4.2.

Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 1

Section 4.2. Lebesgue Integration of a Bounded

Measurable Function over a Set of Finite Measure

Note. In Sections 4.2 through 4.4, we develop Lebesgue integration of measurable

functions. We do so by defining the Lebesgue integral for a class of functions
and then use that class of functions to define the Lebesgue integral for the next
(larger) class of functions. In this section, we start with the class of simple functions
(think of this as “Class 0”). We then use integrals of simple functions to define the
Lebesgue integral of bounded measurable functions over sets of finite measure (think
of this as “Class 1”). In Class 1, we have that everything with which we compute is
bounded (namely, function values are bounded and measures of sets are bounded).
In Section 4.3 we consider nonnegative functions (functions in “Class 2”). We define
the Lebesgue integral of Class 2 functions in terms of integrals of Class 1 functions.
In Class 2, we may encounter unbounded integrals (i.e., integrals which are infinite),
but we will not encounter any sort of ∞ − ∞ situation since nothing is negative
here. In Section 4.4 we consider integrable functions (that is, functions that have a
positive part and a negative part, both of which are nonnegative by definition, and
both of which have a finite Lebesgue integral); these are the functions in “Class 3.”
The positive and negative parts of these functions are in Class 2 and have finite
integrals. We then define the integral of a Class 3 function as the integral of the
positive part minus the integral of the negative part (since both of these integrals
in finite, we have avoided an ∞ − ∞ situation).
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 2

Note. In Section 4.1 we saw that we could define the Riemann integral of a
bounded function on a closed and bounded interval in terms of lower and upper
Riemann integrals:
Z b  Z b 
(R) f = sup (R) ϕ | ϕ is a step function and ϕ ≤ f on [a, b]
a a

and
Z b  Z b 
(R) f = inf (R) ϕ | ϕ is a step function and ϕ ≥ f on [a, b] ,
a a

respectively. When the upper and lower Riemann integrals are equal, the Riemann
integral is the common value of these two integrals. We now follow a similar pro-
cedure in the development of the Lebesgue integral by replacing the step functions
of Riemann with simple function here. Therefore, simple functions play the same
role in Lebesgue integration as step functions play in Riemann integration.

Definition. For simple function ψ on a set of finite measure E with canonical

representation ψ = ni=1 ai χEi , define the (Lebesgue) integral
P

Z Xn
ψ= ai m(Ei ).
E i=1

Note. We have defined the Lebesgue integral in terms of the canonical represen-
tation. In the next result we give the value of the Lebesgue integral of simple ϕ in
the event that the representation of ϕ may not be the canonical one.
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 3

Lemma 4.1. Let {Ei }ni=1 be a finite disjoint collection of measurable subsets of a
set of finite measure E. For 1 ≤ i ≤ n, let ai be a real number. If ϕ = ni=1 ai χEi
P

on E then E ϕ = ni=1 ai m(Ei ).

R P

Note. The following is the first time (of four times; once for each “Class” of
functions) we’ll see a linearity and monotonicity result.

Proposition 4.2. Linearity and Monotonicity of Integration.

Let ϕ and ψ be simple functions defined on a set of finite measure E. Then for any
α, β
Z Z Z
(αϕ + βψ) = α ϕ+β ψ
E E E
R R
and if ϕ ≤ ψ on E then Eϕ ≤ E ψ.

Note. We are ready to define the lower and upper Lebesgue integrals for Class
1 functions. Notice that the definition is very similar to the definition of the
lower and upper Riemann integral of a bounded function on a closed and bounded
interval, but that simple functions are used here whereas step functions were used
for Riemann integrals.

Definition. Let f be a bounded function on a set of finite measure. Define the

lower Lebesgue integral
Z Z 
f = sup ϕ ϕ is simple, ϕ ≤ f .
E E
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 4

Define the upper Lebesgue integral

Z Z 
f = inf ψ ψ is simple, ψ ≥ f .
E E

R R
Note. By monotonicity, since f is bounded, both E f and Ef are finite. Also by
R R
monotonicity E f ≤ E f.

Definition. A bounded function f on a domain E of finite measure is said to be

R R
Lebesgue integrable over E provided E f = E f. The common value is the Lebesgue
R
integral of f over E, denoted E f .

Note. We’ll see below that all Class 1 functions are Lebesgue integrable (see
Theorem 4.4).

Note. Recall from the Riemann-Lebesgue Theorem (Theorem 6-11 in the Riemann-
Lebesgue Theorem supplement) that a bounded function on [a, b] is Riemann in-
tegrable if and only if f is continuous almost everywhere. By Proposition 3.3 a
real-valued function that is continuous on its measurable domain is a measurable
function, and by Proposition 3.5(i) if f is measurable on E and f = g a.e. on
E then g is measurable on E. So a bounded function that is continuous a.e. on
[a, b] is measurable and so both the Riemann integral and the Lebesgue integral
are defined for such a function. We now show that these integrals have the same
values, as expected.
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 5

Theorem 4.3. Let f be a bounded function defined on [a, b]. If f is Riemann

integrable over [a, b] then it is Lebesgue integrable over [a, b] and the two integrals
are equal.

Note. The following result establishes the Lebesgue integrability of the class of
functions studied in this section (the “Class 1” functions).

Theorem 4.4. Let f be a bounded measurable function on a set of finite measure

E. Then f is integrable on E.

Note. The converse of Theorem 4.4 is also true. That is, if bounded f on set of
R
finite measure E is integrable (i.e., E f exists), then f is measurable (see Theorem
5.7 in 5.3. Characterization of Riemann and Lebesgue Integrability).

Note. We now have our second encounter with a linearity and monotonicity result.
Proposition 4.2 addressed linearity and monotonicity for simple functions (i.e.,
Class 0 functions) and the next result addresses bounded functions on sets of finite
measure (i.e., Class 1 functions).
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 6

Theorem 4.5. Linearity and Monotonicity.

Let f and g be bounded measurable functions on a set of finite measure E. Then
for all α, β ∈ R
Z Z Z
(αf + βg) = α f +β g.
E E E
R R
Moreover, if f ≤ g on E, then E f ≤ E g.

Note. Linearity and and monotonicity on Class 1 can be used to prove the following
two corollaries.

Corollary 4.6. Let f be a bounded measurable function on a set E of finite

measure. Suppose A and B are measurable disjoint subsets of E. Then
Z Z Z
f= f+ f.
A∪·B A B

Corollary 4.7. Let f be a bounded measurable function on a set of finite measure

E. Then
Z Z
f ≤ |f |.
E E

Note. Recall that a sequence {fn } of Riemann integrable functions on [a, b] which
converges uniformly to f on [a, b] satisfies
 Z b  Z b  Z b
lim (R) fn (x) dx = (R) lim fn (x) dx = (R) f (x) dx
n→∞ a a n→∞ a

(see Theorem 8-3 of the Riemann-Lebesgue Theorem handout). The next proposi-
tion shows that a similar result holds for Class 1 functions and Lebesgue integration.
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 7

Proposition 4.8. Let {fn } be a sequence of bounded measurable functions on a

set of finite measure on E. If {fn } → f uniformly on E, then
Z  Z   Z
lim fn = lim fn = f.
n→∞ E E n→∞ E

R R R
Note. We are interested in general in when Ef = E (lim fn ) = lim( E fn ) for

a sequence {fn } of measurable functions on set E which converges to measurable

function f pointwise on E. When this result holds, we have a “convergence the-
orem.” We will have three such results, one for each of Class 1, Class 2, and
Class 3 of functions. The next note shows that for Class 1 functions we need some
additional hypothesis on the sequence {fn }.

Note 4.2.A. Define fn on [0, 1] as:


 0 if x = 0 or x ∈ [1/n, 1]
fn (x) =
 n if x ∈ (0, 1/n).
Z
Then fn → f ≡ 0 (pointwise but not uniformly) and fn = 1 for all n ∈ N.
[0,1]
However, Z  Z   Z
lim fn = 1 6= lim fn = 0 = 0.
n→∞ [0,1] [0,1] n→∞ [0,1]
4.2. Lebesgue Integration of a Bounded Measurable Function over a Set of Finite Measure 8

Note. To prove our first convergence theorem, we need Egoroff’s Theorem:

Egoroff ’s Theorem. Assume E has finite measure. Let {fn } be a
sequence of measurable functions on E that converges pointwise on E
to real-valued function f . Then for every ε > 0, there is a closed set
F ⊂ E for which {fn } → f uniformly on F and m(E \ F ) < ε.
See 3.3. Littlewood’s Three Principles, Egoroff’s Theorem, and Lusin’s Theorem.

Note. To deal with the example given in Note 4.2.A, we impose a condition of
“uniform pointwise boundedness” which insures that each function in the sequence
{fn } is bounded by the same constant. Notice that this is not the case for the
example given in Note 4.2.A.

Bounded Convergence Theorem.

Let {fn } be a sequence of measurable functions on a set of finite measure E.
Suppose {fn } is uniformly pointwise bounded on E, that is, there is a number
M ≥ 0 for which |fn | ≤ M on E for all n. If {fn } → f pointwise on E, then
Z  Z   Z
lim fn = lim fn = f.
n→∞ E E n→∞ E

Revised: 11/2/2020