Midterm exam Analysis II, 26-03-2021 Solutions

1. ( 4
y 0 (x) + x3 y = 2
x2
+ ex ,
y(1) = 1.
The ODE is the first order Linear type, which can be solved by using the integrating
factor:
R R 3
R dx 3
µ(x) = e p(x)dx
=e x
dx
= e3 x = e3 ln|x| = eln |x| = |x|3 .
Hence,  
3 3
0
0
y + y x3 = y 0 x3 + yx3 = y 0 x3 + 3yx2 = yx3 .
x x

0 2 4 4
yx3 = x3 ( 2 + ex ) = 2x + x3 ex .
x
Then, by taking the integral of the lats equation we obtain the general solution:
4
ex
yx3 = x2 +


+ C,
4
4
1 ex C
y(x) = + 3 + 3.
x 4x x
Substitution of the initial value to the general solution we get
e
1 = y(1) = 1 + + C.
4
So the solution of this initial value problem is
4
1 ex −e
y(x) = + 3 + 3.
x 4x 4x

2. First solve the homogeneous equation. Substitution of y(x) = erx gives the auxiliary
equation r2 − 6r + 13 = 0, with the solutions r = 3 ± 2i. Hence, the general solution
of the homogeneous equation is

y(x) = C1 e3x cos(2x) + C2 e3x sin(2x), C1 , C2 ∈ R.

A particular solution will be of the form yp (x) = Ax + B + Cex . Then we have

yp0 (x) = A + Cex and yp00 (x) = Cex . Substitution in the nonhomogeneous equation
yields
8C = 1, −6A + 13B = 0, 13A = 1.
1 6
, C = 18 . The general solution for this inhomogeneous diffe-

So, A = 13 , B = 169
rential equation is therefore:
1 6 1
y(x) = C1 e3x cos(2x) + C2 e3x sin(2x) + x+ + ex , C1 , C2 ∈ R.
13 169 8

1
3. a) First consider
∞ ∞
X (n2 + 4n + 2) X (n2 + 4n + 2)
|(−1)n n n
|=
2 e 2n e n
n=1 n=1

Using the ratio test one obtains:

((n + 1)2 + 4n + 6) (n2 + 4n + 2) 1

lim n+1 n+1
: n n
= < 1.
n→∞ 2 e 2 e 2e
Hence, the series converges absolutely.
b)
∞
X ln(n)
(−1)n .
n
n=3

First consider,
∞
X ln(n)
.
n
n=3

For all n ≥ 3 the following holds true:

ln(n) 1
≥ .
n n
P∞ ∞
1 P ln(n)
Since n=3 n diverges, then n also diverges. So there is no absolute
n=3
convergence. Now apply the alternating series test: (i) the series is alternating,
(ii) the general term tends to 0 and (iii) the sequence ln(n)
n is decreasing for all
ln(x) 0
n ≥ 3. For the third item define g(x) = x so g (x) = 1−ln x2
x
≤ 0. (Note that
ln(x) ≥ 1 on the interval [e, ∞) ).
So the series is convergent, but not absolutely convergent, therefore it is condi-
tionally convergent.

4. a) Th pointwise limit is f (x) = x. Since,

 
1
lim x sin(x) + x = x = f (x).
n→∞ n

Now, we will prove uniform convergence. Let ε > 0 be given. Then, for all
n ≥ n∗ we find that for each x ∈ [a, b]:
   
1  1  M
|fn (x) − f (x)| =  x sin(x) + x − x =  x sin(x) ≤
   < ε,
n n n

where M = max{|a|, |b|}, Hence, for all n ≥ n∗ > M

ε the sequence converges
uniformly on the interval [a, b].
b) The point wise limit is f (x) = x, but the sequence is not uniformly convergent
on R. For any n ∈ N, we can always find x ∈ R large enough to guarantee that
1
n |x| sin(x)| > 1 by taking x, such that |x|| sin(x)| > n. Thus for ε = 1 and any
n ∈ N there exists x ∈ R such that

|fn (xn ) − f (xn )| ≥ 1,

2
 so the sequence {fn } does not converge uniformly on R. For instance, set
xn = 2nπ + π2 then one has the following
 π 
|fn (xn ) − f (xn )| = 2π +  ≥ 1, for every n ∈ N.

2n

5.
∞
X x
(n + x2 )2
n=0
−3 x2 +n
Define fn (x) = x
(n+x2 )2
for all n ≥ 0. Then, fn0 (x) = (x2 +n)3
. So fn0 (x) = 0
√ √ √ 5
for xp= ± 33n . Since fn00 ( 3n
3 ) = − 2732 3 n− 2 then fn (x) has its maximum at
x = n3 . Therefore,
√
3 3 −3
|fn (x)| ≤ Mn = n 2.
16
P∞ 3 √3 − 3
Now, we apply the Weierstrass M-test. Since ∞
P
n=1 M n = n=1 16 n
2 is

p-series with p > 1, then the series converges. Weierstrass M-test concludes
∞
P x
that (n+x2 )2
is uniform convergent on R.
n=0

(3x + 1)4n
6. a) Set an = . Using the ratio test one has
n 16n
 an+1  |3x + 1|4 |3x + 1|4
 
n
lim  = lim = .
n→∞ an  16 n→∞ n + 1 16
4 3
If |3x+1|
16 < 1, so if −1 < x < 13 , the series converges absolutely. If |2x+1|
8 > 1,
so if x < −1 or x > 13 , the series diverges. In the next step, we study the
convergence of series at the end points: x = −1 and x = 13 .
∞
P 1
x = −1 gives n , so divergent (p-series with p = 1).
n=1
∞
1 P 1
x= 3 gives n, so divergent (p-series with p = 1).
n=1
The interval of convergence therefore is (−1, 31 ).
b) Let
∞
X (3x + 1)4n
f (x) = .
n 16n
n=1

From part a) we know that f is differentiable on (−1, 13 ) by differentiating of

the series we get term by term:
∞ ∞
X 12n(3x + 1)4n−1 X 12(3x + 1)4n−1
f 0 (x) = = .
n 16n 16n
n=1 n=1

Now put x = 0 to obtain:

∞ ∞  n
X 12 X 1 1/16 12
f 0 (0) = = 12 = 12 × = .
16n 16 1 − (1/16) 15
n=1 n=1