Real Analysis HW 10 Solutions

Problem 47: Show that a function f is absolutely continuous on [a, b] if and only if for
each  > 0, there is a δ > 0 such that for every finite disjoint collection {(ak , bk )}nk=1 of open
intervals in (a, b),
 
Xn  Xn
 [f (bk ) − f (ak )] < , if [bk − ak ] < δ.
 
 
k=1 k=1

Solution: Clearly, by the triangle inequality the usual definition of absolute continuity
implies this one. It remains to show the converse. To see this, let {(ak , bk )}nk=1 be a collection
+ n+
of disjoint open intervals in [a, b], we denote by the invervals {(a+ k , bk )}k=1 the intervals such
− − n − − −
that f (b+ +
k ) − f (ak ) ≥ 0 and {(ak , bk )}k=1 the intervals such that f (bk ) − f (ak ) < 0, and let
+ −
δ and δ be such that
 +
n+ n+

Xn  X X
+ +  + +
 [f (bk ) − f (ak )] = |f (bk ) − f (ak )| < /2, if [b+ +
k − ak ] < δ
+

 
k=1 k=1 k=1

and
 −
n− −

Xn  X n
X
− − 
 [f (bk ) − f (ak )] = |f (b− −
k ) − f (ak )| < /2, if [b− − −
k − ak ] < δ .

 
k=1 k=1 k=1

Summing the two inequalities above it follows that

n
X n
X
|f (bk ) − f (ak )| <  if [bk − ak ] < min{δ + , δ − }.
k=1 k=1


Problem 48: The Cantor Lebesgue function ϕ is continuous and increasing on [0, 1]. Con-
clude from Theorem 10 that ϕ is not absolutely continuous on [0, 1]. Compare this reasoning
with that proposed in Problem 40.
Solution: The Cantor Lebesgue function has the property that ϕ0 (x) = 0 a.e. and ϕ(1) −
ϕ(0) = 1. If it were continuous, this would imply that
Z
0= ϕ0 = ϕ(1) − ϕ(0) = 1,
[0,1]

1
which is clearly a contradiction. This differs from the reasoning proposed in problem 40
since we are using here a characterization of absolute continuity in terms of the Fundamental
theorem of calculus instead of the definition of absolute continuity. 
Problem 49: Let f be continuous on [a, b] and differentiable almost everywhere on (a, b).
Show that Z b
f 0 = f (b) − f (a)
a

if and only if
Z bh i Z b 
lim Diff1/n f = lim Diff1/n f .
a n→∞ n→∞ a

Solution: Since f is continuous and differentiable a.e. on (a, b), we have that
Z b 


lim Diff1/n f = lim Av1/n f (b) − Av1/n f (a) = f (b) − f (a),
n→∞ a n→∞

and Z bh i Z b
lim Diff1/n f = f0
a n→∞ a
for almost all x. The result then follows immediately. 
Problem 52: Let f and g be absolutely continuous on [a, b]. Show that
Z b Z b
0
f · g = f (b)g(b) − f (a)g(a) − f 0 · g.
a a

Solution: Since g is absolutely continuous {Diff1/n g} is uniformly integrable and therefore

{(Diff1/n g) · f } is uniformly integrable. Thus
Z b Z b
0
f · g = lim f · Diff1/n g.
a n→∞ a

however by direct computation we see that

Z b "Z #
b Z b−1/n
f · Diff1/n g = n f (x)g(x + 1/n)dx − f (x + 1/n)g(x + 1/n)dx
a a a−1/n
Z b Z b
=n (f (x) − f (x + 1/n))g(x + 1/n)dx + n f (x + 1/n)g(x + 1/n)dx
a b−1/n
Z a
−n f (x + 1/n)g(x + 1/n)dx
a−1/n
Z b
= Av1/n (f · g)(b) − Av1/n (f · g)(a) − Diff1/n f · g.
a

2
Since f and g are continuous,

Av1/n (f · g)(b) → f (b)g(b), Av(f · g)(a) → f (a)g(a),

and f is absolutly continuous so {Diff1/n f · g} is uniformly integrable, therefore

Z b Z b
Diff1/n f · g → f 0 · g.
a a

Hence we have our result. 

Problem 55: Let f be of bounded variation on [a, b] and define v(x) = T V (f[a,x] ) for all
x ∈ [a, b].
(i) Show that |f 0 | ≤ v 0 a.e. on [a, b], and infer from this that
Z b
|f 0 | ≤ T V (f ).
a

(ii) Show that the above is an equlity if and only if f is absolutely continuous on [a, b].

(iii) Compare parts (i) and (ii) with Corollaries 4 and 12, repectively.
Solution:
(i) Suppose we take a partition P = {u, v}, with u, v ∈ [a, b]

|f (v) − f (u)| = V (f, P ) ≤ T V (f[u,v] ) = T (f[a,v] ) − T (f[a,u] )

therefore
|f (v) − f (u)| T (f[a,v] ) − T (f[a,u] )
≤ ,
v−u v−u
since f is of bounded variation, and T V (f[a,x] ) is increasing, the limits as u → v exists
pointwise a.e., so |f 0 | ≤ v 0 a.e. However,
Z b Z b
0
|f | ≤ v 0 ≤ v(b) − v(a) = T V (f ).
a a

(ii) If f is absolutely continuous, then we know that for any (u, v) ⊆ [a, b]
Z v
f 0 = f (v) − f (u).
u

Also f is of bounded variation, therefore for any partition P of [a, b],

n
X n Z b k
X
 n Z bk Z b
0
 X 0
|f 0 |.

V (f, P ) = |f (bk ) − f (ak )| = 
 f≤ |f | =
k=1 k=1 ak k=1 ak a

3
 Therefore Z b
T V (f ) ≤ |f 0 |
a
and so by the previous inequality,
Z b
T V (f ) = |f 0 |.
a

To show the coverse, suppose that we have equality. We can see from the work in part
(i) that this must imply that
Z b
v 0 = v(b) − v(a).
a

Since v is strictly increasing, this means that v is absolutely continuous on [a, b]. Let
δ > 0 be such that if {(ak , bk )} is a collection of disjoint open intervals then
n
X n
X
|T V (f[a,bk ] ) − T V (f[a,ak ] )| = T V (f[ak ,bk ] ) < 
k=1 k=1
Pn
whenever k=1 [bk − ak ] < δ. By the definition of T V (f[ak ,bk ] ) have

|f (bk ) − f (ak )| ≤ T V (f[ak ,bk ] ).

Therefore n n
X X
|f (bk ) − f (ak )| ≤ T V (f[ak ,bk ] ) < 
k=1 k=1
Pn
whenever k=1 [bk − ak ] < δ.

(iii) If f is increasing as in Corollaries 4 and 12, then we know that total variation is
T V (f ) = f (b) − f (a), and f 0 is positive, Part (i) says that
Z b
f 0 ≤ f (b) − f (a)
a

which is equivalent to Corollary 4, and part (ii) says that f is absolutely continuous if
and only if Z b
f 0 = f (b) − f (a)
a
which is equivalent to Corollary 12.


Problem 56: Let g be strictly increasing and absolutely continuous on [a, b].

4
 (i) Show that for any open subset O of (a, b),
Z
m(g(O)) = g 0 (x)dx.
O

(ii) Show that for any Gδ subset E of (a, b),

Z
m(g(E)) = g 0 (x)dx.
E

(iii) Show that for any subset E of [a, b] that has measure 0, its image g(E) also has measure
0, so that Z
m(g(E)) = 0 = g 0 (x)dx.
E

(iv) Show that for any measurable subset A of [a, b],

Z
m(g(A)) = g 0 (x)dx.
A

(v) Let c = g(a), and d = g(b). Show that for any simple function φ on [c, d],
Z d Z b
φ(y)dy = φ(g(x))g 0 (x)dx.
c a

(vi) Show that for any nonnegative integrable function f over [c, d].
Z d Z b
f (y)dy = f (g(x))g 0 (x)dx.
c a

(vii) Show that part (i) follows from (vi) in the case that f is the characteristic function of
O and the composition is defined.

Solution:
(i) Write O countable collection of disjoint open intervals, OS= ∞
S
k=1 (ak , bk ), since g
is increasing and continuous, we can also write g(O) = ∞ k=1 (g(a k ), g(bk )), where
{(g(ak ), g(bk ))} is also a countable collection of disjoint open intervals. Thus we see
∞
X
m(g(O)) = [g(bk ) − g(ak )]
k=1

which by the absolute continuity of g gives,

X∞ Z bk  Z
0
m(g(O)) = g = g0.
k=1 ak O

5
 T∞
(ii) Write a Gδ sets E as theTintersection of a countable collection of open sets E = k=1 Ok .
Since for any n, En = nk=1 Ok is open and descending, we have
Z
m(g(En )) = g0.
En

Since g is strictly increasing and continuous, we know g(En ) is also open and descend-
ing, therefore, by continuity of measure and continuity of integration
Z
m(g(E)) = g0.
E

(iii) Problem 40.

(iv) Since A is measurable, we may write it as A = E ∪ F , a union of a Gδ set E and F a
set of measure 0. Since g(E ∪ F ) = g(E) ∪ g(F ), and g(F ) is measure 0 and g(E) is
Gδ , Z Z
0
m(g(A)) = m(g(E)) = g = g0.
E A

(v)* We first show this for an indicator function χE of a measurable set E ⊂ [c, d]. That is,
we would like to show that
Z b
m(E) = (χE ◦ g) · g 0 .
a

Note, the difficulty here is that we do not know the measurability of χE ◦ g ≡ χg−1 (E) ,
since the inverse of a strictly increasing absolutely continuous function may not map
measurable sets to measurable sets. This can be seen in a counter-example by Silvia
Spataru where an absolutely continuous strictly increasing fucntion f is constructed
so that f 0 = 0 a.e. on a ‘fat’ cantor set B, and f maps B to a set of measure 0. Hence
g −1 behaves like the Cantor-Lebesgue function, mapping a set of measure 0 to a set of
positive measure. In particular this means that there f −1 maps a measure zero set to
a non-measurable set.

However, it turns out that while χE ◦g may not be measurable, (χE ◦g)·g 0 will always be
measurable. We will show this by approximation with measurable functions. Denote
h := (χE ◦ g) · g 0 and note that since g is continuous, then for any open set O and
compact K, (χO ◦ g) and (χK ◦ g) are both measurable functions. In particular this
implies by (iv) that
Z b Z b
0
m(O) = (χO ◦ g) · g , and m(K) = (χK ◦ g) · g 0 .
a a

We now choose {On } a decreasing sequence of open sets containing E, and {Kn } an
increasing sequence of compact sets contained in E so that,
m(On ) → m(E) and m(Kn ) → m(E).

6
 Define φn := (χKn ◦ g) · g 0 and ψn := (χOn ◦ g) · g 0 . We see that for every n,
φ1 ≤ . . . ≤ φn ≤ h ≤ ψn ≤ . . . ≤ ψ1 .
Since g 0 ≥ 0 a.e., {φn } and {ψn } are monotone sequences of functions bounded by g 0 .
Thus they have pointwise limits a.e. In particular since g 0 is integrable, dominated
convergence gives Z b
lim (ψn − φn ) = m(E) − m(E) = 0.
a n→∞
Since ψn − φn ≥ 0 we conclude
lim (ψn − φn ) = 0
n→∞

pointwise. Therefore
lim |h − φn | ≤ lim (ψn − φn ) = 0,
n→∞ n→∞
pointwise. So h is measurable, being the pointwise limit of measurable functions φn .
We conclude by monotone (or dominated) convergence that
Z b Z b
0
m(E) = lim (χKn ◦ g) · g = (χE ◦ g) · g 0 .
n→∞ a a

Now suppose that ϕ is a simple function over (c, d) given in canonical form by
n
X
ϕ= sk χEk .
k=1

where {Ek } is a finite disjoint collection of measurable sets such that nk=1 Ek = (c, d).
S
We find
Z g(b) Xn Xn Z b Z b
0
ϕ= sk · m(Ek ) = sk (χEk ◦ g) · g = (ϕ ◦ g) · g 0 .
g(a) k=1 k=1 a a

(vi) If f is a non-negative integrable function over [c, d] there is a monotone sequence of

simple functions ϕn that converge to f p.w. Therefore by the previous problem
Z d Z b
ϕn = (ϕn ◦ g) · g 0
c a
so by the monotone convergence theorem,
Z d Z b
f= (f ◦ g) · g 0 .
c a

(vii) Choose f to be the characteristic function of g(O), where O ⊆ (a, b). Then
Z b Z b Z
0 0
m(g(O)) = (χg(O) ◦ g)g = χO g = g0
a a O