Вы находитесь на странице: 1из 10

2

Infinite Series

2.1 Definitions & convergence

Definition 2.1.1. Let {a n } be a sequence of real numbers.

a) An expression of the form

is called an infinite series.

a 1 + a 2 +

+ a n +

b) The number a n is called as the n th term of the series.

c) The sequence {s n }, defined by s n =

n

k=1

a k , is called the sequence of partial sums of the series.

d) If the sequence of partial sums converges to a limit L, we say that the series converges and its sum is L.

e) If the sequence of partial sums does not converge, we say that the series diverges.

Examples 2.1.2.

1)

If 0 < x < 1, then n=0 x n converges to

1

1 x .

Solution. Let us consider the sequence of partial sums {s n }, where s n =

n

k=1

x k . Here

s n =

n

k=1

x k = 1 x n+1 1 x

=

1

1 x

x

n+1

1 x ,

n N.

As, 0 < x < 1, x n+1 0 as n → ∞. Hence s n

1

1 x .

1

1 x . Thus the given series converges to

///

2) The series

1

n=1

n diverges.

Solution. Consider the sequence of partial sums {s n }, where s n =

the subsequence s 2 n of {s n }. Here

n

k=1

k 1 . Now, let us examine

s

s

2

4

=

=

1 + 1/2 = 3/2,

1 + 1/2 + 1/3 + 1/4 > 3/2 + 1/4 + 1/4 = 2.

1

Suppose s 2 n > (n + 2)/2, then

2 n

k=1

1

s 2 n+1 = s 2 n +

n + 2

2

n + 2

2

2 n + k

2 n

1

2 n+1


>

+

k=1

= 2 n+1 = (n + 1) + 2

+

2

n

2

.

Thus the subsequence {s 2 n } is not bounded above and as it is also increasing, it diverges. Hence

the sequence diverges, i.e., the series

1

n=1

n diverges.

///

3) (Telescopic series:) Show that the series

1

n=1

n(n + 1) converges to 1.

Solution. Consider the sequence of partial sums {s n }. Then

s n =

n

k=1

1

k(k + 1) =

k

k=1

k k + 1 = 1 n + 1 1.

1

1

1

Summarizing this observation, one has the following theorem on Telescopic series

Theorem 2.1.3. Suppose {a n } is a sequence of non-negative real numbers such that a n L. Then the series (a n a n+1 ) converges to a 1 L.

Lemma 2.1.4.

1) If n=1 a n converges to L and n=1 b n converges to M, then the series n=1 (a n + b n ) converges to

L + M.

2) If n=1 a n converges to L and if c R, then the series n=1 ca n converges to cL.

Lemma 2.1.5. If n=1 a n converges, then lim

n a n = 0.

Proof. Suppose n=1 a n = L. Then the sequence of partial sums {s n } also converges to L. Now

a n = s n s n1 L L = 0. ///

Example 2.1.6. If x > 1, then the series n=1 x n diverges.

Solution. Assume to the contrary that the series n=1 x n converges. Then the n th term, i.e., x n 0.

But as x > 1, x n 1 for all n N and hence

n x n 1, which is a contradiction. Hence the series

n=1 x n diverges.

lim

///

2

As a first result we have the following comparison theorem:

Theorem 2.1.7. Let {a n }, {b n } be sequences of positive reals such that a n b n . If b n converges then a n converges.

partial sum of a n , b n

respectively. Then s n t n . Since b n converges, we have {t n } converges and is bounded. Now since {s n } is monotonically increasing sequence that is bounded above, we get the convergence of {s n } and

///

hence the convergence of a n .

Proof.

Let s n

=

a 1 + a 2 +

+ a n

and t n

=

b 1 + b 2 +

+ b n

be the

Theorem 2.1.8.

Let {a n }

1

be an decreasing sequence of positive numbers. Then n=1 a n converges if

and only if n=0 2 n a 2 n converges.

Proof. Let s n and t n be the sequence of partial sums of a n and 2 n a 2 n respectively. Then s n and t n are monotonically increasing sequences. We know that such sequences converge if they are bounded from above. proof follows from the observation that

s 2 n =

2 n

k=1

a n = a 1 + a 2 + (a 3 + a 4 ) + (a 5 + a 6 + a 7 + a 8 ) +

a 1 + a 2 + 2a 4 + 4a 8 + 8a 16 +

=

1

a 1 + 2 t n .

2

n1 a 2 n

+

(a 2 n1 +1 +

+

a 2 n )

(2.1)

Therefore, if {s n } converges then {s 2 n } converges and hence bounded from above. Now convergence of {t n } follows from 2.1, {t n }. On the other hand,

s 2 n 1 = a 1 + (a 2 + a 3 ) + (a 4 + a 5 + a 6 + a 7 ) + (a 8 +

a 15 ) + (a 2 n1 +

a 1 + 2a 2 + 4a 4 + 8a 8 +

+ 2 n1 a 2 n1 = a 1 + t n1

+

a 2 n 1 )

So if {t n } converges, then {s 2 n 1 } converges. Now the conclusion follows from s n s 2 n+1 1 and the

///

fact that {s n } is monotonically increasing sequence.

Examples 2.1.9.

1) Consider the series

n=1

1

n p ,

p > 0. Then, we have

p > 1 and diverges for p 1.

1

2 n

n=1

(2 n ) p =

1

n=1

(2 n )

p1 which converges for

2) Consider the series

n=2

given series diverges.

1

n log n . Here

2

n

n=2

1

2 n log 2 n

=

1

log 2

1

n=2

n which diverges.

Hence the

3

2.2

Absolute convergence

Definition 2.2.1.

a) Let n=1 a n be a series of real numbers. If n=1 |a n | converges, we say that

converges absolutely.

b) If n=1 a n converges but n=1 |a n | diverges, we say that n=1 a n converges conditionally.

Examples 2.2.2.

1) The series

(1) n n!

n=1

converges absolutely.

2) The series

n=1 (1) n

n

2

converges absolutely.

Theorem 2.2.3. If n=1 a n converges absolutely, then n=1 a n converges.

n=1 a n

n

Proof. Let t n = k=1 |a k |. As the series converges absolutely, the sequence {t n } is Cauchy. Thus, given

> 0, there exists N N such that

1

|t m t n | < m, n N.

Let m > n. Then

Thus the sequence {s n }

1

|s m s n | =

m

a

i

i=n+1

m

i=n+1

|a i | = |t m t n | < .

is Cauchy and hence converges. Thus

1

a n converges.

///

Theorem 2.2.4. Let

1

a n be a series of real numbers. Let p n = max{a n , 0} and q n = min{a n , 0}.

a) If a n converges absolutely, then both p n and q n converges.

b) If a n converges conditionally then both p n and q n diverges.

Proof.

a) Observe that p n = (a n + |a n |)/2 and q n = (a n − |a n |)/2. Thus the convergence of the two series follows from the hypothesis.

b) From the observation that p n = (a n + |a n |)/2, we have |a n | = 2p n a n . As the series a n converges and |a n | diverges, the series p n diverges. Similarly, the series q n diverges. ///

Tests for absolute convergence

Theorem 2.2.5 (Comparison test). Let a n be a series of real numbers. Then, a n converges absolutely if there is an absolutely convergent series c n with |a n | ≤ |c n | for all n N, N N.

4

Proof follows as in Theorem 2.1.7

Examples 2.2.6.

1)

The series

2)

The series

n=1

n=0

7

7n

2 diverges because

7

7n 2 =

1

n! converges because

1

n!

1

2 n and

1

n 2/7

1

1

n for all n N and n diverges.

1

n=0

2 n converges.

Theorem 2.2.7 (Limit comparison test). Let {a n } and {b n } be two sequences of positive numbers. Then

a) n→∞

if

lim

n

n

a

b c > 0, a n and b n both converge or diverge together;

=

b) n→∞

if

lim

c) n→∞

if

lim

n

n

n

a 0 and b n converges absolutely, a n converges absolutely.

b n

a and |b n | diverges, a n diverges.

b

=

=

Proof. (a) As lim

n→∞

a b = c > 0, for =

n

n

c

2

> 0, there exists N N such that

 

n N

Thus, for n N,

 

c

2

or equivalently

 

cb n

2

=

a

n

b

n

c

<

a n

b

n

c

c

2

a n 3cb 2 n

.

c

2 .

Hence the conclusion follows from the comparison test.

b) Given that lim

n→∞

a b = 0. Hence for = 2 , there exists N N such that

n

n

1

n N

=

a n

b

n

<

1

2

or equivalently,

a n b n

n N

2 .

Thus the desired conclusion follows from the comparison test.

=

c) Here we are given that lim

n→∞

a b = . Hence for any real number M > 0, there exists N N

n

n

such that

 

n

N

=

or equivalently,

a n

b n

M

n N

=

a n Mb n .

Thus if |b n | diverges, then |a n | diverges by comparison test.

///

5

Examples 2.2.8.

1) Consider the series

n=1

2n + 1

(n + 1)

2 . Here a n =

2n 2 + n

1

2 as n → ∞. Further, n

n 2 + 2n + 1

given series diverges.

(n 2n + + 1) 1 2 . Let b n =

1

n . Then

a

n

b

n

=

2

2n + 1

(n + 1)

1

n

=

diverges.

Thus by limit comparison theorem, the

2)

Consider the series

1

2

n

1 1 . Here a n =

1 1 . Let b n =

2 n

1

n converges and hence the given series converges.

2

1 n . Then a n

2

b

n

=

n

2 1 1. Further,

2 n

3)

Consider the series e n .

Further,

n

2

Here a n =

e n

e − n

n

2

and b n =

1 2 .

n

Then a n

b

n

1 2 converges and hence the given series converges.

n

Theorem 2.2.9 (Ratio test). Let

1

a n be a series of real numbers. Let

= e n 0 as n → ∞.

a = lim inf

n→∞

Then

a) a n converges absolutely if A < 1;

1

b) a n diverges if a > 1;

1

a

n+1

a

n

c) the test fails if a < 1 < A.

and A = lim sup

n→∞

a

n+1

a

n

.

Proof. a) If A < 1, choose and also N N such that

B such thata A < B < 1. Then there exists an > 0 such that B = A +

a

n+1

a n

B for all n N. Further, for any k N,

a

N+k

a

N

=

k

i=1

a

N+i

a

N+i1

k

B

i=1

= B k .

Thus |a N+k | ≤ B k |a N |,

k N. But k=0 |a N |B k < as B < 1. Thus by comparison test, the series

1

a n converges.

b) If a > 1, choose b such that 1 < b < a. There exits N N such Further, for any k N,

a

N+k

a

N

=

k

i=1

a

N+i

a

N+i1

k

i=1

b = b k .

that

a

n+1

a

n

b for all n N.

6

k=0

n+1

Thus |a N+k | ≥ |a N |,

the series

1

k N. But, as b > 1,

a n diverges.

1

c) Consider the series n . Here lim

n→∞

a

a

n

converges, again lim

n→∞

a n+1

a n

= 1.

Examples 2.2.10.

a N b k diverges. Thus, again, by the comparison test,

1

1

= 1. But n diverges. For the series n 2 , which

///

a) Consider the series

n

n

n! . Here

1

 

a

n+1

= (n + 1) n+1

n! n

a

n

(n + 1)!

n

= n + 1

n

n = 1 +

n n

1

e,

which is greater than 1. So a = A = e > 1. Thus the given series diverges.

b) Consider the series

x

n

0

n! , x R. Here

a n+1

a

n

=

x

n+1

n!

(n + 1)! x n =

x

n + 1 0.

Therefore a = A = 0 < 1. Thus, for all x R, the given series converges.

Theorem 2.2.11 (Root test). Let

1

a n be a series of real numbers. Let A = lim sup |a n |. Then

n

n→∞

a) the series converges absolutely if A < 1;

b) the series diverges if A > 1;

c) the test fails if A = 1.

Proof. a) If A < 1, choose B such that A < B < 1. Then there exists N N such that

all n N. This implies |a n | < B n for all n N. As B < 1, the series converges by comparison test.

|a n | < B for

n

b) If A > 1, there exists infinitely many n N such that

n |a n | > 1. But this implies that |a n | > 1

for infinitely many values of n and hence a N 0, i.e., a n diverges.

1

c)

1

Consider the series n . Here A = 1 and the series diverges. On the other hand, for the series

///

2 , again A = 1, but the series converges.

n

Examples 2.2.12.

1) Consider the series

x n

1

n

n

, x R. Here a n = x n . Therefore,

n

x

n

n

=

series converges for |x| < 1 and diverges for |x| > 1.

x

n

n

→ |x|. Thus the

7

2) Consider the series

n x n , x R. Here a n

1

n

converges for any x R.

=

n

x

n n . Then,

|a n | =

n

x

n

0. Thus the series

3) Consider the series a n , where a n =

series converges.

n

  n is odd

4 1

n

2 n

n is even

. Then lim sup a n = 1 2 . Therefore the

n

lim sup √ a n = 1 2 . Therefore the n n →∞ Alternating series:

n→∞

Alternating series:

Definition 2.2.13. An alternating series is an infinite series whose terms alternate in sign.

Theorem 2.2.14. Suppose {a n } is a sequence of positive numbers such that

(a)

a n a n+1 for all n N and

(b)

lim

n a n = 0,

then the alternating series n=1 (1) n+1 a n converges.

Proof. Consider the partial sums with odd index, s 1 , s 3 , s 5 ,

Now, for any n N,

s 2n+1 = s 2n1 a 2n + a 2n+1 s 2n1 (by (a)).

Thus the sequence {s 2n1 }

1

forms a non-increasing sequence. Also, notice that

s 2n1 =

n1

i=1

(a 2i1 a 2i ) + a 2n1 .

Since each quantity in the parenthesis is non-negative and a 2n1 > 0, the sequence {s 2n1 } is bounded

below by 0. Hence {s 2n1 }

1

is convergent.

Now, consider the partial sums with even index, s 2 , s 4 , s 6 ,

For any n N,

s 2n+2 = s 2n + a 2n+1 a 2n+2 s 2n (by (a)).

Thus the sequence {s 2n }

1

forms a non-decreasing sequence. Further,

s 2n = a 1

n1

i=1

(a 2i a 2i+1 ) a 2n a 1 ,

which means that s 2n is bounded above by a 1 . Therefore, {s 2n } is convergent. Let L = lim s 2n and M = lim S 2n1 . By ((b)),

0 = lim

a 2n = lim (s 2n s 2n1 ) = L M.

Thus L = M and hence the alternating series n=1 (1) n+1 a n converges.

8

///

Examples 2.2.15.

1) Consider the series n=1 (1) n+1 2 1/n . Here a n = 2 1/n 1 as n → ∞. Hence the above theorem

does not apply. Anyhow, one can show that the series diverges.

2) Consider the series

(1)

n=1

n

n+1

. The a n s of this series satisfies the hypothesis of the above

theorem and hence the series converges.

Examples 2.2.16.

1) The series

2) The series

(1) n+1

n=1

n

(1) 2n1

n=1

2n 1

converges conditionally.

converges conditionally.

The following is a more genreal test than the previous theorem.

Theorem 2.2.17. (Dirichlet test) Let {a n } and {b n } be sequences of real numbers such that

1. the sequence A n =

n

k=1

a k is bounded,

2. the sequence b n is decreasing and b n 0.

Then the series a n b n converges.

Examples 2.2.18.

1) Consider the series cos log nπ n . Here take a n = cos and b n = log n . Then

1

|A n | ≤

n

k=1

| cos | ≤ 1

and b n decreases to 0. Hence the series conveges. In this case we can see that the series does not converge absolutely (apply Cauchy’s test).

2)

2 2n n 2

e n n!

To show the

boundedness of the partial sums of a n , we can apply Ratio test to see that the series a n

converges. Hence the sequence of partial sums converge and so will be bounded. Therefore by Dirichlet test the series a n b n converges.

1

1

n) 2 and a n = 2 2n n 2

e n n!

.

(log

n) 2 .

Take b n

=

(log

Then b n decreases to 0.

9

2.3

Problems

1. If the terms of the convergent series

then prove that lim

n 2 n a 2 n = 0.

n=1 a n are positive and forms a non-increasing sequence,

2. If 0 a n 1 (n 0) and if 0 x 1, then prove that n=1 a n x n converges.

3. Test the convergence of the series (1)

n=2

1

n(log n) p (2)

n=2

1

(log

n) x x

R.

4. Determine which of the following series diverges

(a)

n=1

log n

n

3/2

(b)

n=1

(log n) 2

n

3/2

(c)

n=1

n

n 2 +

1

(d)

n=1

1 n n2 n

(e)

n=1

1

n (f)

n

n

n=1

n

n

n 2 .

5. Determine which of the following series converges

(a)

n=1

n

2

2 n

(b)

n=1

n!

10

n

(c)

n=1 n 2

n

n

(d)

n=1

(log n) n

n

n

(e)

n=1

n! (2n + 1)!

6. Test the convergence of the infinite series

(1)

n=0 sin

n

π

2

(2)

n=2

1

ln(n 3 ) (3)

n

n=1

n sin n (4) n sin(

1

1

1

1

n )

10