12 y 2 −3 and xR = y +1
Z 4
(y + 1) − ( 21 y 2 − 3) dy
A=
−2
3
y y2 4
=− + + 4y = 18
6 2 −2
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 2. Applications of Integrals Fall 2022 22 / 39
Section 3
Volumes
Disk Method
• Lies between x = −r to
x =r
• Cross-section at x is a
circle
√ of radius
r − x 2 , so
2
A(x) = π(r 2 − x 2 ).
• Thus,
R1 R1
Answer: V = 0
A(x)dx = 0
πxdx = π2 .
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 2. Applications of Integrals Fall 2022 29 / 39
Example
Find the volume of the solid obtained by rotating the region bounded
by y = x 3 , y = 8 and x = 0 about the y -axis.
R8 R8 96π
Answer: V = 0
A(y )dy = 0
πy 2/3 dy = 5
.
R1 R1 2π
Answer: V = 0
A(x)dx = 0
π(x 2 − x 4 )dx = 15
.
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 2. Applications of Integrals Fall 2022 31 / 39
Section 5
• Sn is a Riemann sum
• Let n → ∞ then Sn converges to the volume of the solid
Z b
V = lim Sn = 2πxf (x)dx.
n→∞ a
Z b
• To memorize: V = 2π[radius][height]dx
a
Nguyễn Anh Tú (natu@hcmiu.edu.vn) 2. Applications of Integrals Fall 2022 36 / 39
Example
Find the volume of the solid created by rotating about y −axis the
region bounded by y = 2x 2 − x 3 and y = 0.
• Shell has radius x,
interval [0, 2]
• the height
f (x) = 2x 2 − x 3
• Volume
Z 2
16
V = 2πx(2x 2 −x 3 )dx = π
0 5
2x 2 − x 3 = y .
Note: The Disk Method uses integration along the axis of rotation,
while the Shell Method uses integration along the other axis.
Nguyễn Anh Tú
natu@hcmiu.edu.vn
1 Sequences
2 Limits of Sequences
3 Series
4 Tests of Convergence
5 Power Series
Sequences
a1 , a2 , a3 , a4 , . . . , an , . . .
The sequence { n1 }.
Nguyễn Anh Tú natu@hcmiu.edu.vn 3. Sequences and Series Fall 2022 5 / 91
Descriptions of sequences
1 ∞
The sequence n n=1
can be described as
• 1, 21 , 13 , 14 , . . .
• an = n1 , n = 1, 2, . . . .
• a1 = 1, an+1 = ana+1 n
.
an an−1
• a1 = 1, a2 = 12 , an+1 = 2an−1 −an
.
Limits of Sequences
lim an = L or an → L as n → ∞
n→∞
1 (−1)n
Limits of n
and n
.
Suppose that {an } and {bn } are convergent sequences (to finite
limits) and c is a constant. Then,
1 limn→∞ (an + bn ) = limn→∞ an + limn→∞ bn
• Examples: Evaluate
1
−1
lim n tan .
n→∞ n
2 Find
3n2 − 12n + 7
lim cos( ).
n→∞ 5n2 + n − 9
3 Find
lim n1/n .
n→∞
lim an = lim cn = L.
n→∞ n→∞
Then
lim bn = L.
n→∞
an ≥ m for all n.
Theorem
An unbounded sequence is divergent.
lim an = ∞ or an → ∞.
n→∞
lim an = −∞ or an → −∞.
n→∞
Series
a1 + a2 + a3 + · · · + an + · · ·
is divergent.
Theorem
If an ̸→ 0, i.e. if either
• limn→∞ an does not exist
• or limn→∞ an = L ̸= 0
P
then the series an diverges.
are divergent.
P P
If
P the series a n and bn are P
convergent, then P
so are the series
can (where c is a constant), (an + bn ), and (an − bn ).
Furthermore,
P P
(i) can = c an ;
P P P
(ii) (an + bn ) = an + bn ;
P P P
(iii) (an − bn ) = an − bn .
Tests of Convergence
Definition
A nonnegative series is a series whose terms are all nonnegative.
Theorem
A nonnegative series converges if and only if its partial sums are
bounded from above.
for convergence.
is called the pP
-series,
When p = 1, ∞ 1 1 1
n=1 n = 1 + 2 + 3 + · · · is also called the harmonic
series.
Theorem
P∞ 1
The series n=1 np converges if p > 1 and diverges if p ≤ 1.
P P
Suppose that an and bn are series with nonnegative terms. If
an
lim =K
n→∞ bn
where K is a finite number and K > 0, then either both series
converge or both diverge.
Example: Test the given series for convergence or divergence
∞ √ ∞
X 2n2 − 5 n + 1 X 1
, n sin( ).
n=1
7n4 − 3n3 + 9n − 1 n=1 n2
Suppose that
(i) 0 ≤ cn+1 ≤ cn for all n and
(ii) limn→∞ cn = 0,
then the series is convergent.
Furthermore, the sum s of the series satisfies
|s − sn | ≤ cn+1 .
satisfies
1
i) 0 < cn+1 < cn because n+1 < n1 ;
ii) cn = n1 → 0.
Thus, it converges by the Alternating Series Test.
• Similarly, the alternating p−series
∞
X (−1)n+1
n=1
np
or
n ≥ 9.
• The series
P
an converges Pabsolutely (is absolutely
convergent) if the series |an | converges.
• Absolutely convergence is stronger than convergence:
X X
|an | converges =⇒ an converges .
is convergent or divergent.
• The series
P
an converges conditionally P (is conditionally
convergent) if it converges but the series |an |, diverges.
• In other words, Conditional Convergence = Convergence, but
not Absolutely Convergence.
Example: The alternating p-series
∞
X (−1)n+1 1 1 1
=1− + p − p + ··· , p>0
n=1
np 2p 3 4
converges
• absolutely if p > 1,
• conditionally if 0 < p ≤ 1.
an+1
ρ = lim .
n→∞ an
• Then
(a) if ρ < 1, the series absolutely converges.
(b) if ρ > 1, the series diverges.
• In the case ρ = 1, no conclusion can be drawn.
Example: Prove that ∞ 1
P
n=0 n! converges.
Solution Since an = 1/n!,
an+1 1
ρ = lim = lim = 0 < 1.
n→∞ an n→∞ n + 1
• Then
(a) if ρ < 1, the series absolutely converges.
(b) if ρ > 1, the series diverges.
• In the case ρ = 1, no conclusion can be drawn.
• If ρ = 1 in the Ratio Test, don’t try the Root Test because ρ will
again be 1.
2n+3 n
Example: Determine whether ∞
P
n=1 3n+2 converges.
2(n+1)/n
ρ = lim =0<1
n→∞ n
so the series converges absolutely.
• For the series n (1 + n1 )n2
P
1
ρ = lim (1 + )n = e > 1
n→∞ n
so the series diverges.
• Determine whether ∞ 2n+3 n
P
n=1 3n+2
converges.
P (−1)n n2
1 n 4n
.
P 2n−1 n
2 n 4n+3
P (−1)n √n3 +1
3 n n2 +9
P∞ 3n−1
4 n=1 2
P∞ 2n 1+1
5 n=2 n(ln n)s
P 3n
6 n n
P 5 −4n+1
7
2n3 −2n+3
P n4
8 n
Pn eln n
9 n n+10
Power Series
Then R = L1 .
Nguyễn Anh Tú natu@hcmiu.edu.vn 3. Sequences and Series Fall 2022 55 / 91
Interval of Convergence
• The set of all values of x for which the power series
P∞ n
n=0 cn (x − a) converges is called interval of convergence
of the power series.
• It is an interval of length R centered at a. The interval can be
open, closed, or half-open.
• In case (i) the interval consists of just a single point a.
• In case (ii) the interval is (−∞, ∞).
• In case (iii) there are four possibilities for the interval of
convergence:
(a − R, a + R), (a − R, a + R], [a − R, a + R) or [a − R, a + R].
1.
Find the radius of convergence R, using the formula.
If R = 0, the interval of convergence is {a}.
2.
If R = ∞, the interval of convergence is (−∞, ∞).
3.
If R < ∞, test for convergence at each endpoint a ± R, using
4.
the Comparison Test, the Integral Test, or the Alternating Series
Test.
Example: Determine the center, radius, and interval of
convergence of
∞
X (2x + 5)n
2 + 1)3n
.
n=0
(n
• If f (x) = ∞ n
P
n=0 cn (x − a) for all x in the interval
(a − R, a + R), we say that the power series is a representation
of f (x) on that interval
P∞ or f is expanded into/represented by
n
the power series n=0 cn (x − a) . P
1
E.g. f (x) = 1−x is represented by ∞ n
n=0 x on (−1, 1).
• To find the power series representation of a function, or vice
versa, we use
1 Algebraic manipulations
2 Differentiation and Integration
3 Taylor series
replacing x by −x we get
∞
1 X
= (−1)n x n .
1+x n=0
Solution: By differentiation,
∞
′
X 1
f (x) = (−1)n x 2n = , −1 < x < 1.
n=0
1 + x2
Thus,
1
Z Z
′
f (x) = f (x)dx = dx = arctan x + C .
1 + x2
Substituting x = 0, we obtain C = 0. Thus,
3 5
x − x3 + x5 − · · · = arctan x, −1 ≤ x ≤ 1.
Nguyễn Anh Tú natu@hcmiu.edu.vn 3. Sequences and Series Fall 2022 63 / 91
Section 7
Theorem
If f has a power series representation
∞
X
f (x) = cn (x − a)n
n=0
f (n) (a)
cn = .
n!
Definition 10.1
The Taylor polynomial of order n of f at a is
f ′′ (a)
Tn (x) = f (a) + f ′ (a)(x − a) + (x − a)2 +
2!
f (n) (a)
+ ··· + (x − a)n .
n!
Thus, Tn is the n-th partial sum of the Taylor series of f at a.
f (x) = e x and and its Taylor polynomials T1 (x), T2 (x), and T3 (x).
If we let
Rn (x) = f (x) − Tn (x)
then Rn (x) is called the remainder of the Taylor series and (1) is
equivalent to
Theorem
The following holds for the Taylor remainders of f
• (The integral form of the remainder term)
x
f (n+1) (t)
Z
Rn (x) = (x − t)n .
a n!
• (Lagrange’s form of the remainder term)
f (n+1) (s)
Rn (x) = (x − a)n+1
(n + 1)!
Corollary
Let f (x) be an infinitely differentiable function on the open interval
I = (a − R, a + R) with R > 0. Assume there exists a constant
K ≥ 0 such that for all n,
Let Z x
2
f (x) = e −t dt
0
By the Alternating Series Test, the error when using the first n terms
is less than
1
.
(2n + 3)(n + 1)!
This is less than 0.0005 for n = 5.
Thus, we obtain
5
X (−1)n
f (1) ≈ ≈ 0.746729...
n=0
(2n + 1)n!
f (n) (0)
k(k − 1) · · · (k − n + 1) k
= = .
n! n! n
∞
X k k(k − 1) 2 k(k − 1)(k − 2) 3
x n = 1 + kx + x + x +
n 2! 3!
n=0
k(k − 1)(k − 2) · · · (k − n + 1) n
+ ··· + x + ···
n!
The Ratio Test shows that this series has radius of convergence
R = 1. Furthermore, the binomial series converges to (1 + x)k for
|x| < 1.
Theorem
If k ∈ R and |x| < 1, then
∞
X k
k
(1 + x) = xn
n
n=0
∞
X k(k − 1) · · · (k − n + 1) n
= x
n=0
n!
1
1 Expand in powers of x + 2.
2 + 3x
x
2 Expand in powers of x.
1 + 2x 2
3 Expand ln x in powers of x − 2. Where does the expansion holds?
4 Expand cos(πx) in powers of x − 1.