Sunteți pe pagina 1din 17

Numerical Sequences and Series

Written by Men-Gen Tsai


email: b89902089@ntu.edu.tw
1. Prove that the convergence of {s
n
} implies convergence of {|s
n
|}. Is
the converse true?
Solution: Since {s
n
} is convergent, for any > 0, there exists N such
that |s
n
s| < whenever n N. By Exercise 1.13 I know that
||s
n
| |s|| |s
n
s|. Thus, ||s
n
| |s|| < , that is, {s
n
} is convergent.
The converse is not true. Consider s
n
= (1)
n
.
2. Calculate lim
n
(

n
2
+ n n).
Solution:

n
2
+ n n =
n

n
2
+ n + n
=
1
_
1/n + 1 + 1

1
2
as n .
3. If s
n
=

2 and
s
n+1
=
_
2 +

s
n
(n = 1, 2, 3, ...),
prove that {s
n
} converges, and that s
n
< 2 for n = 1, 2, 3, ....
Proof: First, I show that {s
n
} is strictly increasing. It is trivial that
s
2
=
_
2 +

s
1
=
_
2 +
_

2 >

2 = s
1
. Suppose s
k
> s
k1
when
1
k < n. By the induction hypothesis,
s
n
=
_
2 +

s
n1
>
_
2 +

s
n2
= s
n1
By the induction, {s
n
} is strictly increasing. Next, I show that {s
n
} is
bounded by 2. Similarly, I apply the induction again. Hence {s
n
} is
strictly increasing and bounded, that is, {s
n
} converges.
4.
5.
6.
7. Prove that the convergence of

a
n
implies the convergence of

a
n
n
if a
n
0.
Proof: By Cauchys inequality,
k

n=1
a
n
k

n=1
1
n
2

k

n=1
a
n

a
n
n
for all n N. Also, both

a
n
and

1
n
2
are convergent; thus

k
n=1
a
n

a
n
n
is bounded. Besides,

a
n
n
0 for all n. Hence

a
n
n
is convergent.
8.
9. Find the radius of convergence of each of the following power series:
(a)

n
3
z
n
, (b)

2
n
n!
z
n
, (c)

2
n
n
2
z
n
, (d)

n
3
3
n
z
n
.
2
Solution: (a)
n
= (n
3
)
1/n
1 as n . Hence R = 1/ = 1.
(b)
n
= (2
n
/n!)
1/n
= 2/(n!)
1/n
0 as n . Hence R = +.
(c)
n
= (2
n
/n
2
)
1/n
2/1 = 2 as n . Hence R = 1/ = 1/2.
(d)
n
= (n
3
/3
n
)
1/n
1/3 as n . Hence R = 1/ = 3.
10.
11. Suppose a
n
> 0, s
n
= a
1
+ ... + a
n
, and

a
n
diverges.
(a) Prove that

a
n
1+a
n
diverges.
(b) Prove that
a
N+1
s
N+1
+ ... +
a
N+k
s
N+k
1
s
N
s
N+k
and deduce that

a
n
s
n
diverges.
(c) Prove that
a
n
s
2
n

1
s
n1

1
s
n
and deduce that

a
n
s
2
n
converges.
(d) What can be said about

a
n
1 + na
n
and

a
n
1 + n
2
a
n
?
Proof of (a): Note that
a
n
1 + a
n
0
1
1
a
n
+ 1
0

1
a
n

a
n
0
as n . If

a
n
1+a
n
converges, then a
n
0 as n . Thus for
some

= 1 there is an N
1
such that a
n
< 1 whenever n N
1
. Since

a
n
1+a
n
converges, for any > 0 there is an N
2
such that
a
m
1 + a
m
+ ... +
a
n
1 + a
n
<
3
all n > m N
2
. Take N = max(N
1
, N
2
). Thus
>
a
m
1 + a
m
+ ... +
a
n
1 + a
n
>
a
m
1 + 1
+ ... +
a
n
1 + 1
=
a
m
+ ... + a
n
2
for all n > m N. Thus
a
m
+ ... + a
n
< 2
for all n > m N. It is a contradiction. Hence

a
n
1+a
n
diverges.
Proof of (b):
a
N+1
s
N+1
+ ... +
a
N+k
s
N+k

a
N+1
s
N+k
+ ... +
a
N+k
s
N+k
=
a
N+1
+ ... + a
N+k
s
N+k
=
s
N+k
s
N
s
N+k
= 1
s
N
s
N+k
If

a
n
s
n
converges, for any > 0 there exists N such that
a
m
s
m
+ ... +
a
n
s
n
<
for all m, n whenever n > m N. Fix m = N and let n = N + k.
Thus
>
a
m
s
m
+ ... +
a
n
s
n
=
a
N
s
N
+ ... +
a
N+k
s
N+k
1
s
N
s
N+k
4
for all k N. But s
N+k
as k since

a
n
diverges and
a
n
> 0. Take = 1/2 and we obtain a contradiction. Hence

a
n
s
n
diverges.
Proof of (c):
s
n1
s
n

1
s
2
n

1
s
n
s
n1

a
n
s
2
n

a
n
s
n
s
n1
=
s
n
s
n1
s
n
s
n1

a
n
s
2
n

1
s
n1

1
s
n
for all n.
Hence
k

n=2
a
n
s
2
n

n=2
(
1
s
n1

1
s
n
)
=
1
s
1

1
s
n
.
Note that
1
s
n
0 as n since

a
n
diverges. Hence

a
n
s
2
n
converges.
Proof of (d):

a
n
1+na
n
may converge or diverge, and

a
n
1+n
2
a
n
con-
verges. To see this, we put a
n
= 1/n.
a
n
1+na
n
=
1
2n
, that is,

a
n
1+na
n
=
2

1/n diverges. Besides, if we put


a
n
=
1
n(log n)
p
where p > 1 and n 2, then
a
n
1 + na
n
=
1
n(log n)
2p
((log n)
p
+ 1)
<
1
2n(log n)
3p
for large enough n. By Theorem 3.25 and Theorem 3.29,

a
n
1+na
n
converges. Next,

a
n
1 + n
2
a
n
=

1
1/a
n
+ n
2
5
<

1
n
2
.
for all a
n
. Note that

1
n
2
converges, and thus

a
n
1+n
2
a
n
converges.
12. Suppose a
n
> 0 and

a
n
converges. Put
r
n
=

m=n
a
m
.
(a) Prove that
a
m
r
m
+ ... +
a
n
r
n
> 1
r
n
r
m
if m < n, and deduce that

a
n
r
n
diverges.
(b) Prove that
a
n

r
n
< 2(

r
n

r
n+1
)
and deduce that

a
n

r
n
converges.
Proof of (a):
a
m
r
m
+ ... +
a
n
r
n
>
a
m
+ ... + a
n
r
m
=
r
m
r
n
r
m
= 1
r
n
r
m
if m < n. If

a
n
r
n
converges, for any > 0 there exists N such that
a
m
r
m
+ ... +
a
n
r
n
<
for all m, n whenever n > m N. Fix m = N. Thus
a
m
r
m
+ ... +
a
n
r
n
> 1
r
n
r
m
= 1
r
n
r
N
6
for all n > N. But r
n
0 as n ; thus
a
m
r
m
+... +
a
n
r
n
1 as n .
If we take = 1/2, we will get a contradiction.
Proof of (b): Note that
r
n+1
< r
n


r
n+1
<

r
n


r
n
+

r
n+1
< 2

r
n

r
n
+

r
n+1

r
n
< 2
(

r
n

r
n+1
)

r
n
+

r
n+1

r
n
< 2(

r
n

r
n+1
)

r
n
r
n+1

r
n
< 2(

r
n

r
n+1
)

a
n

r
n
< 2(

r
n

r
n+1
)
since a
n
> 0 for all n.
Hence,
k

n=1
a
n

r
n
<
k

n=1
2(

r
n

r
n+1
)
= 2(

r
1

r
k+1
)
Note that r
n
0 as n . Thus

a
n

r
n
is bounded. Hence

a
n

r
n
converges.
Note: If we say

a
n
converges faster than

b
n
, it means that
lim
n
a
n
b
n
= 0.
According the above exercise, we can construct a faster convergent se-
ries from a known convergent one easily. It implies that there is no
perfect tests to test all convergences of the series from a known con-
vergent one.
7
13. Prove that the Cauchy product of two absolutely convergent series con-
verges absolutely.
Note: Given

a
n
and

b
n
, we put c
n
=

n
k=0
a
k
b
nk
and call

c
n
the Cauchy product of the two given series.
Proof: Put A
n
=

n
k=0
|a
k
|, B
n
=

n
k=0
|b
k
|, C
n
=

n
k=0
|c
k
|. Then
C
n
= |a
0
b
0
| +|a
0
b
1
+ a
1
b
0
| + ... +|a
0
b
n
+ a
1
b
n1
+ ... + a
n
b
0
|
|a
0
||b
0
| + (|a
0
||b
1
| +|a
1
||b
0
|) + ...
+(|a
0
||b
n
| +|a
1
||b
n1
| + ... +|a
n
||b
0
|)
= |a
0
|B
n
+|a
1
|B
n1
+ ... +|a
n
|B
0
|a
0
|B
n
+|a
1
|B
n
+ ... +|a
n
|B
n
= (|a
0
| +|a
1
| + ... +|a
n
|)B
n
= A
n
B
n
AB
where A = limA
n
and B = limB
n
. Hence {C
n
} is bounded. Note
that {C
n
} is increasing, and thus C
n
is a convergent sequence, that
is, the Cauchy product of two absolutely convergent series converges
absolutely.
14. If {a
n
} is a complex sequence, dene its arithmetic means
n
by

n
=
s
0
+ s
1
+ ... + s
n
n + 1
(n = 0, 1, 2, ...).
(a) If lims
n
= s, prove that lim
n
= s.
(b) Construct a sequence {s
n
} which does not converge, although lim
n
=
0.
(c) Can it happen that s
n
> 0 for all n and that limsup s
n
= ,
although lim
n
= 0?
8
(d) Put a
n
= s
n
s
n1
, for n 1. Show that
s
n

n
=
1
n + 1
n

k=1
ka
k
.
Assume that lim(na
n
) = 0 and that {
n
} converges. Prove that {s
n
}
converges. [This gives a converse of (a), but under the additional as-
sumption that na
n
0.]
(e) Derive the last conclusion from a weaker hypothesis: Assume M <
, |na
n
| M for all n, and lim
n
= . Prove that lims
n
= , by
completing the following outline:
If m < n, then
s
n

n
=
m + 1
n m
(
n

m
) +
1
n m
n

i=m+1
(s
n
s
i
).
For these i,
|s
n
s
i
|
(n i)M
i + 1

(n m1)M
m + 2
.
Fix > 0 and associate with each n the integer m that satises
m
n
1 +
< m + 1.
Then (m + 1)/(n m) 1/ and |s
n
s
i
| < M. Hence
limsup
n
|s
n
| M.
Since was arbitrary, lims
n
= .
Proof of (a): The proof is straightforward. Let t
n
= s
n
s,
n
=
n
s.
(Or you may suppose that s = 0.) Then

n
=
t
0
+ t
1
+ ... + t
n
n + 1
.
9
Choose M > 0 such that |t
n
| M for all n. Given > 0, choose N
so that n > N implies |t
n
| < . Taking n > N in
n
= (t
0
+ t
1
+ ... +
t
n
)/(n + 1), and then
|
n
|
|t
0
| + ... +|t
N
|
n + 1
+
|t
N+1
+ ... +|t
n
n + 1
<
(N + 1)M
n + 1
+ .
Hence, limsup
n
|
n
| . Since is arbitrary, it follows that lim
n
|
n
| =
0, that is, lim
n
= s.
Proof of (b): Let s
n
= (1)
n
. Hence |
n
| 1/(n + 1), that is,
lim
n
= 0. However, lims
n
does not exists.
Proof of (c): Let
s
n
=
_

_
1 , n = 0,
n
1/4
+ n
1
, n = k
2
for some integer k,
n
1
, otherwise.
It is obvious that s
n
> 0 and limsup s
n
= . Also,
s
0
+ ... + s
n
= 1 + nn
1
+
_

n
_
n
1/4
= 2 +
_

n
_
n
1/4
.
That is,

n
=
2
n + 1
+

n n
1/4
n + 1
The rst term 2/(n + 1) 0 as n . Note that
0

n n
1/4
n + 1
< n
1/2
n
1/4
n
1
= n
1/4
.
It implies that the last term 0. Hence, lim
n
= 0.
10
Proof of (d):
n

k=1
ka
k
=
n

k=1
k(s
k
s
k1
) =
n

k=1
ks
k

k=1
ks
k1
=
n

k=1
ks
k

n1

k=0
(k + 1)s
k
= ns
n
+
n1

k=1
ks
k

n1

k=1
(k + 1)s
k
s
0
= ns
n

n1

k=1
s
k
s
0
= (n + 1)s
n

k=0
s
k
= (n + 1)(s
n

n
).
That is,
s
n

n
=
1
n + 1
n

k=1
ka
k
.
Note that {na
n
} is a complex sequence. By (a),
lim
n
_
1
n + 1
n

k=1
ka
k
_
= lim
n
na
n
= 0.
Also, lim
n
= . Hence by the previous equation, lims = .
Proof of (e): If m < n, then
n

i=m+1
(s
n
s
i
) + (m + 1)(
n

m
)
= (n m)s
n

i=m+1
s
i
+ (m + 1)(
n

m
)
= (n m)s
n

_
n

i=0
s
i

i=0
s
i
_
+ (m + 1)(
n

m
)
= (n m)s
n
(n + 1)
n
+ (m + 1)
m
+ (m + 1)(
n

m
)
= (n m)s
n
(n m)
n
.
11
Hence,
s
n

n
=
m + 1
n m
(
n

m
) +
1
n m
n

i=m+1
(s
n
s
i
).
For these i, recall a
n
= s
n
s
n1
and |na
n
| M for all n,
|s
n
s
i
| =

k=i+1
a
k

k=i+1
|a
k
|
n

k=i+1
M
i + 1
=
(n i)M
i + 1

(n (m + 1))M
(m + 1) + 1
=
(n m1)M
m + 2
.
Fix > 0 and associate with each n the integer m that satises
m
n
1 +
< m + 1.
Thus
n m
m + 1
and
n m1
m + 2
< ,
or
m + 1
n m

1

and |s
n
s
i
| < M.
Hence,
|s
n
| |
n
| +
1

(|
n
| +|
m
|) + M.
Let n and thus m too, and thus
limsup
n
|s
n
| M.
Since was arbitrary, lims
n
= .
15.
12
16. Fix a positive number . Choose x
1
>

, and dene x
2
, x
3
, x
4
, ...,
by the recursion formula
x
n+1
=
1
2
(x
n
+

x
n
).
(a) Prove that {x
n
} decreases monotonically and that limx
n
=

.
(b) Put
n
= x
n

, and show that

n+1
=

2
n
2x
n
<

2
n
2

so that, setting = 2

n+1
< (

)
2
n
(n = 1, 2, 3, ...).
(c) This is a good algorithm for computing square roots, since the
recursion formula is simple and the convergence is extremely rapid.
For example, if = 3 and x
1
= 2, show that
1
/ <
1
10
and therefore

5
< 4 10
16
,
6
< 4 10
32
.
Proof of (a):
x
n
x
n+1
= x
n

1
2
(x
n
+

x
n
)
=
1
2
(x
n


x
n
)
=
1
2
(
x
2
n

x
n
)
> 0
since x
n
> . Hence {x
n
} decreases monotonically. Also, {x
n
} is
bounded by 0; thus {x
n
} converges. Let limx
n
= x. Hence
limx
n+1
= lim
1
2
(x
n
+

x
n
) x =
1
2
(x +

x
)
x
2
= .
13
Note that x
n
> 0 for all n. Thus x =

. limx
n
=

.
Proof of (b):
x
n+1
=
1
2
(x
n
+

x
n
)
x
n+1

=
1
2
(x
n
+

x
n
)

x
n+1

=
1
2
x
2
n
2x
n

+
x
n
x
n+1

=
(x
n

)
2
2x
n

n+1
=

2
n
2x
n
<

2
n
2

.
Hence

n+1
< (

)
2
n
where = 2

by induction.
Proof of (c):

=
2

3
2

3
=
1
2

3(2 +

3)
=
1
6 + 4

3
<
1
10
.
Thus

5
< (

)
2
4
< 2

3 10
16
< 4 10
16
,

6
< (

)
2
5
< 2

3 10
32
< 4 10
32
.
Note: It is an application of Newtons method. Let f(x) = x
2
in
Exercise 5.25.
17.
18.
14
19.
20.
21.
22. Suppose X is a complete metric space, and {G
n
} is a sequence of dense
open subsets of X. Prove Baires theorem, namely, that

1
G
n
is not
empty. (In fact, it is dense in X.) Hint: Find a shrinking sequence of
neighborhoods E
n
such that E
n
G
n
, and apply Exercise 21.
Proof: Ive proved it in Chapter 2 Exercise 30.
23. Suppose {p
n
} and {q
n
} are Cauchy sequences in a metric space X.
Show that the sequence {d(p
n
, q
n
)} converges. Hint: For any m, n,
d(p
n
, q
n
) d(p
n
, p
m
) + d(p
m
, q
m
) + d(q
m
, q
n
);
it follows that
|d(p
n
, q
n
) d(p
m
, q
m
)|
is small if m and n are large.
Proof: For any > 0, there exists N such that d(p
n
, p
m
) < and
d(q
m
, q
n
) < whenever m, n N. Note that
d(p
n
, q
n
) d(p
n
, p
m
) + d(p
m
, q
m
) + d(q
m
, q
n
).
It follows that
|d(p
n
, q
n
) d(p
m
, q
m
)| d(p
n
, p
m
) + d(q
m
, q
n
) < 2.
Thus {d(p
n
, q
n
)} is a Cauchy sequence in X. Hence {d(p
n
, q
n
)} con-
verges.
15
24. Let X be a metric space. (a) Call two Cauchy sequences {p
n
}, {q
n
} in
X equivalent if
lim
n
d(p
n
, q
n
) = 0.
Prove that this is an equivalence relation.
(b) Let X

be the set of all equivalence classes so obtained. If P X

,
Q X

, {p
n
} P, {q
n
} Q, dene
(P, Q) = lim
n
d(p
n
, q
n
);
by Exercise 23, this limit exists. Show that the number (P, Q) is
unchanged if {p
n
} and {q
n
} are replaced by equivalent sequences, and
hence that is a distance function in X

.
(c) Prove that the resulting metric space X

is complete.
Proof of (a): Suppose there are three Cauchy sequences {p
n
}, {q
n
},
and {r
n
}. First, d(p
n
, p
n
) = 0 for all n. Hence, d(p
n
, p
n
) = 0 as n .
Thus it is reexive. Next, d(q
n
, p
n
) = d(p
n
, q
n
) 0 as n . Thus it
is symmetric. Finally, if d(p
n
, q
n
) 0 as n and if d(q
n
, r
n
) 0
as n , d(p
n
, r
n
) d(p
n
, q
n
) + d(q
n
, r
n
) 0 + 0 = 0 as n .
Thus it is transitive. Hence this is an equivalence relation.
Proof of (b):
Proof of (c): Let {P
n
} be a Cauchy sequence in (X

, ). We wish to
show that there is a point P X

such that (P
n
, P) 0 as n .
For each P
n
, there is a Cauchy sequence in X, denoted {Q
k
n}, such
that (P
n
, Q
k
n) 0 as k . Let
n
> 0 be a sequence tending
to 0 as n . From the double sequence {Q
k
n} we can extract a
subsequence Q

n
such that (P
n
, Q

n
) <
n
for all n. From the triangle
inequality, it follows that
(Q

n
, Q

m
) (Q

n
, P
n
) +(P
n
, P
m
) +(P
m
, Q

m
). (1)
16
Since {P
n
} is a Cauchy sequence, given > 0, there is an N > 0 such
that (P
n
, P
m
) < for m, n > N. We choose m and n so large that

m
< ,
n
< . Thus (1) shows that {Q

n
} is a Cauchy sequence in X.
Let P be the corresponding equivalence class in S. Since
(P, P
n
) (P, Q

n
) +(Q

n
, P
n
) < 2
for n > N, we conclude that P
n
P as n . That is, X

is
complete.
25.
17

S-ar putea să vă placă și