Homework Set 12: Alberto Rendon Vargas // Jose Daniel Ambriz Hernández

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

HOMEWORK SET 12

Alberto Rendon Vargas // Jose Daniel Ambriz Hernández

Noviembre 2021

Problem 12.1 Let K : [0, 1] × [0, 1] → R be a continuous function. Show that


if f : [0, 1] → R is Riemann-integrable, the function
Z 1
g(x) = K(x, y)f (y)dy
0
makes sense and is in fact continuous.
Solution
We have a continuous K : [0, 1]×[0, 1] → R which is continuous. For each fixed
x ∈ [0, 1], the function y → K(x, y) is thus a continuous function from [0, 1] →
R, hence is Riemann integrable. Since f : [0, 1] → R is Riemann integrable
by assumption and the product of two integrable functions is integrable, we
conclude that
Z 1
g(x) = K(x, y)f (y)dy
0
Is well-defined. We will show that it is continuous. Let  > 0. Pick M ∈ R
sufficiently large that
Z 1
|f (y)|dy < M
0

K is a continuous function on the compact set [0, 1] × [0, 1], hence is unifor-
mally continuous. Thus, there exists δ > 0 such that
p 
(x0 − x)2 + (y 0 − y)2 < δ ⇒ |K(x0 , y 0 ) − K(x, y)| <
M
For {x, x0 , y, y 0 } ⊂ [0, 1]. So that,

|x0 − x| < δ ⇒ |K(x0 , y) − K(x, y)| <
M
Thus, for |x0 − x| < δ, we have
Z 1 Z 1
|g(x0 ) − g(x)| = | K(x0 , y)f (y)dy − K(x, y)f (y)dy|
0 0
Z 1 Z
0
=| (K(x , y) − K(x, y))f (y)dy| ≤ |K(x0 , y) − K(x, y)| · |f (y)|dy
0
Z 1

< |f (y)|dy < 
M 0

Which proves the (uniform) continuity of g.


Problem 12.2 Prove that if f, g : [a, b] → R are Riemann-Stieltjes integrable
(for some α) , then so is the function máx(f, g) .
Solution

1
Let a, b ∈ R. We claim that máx(a, b) = (a + b + |a − b|)/2. So, we have three
cases here:
a = b: Then (a + b + |a − b|)/2 = a = máx(a, b) .
a > b: Then (a + b + |a − b|)/2 = (a + b + a − b)/2 = a = máx(a, b) .
a < b: Then (a + b + |a − b|)/2 = (a + b + b − a)/2 = b = máx(a, b) .
Now suppose f and g are Riemann-Stieltjes integrable for some α. Then f − g
is integrable, hence so is |f − g|, hence so is (f + g + |f − g|)/2 = máx(f, g) .
Problem 12.3 (i) Prove that if f : [a, b] → R is a continuous function and
Z b
not everywhere zero, then f (x)2 dx > 0.
a Z b
(ii) Using that, prove that if f : [a, b] → R is a continuous function and xn f (x)dx =
a
0 for all n>0, then f is everywhere zero.
Solution
i) Suppose f : [a, b] → R is continuous. Then f 2 ∈ R. By assumption there
exists x0 ∈ [a, b] with f (x0 ) 6= 0. Because f 2 is continuous at x0 there exists a
δ > 0 such that

x ∈ [a, b], |x − x0 | < δ ⇒ |f 2 (x) − f 2 (x0 )| < |f 2 (x0 )|/2.


Thus

x ∈ [a, b], |x − x0 | < δ ⇒ |f 2 (x)| > |f 2 (x0 )|/2.


By definition
Z b n
X
f 2 (x)dx = sup L(P, f 2 ) and L(P, f 2 ) = mi 4xi .
a i=1
If we choose a partition P with two points in [a, b]∩(x0 −δ, x0 +δ) then at least
one of the mi ’s is strictly greater than zero and the rest are all greater
Z than or
equal to zero. Thus for this partition L(P, f ) > 0 and we obtain f 2 > 0.

Problem 12.4 Give the proofs to the following statements:


R∞
1. ζ = s 1 x[x]
s+1 dx. Hint: Compute the difference between the integral over

[1,N] and the N th partial sum of the series that defines ζ(s)
Solution
Let b a real number such that b>2, and let N be a integer such that N≤
b < N+1. Then N ≥ 2 and we see that
Z b Z N
[x] [x]
s s+1
dx ≥ s dx
1 x 1 xs+1
N −1 Z k+1
X x
=s s+1
dx
k x
k=1

2
N −1 Z k+1
X k
≥s dx
k xs+1
k=1
N
X −1 Z k+1
=s k x−s−1 dx
k=1 k

N −1
X (k + 1)−s − k −s
=s k
−s
k=1
N −1  
X 1 1
= k s
− (1)
k (k + 1)s
k=1

On the other hand,


Z b Z N +1
[x] [x]
s s+1
dx ≤ s dx
1 x 1 xs+1
N Z k+1
X [x]
=s s+1
dx
k x
k=1
N Z k+1
X k+1
≤s dx
k xs+1
k=1
N
X Z k+1
=s (k + 1) x−s−1 dx
k=1 k

N
X (k + 1)−s − k −s
=s (k + 1)
−s
k=1
N  
X 1 1
= (k + 1) −
ks (k + 1)s
k=1
N   XN  
X 1 1 1 1
= k − + −
ks (k + 1)s ks (k + 1)s
k=1 k=1
N    
X 1 1 1
= k − + 1− (2)
ks (k + 1)s (N + 1)s
k=1

By combining (1) and (2) we obtain

N   Z b
X 1 1 [x]
k − ≤ s dx (3)
ks (k + 1)s 1 x s+1
k=1

3
N    
X 1 1 1
≤ k − + 1 −
ks (k + 1)s (N + 1)s
k=1

For every number b > 2 and for the natural number N such that N ≤ b
< N+1,(i.e.N=[b]) and so N ≥ 2.
Conversely, for every natural number N > 2, we can find a real number
B such that N ≤ b < N+1, and so (3) Holds.
R∞
s
2. ζ(s) = s−1 − s 1 x−[x]
xs+1 dx,Where [x] denotes the greatest integer ≤ x.
Solution
We now assume that
Z ∞
[x]
ζ(s) = s dx (4)
1 xs+1

holds.Then
∞ ∞ ∞
x − [x]
Z Z Z
s s 1 [x]
−s dx = −s dx + s dx
s−1 1 xs+1 s−1 1 xs 1 xs+1
Z b
s 1
= − s lı́m dx + ζ(s) [using(4)]
s−1 b→∞ 1 xs
s b−s+1 − 1
= − s lı́m + +ζ(s)
s−1 b→∞ −s + 1

s lı́mb→∞ b−s+1 − 1
= −s + ζ(s)
s−1 −s − 1
s 0−1
= −s + ζ(s) [because s > 1]
s−1 −s + 1
= ζ(s)

Problem 12.5 Change of variables in a Riemann-Stieltjes integral


Theorem: Suppose that ϕ is strictly increasing and continuos, and maps
|A, B| to |a.b|. Then if f:|a, b|=⇒ R is RS-integrable for some α,g=f(ϕ) such
that |A, B| =⇒ R is RS-integrable for β = α(ϕ) and
Z b Z B
f dα = gdβ
a A

Proof To each partionP = {x0 , ..., xn } of [a,b] correspons a partion Q =


y0 , ...yn of [A,B], so that xi = ϕ(yi ). All partions of [A,B] are obtained in this
way.Since the values taken by f on [xi−1 , x − i] are exactly the same as those
taken by g on [yi−1 , y − i], then

U (Q, g, β) = U (P, f, α), L(Q, g, β) = L(P, f, α)

4
The fact that f is RS-integrable onR [a,b] impies that ∀ > 0 and given P, such
that both U (P, f, α) are close to f dα.
Now, the equalities in the above expressions
R and the partition Q of [A,B] imply
U (Q, g, β) and L(Q, g, β are close to gdβ, hence = f (ϕ) : [A,B] =⇒ R i.e.,
RS-integrable thus
Z B Z b
gdδ = f dα
A a

You might also like