1

Chapter 3 Power Series

Definition 3.1: A power series is a series of the form


 cn ( x  a)n  c0  c1( x  a)  c2 ( x  a)2  ...  ck ( x  a)k  ... (1)
n 0
where a , c0 , c1, c2 ,..., ck ,... are constant and x is a
variable.
The constant “a” is called the center of the
power series, while the constants “ c0 , c1, c2 ,..., ck ,...”
are called coefficients of the power series.

If a  0 , we call (1) a power series in x.

 xn   2n 1
n x
Example:  ,  n! x and  (1)
n
are
n 0 n! n 0 n 0 (2n  1)!
all power series in x

If a  0 , we call (1) a power series in x  a or a

power series centered at a,
 ( x  1) n
Example:  is a power series in x  1, and
n 1 n  1
 n
( x  3)
 (1)n is a power series centered at -3 .
n 0 n!
 2

Since x can be any number, the power series (1)

may be either convergent or divergent depending on
the value of x.
 ( x  5) n
For example, consider the power series  2
.
n 1 n
 1
If x  6 , the above series becomes  2
which is a
n 1 n
p-series with p = 2. It then converges.
 ( 2) n  (1) n 2n
If x  3 , the series becomes  2  2
n 1 n n 1 n
which is a divergent alternating series.


Question? For a power series  cn ( x  a)n , what are
n 0
the values of x that make this series converges?

Observe that any power series


 cn ( x  a)n  c0  c1( x  a)  c2 ( x  a)2  ...  ck ( x  a)k  ...
n 0
always converges at its center x  a . (Why?)

Then where else does it converge? The following

theorem states about the convergence of a power
series.
 3

Theorem 3.2: (Radius of convergence)


Let F ( x)   cn ( x  a)n be a power series in x-a.
n 0

Then there are only three possibilities:

(i) F ( x) converges only at x  a , or
(ii) F ( x) converges for all x , or
(iii) there is a real number R0 such that F ( x)
converges absolutely if x  a  R ( x  (a  R, a  R) ) and
diverges if x  a  R ( x  (, a  R)  (a  R, ) ). It may
or may not converge at the end points x  a  R
( x  a  R or x  a  R ).
In case (i), set R  0 , and in case (ii), set R   .
We call R the radius of convergence of the power

series F ( x)   cn ( x  a)n .
n 0
By the above theorem, we note that the set of all
x that make the series converge in each case can be
expressed as an interval. Namely, a  [a, a] for case
(i), (, ) for case (ii), and one of the following
intervals (a  R, a  R) , (a  R, a  R] , [a  R, a  R) , or
 4

[a  R, a  R] for case (iii). This leads to the following

definition.
Definition 3.3: The set of all values of x that make a
power series converge is called the interval of
convergence of the power series.

The procedure used to find the interval of

convergence of a power series.
This procedure contains 2 steps.
Step 1: Finding the radius of convergence of the
power series by applying the ratio test or the n th root
test so that we get the convergence interval of x
where x  a  R or a  R  x  a  R . To do this, we
set
an 1
lim  1, or lim n an  1,
n  n 
an
where an is the n-th term of a given power series.
(Here, an  cn ( x  a)n ).
 5

Step 2: Test the convergence of this power series at

the end points x = a + R and x = a – R by the
methods we know such as comparison test, integral
test and alternating series test.

Example: Find x which makes the series converge.

 xn 
a.  b.  n! x n
n 0 n! n 0
 6

Problem: Find the intervals of convergence of the

following series.

 xn  ( x  2)n
a.  (1)n
b. 
n 1 n(n  1) n 1 n  3
n
 7

Exercise 3.1
Find the radius of convergence and the interval of
convergence of the following power series.
 8

Solution of Exercise 3.1

Taylor and Maclaurin Series

Both series are power series used to approxi-

mate other functions.

Definition 3.4 (Taylor’s Series )

Let f be a function that has derivatives of all orders
at a point a. The Taylor Series of f about x=a is
( x  a)2 ( x  a)n
f ( x)  f (a)  f (a)( x  a)  f (a)  f ( n)
(a)  (1)
2! n!
 9

Definition 3.5 (Taylor’s Polynomial)

Let f be a function that has derivatives up to
order n at a point a, the n-th Taylor Polynomial of f
about x = a is
( x  a)2 ( x  a)n
f (a)  f (a)( x  a)  f (a)  . . .  f (a)
( n )
. (2)
2! n!

We can define the polynomial (of degree n) Pn(x) as

follows
( x  a)2 ( x  a)n
Pn ( x)  f (a)  f (a)( x  a)  f (a)  . . .  f ( n) (a ) .
2! n!

If a = 0 in either (1) or (2), we call it Maclaurin

Series or Maclaurin Polynomial, respectively.

Example: Given f ( x)  e x , write down the Maclaurin

polynomial of f.
Solution Let f ( x)  e x .
Then f ( x)  f ( x)  f ( x)  ...  f (n) ( x)  e x .

Hence, f (0)  f (0)  f (0)  ...  f (n) (0)  e0  1 .

 10

 P0 ( x)  f (0)  1

P1( x)  f (0)  f (0) x  1  x

f (0) 2 x2 1
P2 ( x)  f (0)  f (0) x  x 1 x   1  x  x2
2! 2! 2
f (0) 2 f (0) 3
P3 ( x)  f (0)  f (0) x  x  x
2! 3!
x 2 x3 1 1
1 x    1  x  x 2  x3
2! 3! 2 6

f (0) 2 f (0) 3 f (n) (0) n

Pn ( x)  f (0)  f (0) x  x  x  ... x
2! 3! n!
x 2 x3 xn
1 x    ... 
2! 3! n!

Figure 1

Figure 1 shows that the higher order of n, the

closer polynomial Pn(x) gets to the function f ( x)  e x .
 11

The Maclaurin Polynomial is just an approxi-

mation of a function since we cut off the tail of an
infinite series. Thus, there exists a truncation error.
We have also found that the Maclaurin series is
more accurate when x gets closer to zero. That’s
why we have better use a Taylor series when x is far
from zero.

Exercise: Find the 3rd Taylor polynomial of


f ( x)  sin x about x  .
3

Solution: Let f ( x)  sin x .

   3
Since f ( x)  sin x , f    sin 
3 3 2
   1
f ( x)  cos x , f     cos 
3 3 2
   3
f ( x)   sin x , f      sin  
3 3 2

   1
f ( x)   cos x , f      cos  
3 3 2
Therefore,

3  3  3
f    
f  3
3  
P3 ( x)  f   f   x   
2
x
3
3  x
2! 3! 3


3 1
 x 
2 2 3 
3
2.2!
x 
3 2.3!

 2  1 x   3.
3    
 12

3.6 Differentiation and Integration of Power

Series

We can do it term by term:


Let f ( x)   cn ( x  a) n be a power series with
n 1

radius of convergence R > 0. Then we have

1. f(x) is continuous for all x  a  R
 c ( x  a)
n 1

2.  f ( x)dx   n  C; x  a  R
n 0 n 1

3. f '( x)   ncn ( x  a)n1 ; x  a  R
n 0

Well known Maclaurin Series


1
1.  1  x  x  ...  x  ...   x n ,  1  x  1
2 n

1 x n 0

x2 xn 
 xn 
2. e  1  x   ...   ...     ,    x  
x

2! n! n 0  n ! 

x3 x5 (1)n x 2 n 1 
 (1)n x 2 n 1 
3. sin x  x    ...   ...     ,   x  
3! 5! (2n  1)! n 0  (2 n  1)! 
x2 x4 (1)n x 2 n 
 (1)n x 2 n 
4. cos x  1    ...   ...     ,   x  
2! 4! (2n)! n 0  (2n)! 
x 2 x3 (1)n x n1 
(1)n x n1
5. ln 1  x  x    ...   ...   , 1  x  1
2 3 n 1 n 0 n  1
 13

1
Example Find Maclaurin series of f ( x)  tan x .
Solution Since we know
1
tan 1 x   dx
1 x 2

  (1  x 2  x 4  x 6  x 8  ...)dx
x3 x5 x7
Cx    ...
3 5 7
Evaluate C by plugging in x  0
Thus C  tan 1 (0)  0

1 x3 x5 x7
Then tan x  x     ...
3 5 7
 2 n 1
x
  (1) n
n 0 2n  1
1
The radius of convergence of f ( x ) 
1  x 2 is 1.
1
Hence, the radius of convergence of f ( x)  tan x is
also 1, and its interval of convergence is  1,1 .
 14

3.7 The Binomial Series

Let k be any real number and x  1.

x2 x3
(1  x)  1  kx  k (k  1)  k (k  1)(k  2)  ...
k

2! 3!

k  n
   x
n 0  n 

 k  k (k  1)...(k  n  1) k 
where    , n  1 and    1
 
n n ! 0
calculated from Maclaurin series of f ( x)  (1  x) k
as follows:
f ( x)  (1  x) k and f (0)  1
f ( x)  k (1  x)k 1
f ( x)  k (k  1)(1  x)k 2
f ( x)  k (k  1)(k  2)(1  x)k 3
f ( n ) ( x)  k (k  1)...(k  n  1)(1  x)k n
and f ( n ) (0)  k (k  1)...(k  n  1) .
Hence, Maclaurin series of f ( x)  (1  x) k is
 
xn xn
f
n 0
(n)
(0)   k (k  1)...(k  n  1) .
n! n 0 n!
 15

Find the interval of convergence:

k (k  1)...(k  n  1)(k  n) x n1
an1 (n  1)!
lim  lim
n  a
n
n k (k  1)...(k  n  1) x n
n!
k n
 x lim  x 1
n n  1

Example Find the power series of f ( x)  1/(1  x)2

and its interval of convergence.
Solution Apply binomial series when k  2 :
 2  2(3)...(2  n  1)
 n  ,n 1
  n !
(1)n 2(3)...n(n  1)
  (1)n (n  1)
n!
Thus f ( x)  1/(1  x)2


 2  n 
    x   (1) n (n  1) x n , x  1
n 0  n  n 0

Example Find the power series of f ( x)  1/ 4  x

and its radius of convergence.
 16

Solution Consider
1/ 2
1 1 x
f ( x)  1/ 4  x   1   .
x 2 4
4(1  )
4
Apply binomial series at k  1/ 2 and replace
x by  x / 4 . We obtain
1/ 2
1   1/ 2   x 
n
1 1 x
f ( x)   1      
4 x 2 4 2 n0  n   4 
  1  3 
  
1   1  x   2 
2
 2 
  x 
 1          
2   2  4  2!  4 

 1  3  5 
       x 3
 2  2  2      ...
 
3!  4
 1  3  5   1  
       
...   n  1  x n 
 2  2  2   2      ...
 
n!  4 

1  x 1 3x 2 1  3  5  (2n  1) x n 
 1   2
 ...  n
 ...
2  8 2!8 n!8 
where  x / 4  1, i.e., x  4 .
Thus its radius of convergence is 4 .
 17

Exercise 3.2
For each of the problem 1-8, find the Taylor’s series
of the function f about the given point x = a and
specify its radius of convergence.

For each of the problem 9-14, find the Macluarin’s

series of the function f and its radius of con-
vergence by using the well-known Macluarin’s
series including differentiation and integration of the
series.
 18

For each of the problem 15-22, apply any techniques

to find the Macluarin’s series of the function f and
its radius of convergence.

18. f ( x)  sin x
2
( Hint: )

-------------------------

Answer to exercise 3.2

19