HOMEWORK #7 - MA 504

PAULINHO TCHATCHATCHA

P
ChapterP3, problem 8. If an converges, and if {bn } is monotonic and bounded,
prove that an bn converges.

Solution.
Since {bn } is monotonic and bounded, {bn } converges by Theorem 3.14, say bn → b ∈ R.
Since {bn } is monotonic, we either bn+1 ≥ bn for all n (non-decreasing) or bn+1 ≤ bn for
all n (non-increasing). Assume, without loss of generality, that {bn } is monotonic non-
decreasing, otherwise consider −bn . If {bn } is non-decreasing, then bn ≤ b for all n. Then
consider cn = b − bnP . P
We have that since an converges, the partial sums An of an form a bounded sequence;
c0 ≥ c1 ≥ c2 ≥ · · ·, and cn →P0.
Therefore by Theorem 3.42 an bn converges.

Chapter 3, problem 9. Find the radius of convergence of each of the following power
series:
X
(a) n3 z n ,
X 2n
(b) zn,
X 2n!n
(c) zn,
n2
X n3
(d) n
zn.
3
Solution.
1 p
cn z n is R = , where α = lim sup n |cn |.
P
By Theorem 3.39, the radius of convergence of
α n→∞
cn+1 cn+1
Note that by Theorem 3.37, if lim exists, then lim = α.
n→∞ cn n→∞ cn

(a) We have by theorems 3.3 and 3.20

p √ √
α = lim sup n |n3 | = lim sup( n n)3 = (lim sup n n)3 = 1.
n→∞ n→∞ n→∞

Hence R = 1.

1
2 PAULINHO TCHATCHATCHA

(b) We have
2n+1
(n+1)! 2
lim 2n = lim = 0.
n→∞ n→∞ n
n!
Hence R = ∞.

(c) We have
2n+1
(n+1)2 2n2
lim 2n = lim = 2.
n→∞ n→∞ (n + 1)2
n2
1
Hence R = .
2
(d) We have
(n+1)3  3
n+1 1 1 1
lim 3 n3 = lim 1+ = .
n→∞ n→∞ 3 n 3
3n
Hence R = 3.
P
Chapter 3, problem 11. Suppose an > 0, sn = a1 + ... + an , and an diverges.
X an
(a) Prove that diverges.
1 + an
(b) Prove that
aN +1 aN +k sN
+···+ ≥1−
sN +1 sN +k sN +k
P an
and deduce that sn
diverges.
(c) Prove that
an 1 1
2
≤ −
sn sn−1 sn
X an
and deduce that converges.
s2n
(d) What can be said about
X an X an
and ?
1 + nan 1 + n2 an
Solution.
(a) If {an } is not bounded, then an → ∞. Then
 
an 1
lim 1 − = lim = 0.
n→∞ 1 + an n→∞ 1 + an

an X an
Hence → 1 and diverges by theorem 3.23.
1 + an 1 + an
So assume that {an } is bounded. We have then that there exists M > 0 such that an ≤ M .
 HOMEWORK #7 - MA 504 3

1 1
Then ≥ , and
1 + an 1+M
X an X an 1 X
≥ = an .
1 + an 1+M 1+M
P X an
Since an diverges, by theorem 3.25 diverges.
1 + an
(b) Since an > 0, sn+1 > sn for all n. Then
aN +1 aN +k aN +1 aN +k aN +1 + · · ·aN +k sN +k − sN sN
+···+ ≥ +···+ = = =1− .
sN +1 sN +k sN +k sN +k sN +k sN +k sN +k
an
Suppose that converges. Then given 0 <  < 1, there exists N ∈ N such that
sn
aN +1 aN +k sN
> +···+ ≥1− ,
sN +1 sN +k sN +k
for all P
k.
Since an diverges, {sn } is not bounded, so sn → ∞. Hence since in the inequality above
P an
k is arbitrary, if we let k → ∞, we get  ≥ 1 a contradiction with  < 1. Therefore sn
diverges.

(c) Similarly as before, since sn > sn−1 for all n ≥ 1, we have

an an sn − sn−1 1 1
≤ = = − .
s2n sn−1 sn sn−1 sn sn−1 sn
Hence
N N  
X an X 1 1 1 1 1
2
≤1+ − =1+ − ≤1+ .
s
n=0 n n=1
sn−1 sn a0 s N a0
N
PN an
X an
Then we get that all the partial sums of n=0 s2n are bounde by 1+1/a0 , so converges.
n=0
s2n
(d) We have that
an an 1
2
≤ 2 = 2.
1 + n an n an n
X 1 X an
Since 2
converges, converges.
n 1 + n 2 an
1
For an = , we have that
n
X an X 1
= diverges.
1 + nan 2n
Now consider an = 1 when n = 2k for some k ∈ N, and an = 2−n otherwise. Then
X X∞
an ≥ 1 diverges,
k=0
4 PAULINHO TCHATCHATCHA

but
∞ ∞
X an X 1 X 1
≤ + coverges.
1 + nan k=0
1 + 2k n=1 2n + n
an
Therefore might converge or diverge.
1 + nan
Chapter 3, problem 13. Prove that the Cauchy product of two absolutely convergent
series converges absolutely.

Solution. n
P P X
Let an and bn be absolutely convergent series. Consider cn = ak bn−k . We want to
P k=0
show that cn converges absolutely.
Indeed we have
N
X n
XX
|cn | ≤ |ak ||bn−k | = |a0 |BN +|a1 |BN −1 +···+|aN |B0 ≤ (|a0 |+···|aN |)BN = AN BN ,
n=0 k=0
Pn Pn
whereP An = k=0 P |a k |, B n = k=0 |bk |.
Since an and bn P are absolutely convergent, {An } and {Bn } are bounded. Therefore by
the inequality above |cn | converges.

Problem A. Prove Theorem 4.4 using: 1) The definition of limit; 2) Theorems 4.2 and 3.3.

Solution.
Suppose E ⊂ X, a metric space, p is a limit point of E, f and g are complex functions on
E, and
lim f (x) = A, lim g(x) = B.
x→p x→p

First we want to show

lim (f + g)(x) = A + B.
x→p

Indeed, given  > 0, there exists δ1 > 0 such that

|f (x) − A| < , if |x − p| < δ1 .
Similarly, there exists δ2 > 0 such that
|g(x) − B| < , if |x − p| < δ2 .
Let δ = min{δ1 , δ2 }. Then
|(f + g)(x) − (A + B)| ≤ |f (x) − A| + |g(x) − B| < 2 if |x − p| < δ.
Since  > 0 is arbitrary, we have that
lim (f + g)(x) = A + B.
x→p
 HOMEWORK #7 - MA 504 5

Or take a sequence {pn } such that pn → p with pn 6= p.

Then by theorem 4.2
lim f (pn ) = A, lim g(pn ) = B.
n→∞ n→∞
So by theorem 3.3
lim (f + g)(pn ) = lim f (pn ) + lim g(pn ) = A + B.
n→∞ n→∞ n→∞

Since {pn } is an arbitrary sequence such that pn → p with pn 6= p, it follows from theorem
4.2 that
lim (f + g)(x) = A + B.
x→p

Now we want to show

lim (f g)(x) = AB.
x→p

Indeed, let  > 0 be given. Similarly as before we can take δ > 0 such that
|f (x) − A| <  and |g(x) − B| <  if |x − p| < δ.
We have then
|(f g)(x)−AB| = |(f (x)−A)(g(x)−B)+A(g(x)−B)+B(f (x)−A)| ≤ |f (x)−A||g(x)−B|+|A||g(x)−B|+|B|
Since  > 0 is arbitrary, we can take 1 >  > 0 as small as we want, so that ( + |A| + |B|) <
(1 + |A| + |B|) can be as small as we want. So
lim (f g)(x) = AB.
x→p

Or take take a sequence {pn } such that pn → p with pn 6= p.

Then by theorem 4.2
lim f (pn ) = A, lim g(pn ) = B.
n→∞ n→∞
So by theorem 3.3
lim (f g)(pn ) = ( lim f (pn ))( lim g(pn )) = AB.
n→∞ n→∞ n→∞

As, before since {pn } is an arbitrary sequence such that pn → p with pn 6= p, it follows from
theorem 4.2 that
lim (f g)(x) = AB.
x→p

Finally we want to show  

f A
lim (x) = , if B 6= 0.
x→p g B
Indeed, let  > 0 be given. Similarly as before we can take δ > 0 such that
|f (x) − A| <  and |g(x) − B| <  if |x − p| < δ.
Since B 6= 0, we might assume  < |B|/2. Then
|B|
|g(x)| ≥ |B| − |g(x) − B| > |B| −  > if |x − p| < δ.
2
6 PAULINHO TCHATCHATCHA

So    
 1 1   B − g(x)  2 2
 g(x) − B  =  Bg(x)  ≤ |B|2 |B − g(x)| < |B|2 , |x − p| < δ.
   
      
 f (x) A   1 1 1  |f (x) − A|  1 1 
 − = (f (x) − A) + A − ≤ + |A|  − 
 g(x) B   g(x) g(x) B  |g(x)|  g(x) B 
 
 f (x) A  2 2|A|
⇒  −  ≤ + .
g(x) B |B| |B 2 |
Since  > 0 is arbitrary,  
f A
lim (x) = .
x→p g B
Or take take a sequence {pn } such that pn → p with pn 6= p.
Then by theorem 4.2
lim f (pn ) = A, lim g(pn ) = B.
n→∞ n→∞
So by theorem 3.3  
f limn→∞ f (pn ) A
lim (pn ) = = .
n→∞ g limn→∞ g(pn ) B
As, before since {pn } is an arbitrary sequence such that pn → p with pn 6= p, it follows from
theorem 4.2 that  
f A
lim (x) = .
x→p g B