Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Assignment II

REAL ANALYSIS II


June 12, 2016

Submitted to

Dr. Legesse Lemecha

Name
ID No
1.
Miliyon Tilahun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GSR/1401/08

Real Analysis Problems


1. Prove that the limit exists and find its value
Z 1
1 + nx2
lim
log(2 + cos(x/n))dx
n 0 (1 + x2 )n
Solution. Let
fn =

1 + nx2
log(2 + cos(x/n))
(1 + x2 )n

Now Apply LDCT


2. Let a > 0. Prove that
Z

lim

n2 xen x
dx
1 + x2

exists and find its value. Justify your argument.


Solution. Let
2

n2 xen x
fn =
1 + x2

Now

2 n2 x2

n xe

n2 x




=
|fn | =
2 x2

2
2
n
1+x
(1 + x )e



n2 x
x

(1 + x2 )(1 + n2 x2 ) x



n2 x2
1

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

:= g(x)
x(1 + x2 )
Use Partial fraction to evaluate the integral of g(x)
3. Let g : R R be integrable and f : R R be bounded, measurable and continuous at 1. Prove that
Z n
x
lim
f 1 + 2 g(x)dx
n n
n
exists and determine its value.
Solution. Let
hn = [n,n] (x)f (1 +

x
)g(x)
n2

Now we apply Dominated Convergence theorem


4. Show that if f (x) = ln(1/x) for x (0, 1], then f belongs to Lp (0, 1] for all 1 p but doesnt
belongs to L (0, 1].
Solution. Clearly f (x) = ln(1/x) is not in L (0, 1] since it is not bounded near 0 and, in particular
m{x (0, 1] : f (x) > M } = m{x [0, 1) : x < eM } > min{1, eM } > 0,

for every M > 0, so it cannot be essentially bounded. To see that it is in Lp (0, 1] for 1 p < , we
note that for x [0, 1] we have the elementary bound
0 ln(x) 1/x.
Whereby, using the properties of the log we conclude that for any > 0,
0 ln(x)
Therefore
Z

| ln(1/x)|p dx =

1
.
x

| ln(x)|p p

xp dx.

Since for any 1 p < , we can choose always < 1/p, the right-hand-side is integrable, and therefore
f Lp (0, 1].
5. Assume that E has a finite measure and 1 p1 p2 . Show that if {fn } f in Lp2 (E), then
{fn } f in Lp1 (E).
Solution. Since E has finite measure, we know that if f fn is in Lp2 (E), then there exists a constant
c such that
kf fn kp1 ckf fn kp2 .
The proof follows trivially from this.
6. For 1 p find the value the parameter for which
Z
1
lim
f = 0, for all f Lp ([0, 1])
0+ 0
Solution. It suffices to consider f non-negative, since we may always consider f + and f separately.
We consider the case p < first. Suppose that 1/q, then by Holders inequality,
Z 1/p
Z
1
1/q
f

fp
.
0
0
Since f is in Lp , we have

f p = 0.

lim

Thus,
1
lim
0

f = 0.
0

Now suppose > 1/q and choose f = x1 , note that since p(1) > 1, then f Lp [0, 1]. However
Z
1
x1 = 1,
0
and therefore does not vanish as goes to 0.
When p = , we see that when < 1,
Z
1
f 1 kf k
0
which goes to 0 as 0. When = 1, let f = 1, then
Z
1
f = 1,
0
and does not go to zero as 0.
3

7. Suppose that f 0 on (0, ), f Lp (0, ) and

Prove that for 1 < p <

(F (x)) dx
0

Hint: Write xF (x) =

Rx
0

1
F (x) =
x

f (t)dt
0

p
p1

F (x)p dx =

1
xp

Now use the hint, and write

f (t)ta ta dt

= 1 p1 . We apply Holders Inequality to this expression to get that

Z
f (t)dt

 p1 Z
f (t) t dt

1
xp
1

x q a

(1 aq) q

(1 aq) q

f (t)p tap

Z

x q pap dx =


f (t) t dt dx


p
x q pap dx dt.

x1pa dx = tpa
Z

 p1
f (t) t dt .
p ap

F (x) dx

p ap

Z

1
p+q ,

 q1
dt =

p
f (t)dt dx
0
Z
Z
p
pap
xq

Z

(1 aq) q 0
Z
1

and so we have

aq

Now note that

p ap

Using this estimate we find


Z
Z
F (x)p dx =

Choose now a =

p
f (t)dt dx.

Z

(f (t))p dt

Z
f (t)dt =

1
q

f (t)ta ta dt and use Holders inequality.

Solution. We start with

where 0 < a <

p Z

ap(1 aq)

p
q

1
,
ap

f (t)p dt.

which is found by solving a minimization problem, and then simple algebra gives


1
p

ap(1 aq) q


p
.
p1

8. Let f, {fk } Lp where 1 p < . Show that if kf fk kp 0, then kfk kp kf kp . Conversely if


fk f a.e and kfk kp kf kp , show that kf fk kp 0.





Solution. By Minkowskis inequality, kf kp kfk kp kf fk kp , so kf fk kp 0 implies kfk kp





kf kp . (If p < 1, we have kf kp kfk kp ckf fk kp for some c > 1, so the argument still holds.) For
the converse, since p 1, we have
|f fk |p 2p1 (|f |p + |fk |p ),
4

so by Fatous lemma and the fact that fk f a.e., we have


Z
Z
p
p
2 |f | = lim inf 2p1 (|f |p + |fk |p ) |f fk |p
k
Z
Z
lim inf 2p1 (|f |p + |fk |p ) + lim inf |f fk |p
k
k
Z
Z
p1
p
p
= lim 2 (|f | + |fk | ) lim sup |f fk |p
k
k
Z
Z
p
p
p
= 2 |f | lim sup |f fk | ,
k

where the last Requality follows by the dominated convergence theorem since |f |p and |fk |p are in L1 .
Thus, lim supk |f fk |p 0, so kf fk kp 0.
9. Let (X, M, ) be a -finite measure space and let be another -finite measure on (X, M) such that
 . Show that for any measurable function f : X [0, +] we have
Z
Z
d
f d =
f d
d
X
X
Solution. The equality holds for f = A for any measurable set A by definition of the Radon-Nikodym
derivative. Hence it holds for all simple positive measurable functions. Then it holds for all positive
measurable functions by the monotone convergence theorem.
10. Let 1 , 2 and be measures on a measure space (X, M, ). Show that if 1  and 2  then
(1 + 2 )  .
Solution. Let E be a measurable set in M such that (E) = 0. Since 1  and 2  ,
1 (E) = 0 = 2 (E). Thus, (1 + 2 )(E) = 1 (E) + 2 (E) = 0. Therefore, (1 + 2 )  .
11. Let be a finite measure and be a -finite measure on the same measurable space (X, ) such that
for all E M, (E) (E)
a) Show that  .
b) Show that there is a function f L1 () such that for all E M
Z
(E) =
f d
E

c) Show that the function in (b) must satisfy 0 f (x) 1 for a.e x X.
Solution. For (a) simply use the definition of absolute continuity. For (b), read up on the LesbsegueRadon-Nikodym theorem. For (c), assume f > 1 on a set of measure non-zero, and then assume f < 0
on a set of measure non-zero, and these will lead to contradictions.
12. Use Fubinis theorem to prove that
Z

lim

sin x

dx =
x
2

Solution. I believe this can also be solved using double integrals.


It is possible (if I remember correctly) to justify switching the order of integration to give the equality:
!
!
Z
Z
Z
Z

exy sin x dy dx =

exy sin x dx dy

Notice that

sin x
x

exy sin x dy =

This leads us to
Z

 sin x 
x

dx =

xy

sin x dx dy

Now the right hand side can be found easily, using integration by parts.
Z
Z
xy
xy
I= e
sin x dx = e
cos x y exy cos x dx
Z


= exy cos x y exy sin x + y exy sin x dx
=

yexy sin x exy cos x


.
1 + y2

Thus

1
1 + y2

exy sin x dx =

Thus

 sin x 
x

dx =
0

1
dy = .
2
1+y
2

13. Find the value of the following iterated integrals


Z

Z
2

2x yex yy dy dx
[0,)

[0,)

Solution.
14. Prove that
Z
0

Z
0

x2 y 2
log(4 + sin x)dydx =
(x2 + y 2 )3/2

Z
0

Z
0

x2 y 2
log(4 + sin x)dxdy
(x2 + y 2 )3/2

Solution.
15. Compute
Z

y sin(x)exy dxdy

where E = (0, ) (0, 1) and explain why Fubinis theorem is applicable.

Solution. Since the integrand is a continuous function, it is measurable. We have by integration by


parts
Z
F (y) =
y sin(x)exy dx
0

y
y2 + 1

Thus,
1

F (y)dy =
0
xy

Now, since |f (x, y)| ye

1
log 2.
2

, we may apply Fubini I to see that


Z
Z
|f (x, y)|dxdy
yexy dxdy
E

E
1

Z
=

yexy dx

dy
0

Z
=

0
1

dy
0

= 1.
Doing integration with respect to y first yields


Z 1
sin(x) 1 ex
ex .
y sin(x)exy dy =
x
x
0
Thus, Fubinis theorem shows that
Z
0



sin(x) 1 ex
1
x
e
dx = log 2
x
x
2

16. Show that

ey

Hint: Take f (x, y) = sin(2xy)e

log 5
sin2 (y)
dy =
y
4

and apply Fubinis theorem.

Solution. Consider
Z

sin2 y
y

sin(2xy)dx =
0

Then
Z
0

ey

sin2 (y)
dy =
y

ey

sin(2xy)dxdy
0

Now let f (x, y) = sin(2xy)ey . Then


|f (x, y)| ey

1
1 + y2

You might also like