Skip to Content
Math4121Introduction to Lebesgue Integration (Lecture 35)

Math4121 Lecture 35

Continue on Lebesgue Integration

Lebesgue Integration

Definition of Lebesgue Integral

For simple functions ϕ=i=1naiχSi\phi = \sum_{i=1}^{n} a_i \chi_{S_i}, given a measure EE, the Lebesgue integral is defined as:

Rnϕdm=i=1naim(SiE)\int_{\mathbb{R}^n} \phi \, dm = \sum_{i=1}^{n} a_i m(S_i\cap E)

Given a non-negative measurable function ff and a measurable set EE.

Define Efdm=sup{Eϕdm:ϕ is a simple function and ϕf}\int_E f \, dm = \sup \left\{ \int_E \phi \, dm : \phi \text{ is a simple function and } \phi \leq f \right\}

(We do allows Efdm=\int_E f \, dm = \infty)

For general measurable function ff, we can define f(x)=max{0,f(x)}f^-(x)=\max\{0,-f(x)\}, f+(x)=max{0,f(x)}f^+(x)=\max\{0,f(x)\}. (The positive part of the function and the negative part of the function, both non-negative)

Then f=f+ff=f^+-f^-.

We say ff is integrable if Ef+dm<\int_E f^+ \, dm < \infty and Efdm<\int_E f^- \, dm < \infty. (both finite) If at least one is finite, define

Efdm=Ef+dmEfdm\int_E f \, dm = \int_E f^+ \, dm - \int_E f^- \, dm

We allow for A=A-\infty = -\infty and A+=A+\infty = \infty for any ARA\in \mathbb{R}. But not \infty-\infty.

Immediate Properties of Lebesgue Integral

If ff is measurable and m(E)=0m(E)=0, then Efdm=0\int_E f \, dm = 0.

If E=E1E2E=E_1\cup E_2 and E1E2=E_1\cap E_2=\emptyset, then Efdm=E1fdm+E2fdm\int_E f \, dm = \int_{E_1} f \, dm + \int_{E_2} f \, dm.

Corollary

If fgf\leq g almost everywhere, (fgf\leq g except for a set of measure 0), then EfdmEgdm\int_E f \, dm \leq \int_E g \, dm.

Proof:

Let F={xE:f(x)>g(x)}F=\{x\in E: f(x)>g(x)\}. Then m(F)=0m(F)=0.

Efdm=EFfdm+FfdmEFgdm+Egdm=Egdm\begin{aligned} \int_E f \, dm &= \int_{E\setminus F} f \, dm + \int_F f \, dm\\ &\leq \int_{E\setminus F} g \, dm+\int_E g \, dm\\ &= \int_E g \, dm \end{aligned}

QED

Proposition 6.13

If ff is non-negative and Efdm=0\int_E f \, dm =0, then f=0f=0 almost everywhere on EE, f(x)=0f(x)=0 xEF\forall x\in E\setminus F, where m(F)=0m(F)=0.

Proof:

Let En={xE:f(x)1n}E_n=\{x\in E: f(x)\geq \frac{1}{n}\}. Then 1nχEn(x)f(x)\frac{1}{n}\chi_{E_n}(x)\leq f(x) for all xEx\in E.

By definition 1nm(En)=E1nχEndmEfdm=0\frac{1}{n}m(E_n)=\int_E \frac{1}{n}\chi_{E_n} \, dm \leq \int_E f \, dm =0.

Therefore, m(En)=0m(E_n)=0 for all nn.

Now U={xE:f(x)>0}=n=1EnU=\{x\in E: f(x)>0\}=\bigcup_{n=1}^{\infty} E_n, and EnEn+1E_n\subseteq E_{n+1} for all nn.

Therefore, m(U)=m(n=1En)=limnm(En)=0m(U)=m(\bigcup_{n=1}^{\infty} E_n)=\lim_{n\to\infty} m(E_n)=0.

QED

Convergence Theorems

When does limnEfndm=Elimnfndm\lim_{n\to\infty} \int_E f_n \, dm = \int_E \lim_{n\to\infty} f_n \, dm?

Theorem 6.14 Monotone Convergence Theorem

Let {fn}\{f_n\} be a monotone increasing sequence of measurable functions on EE and fnff_n\to f almost everywhere on EE. (fn(x)fn+1(x)f_n(x)\leq f_{n+1}(x) for all xEx\in E and nn)

If there exists A>0A>0 such that EfndmA\left|\int_E f_n \, dm\right|\leq A for all nNn\in \mathbb{N}, then f(x)=limnfn(x)f(x)=\lim_{n\to\infty} f_n(x) exists for almost every xEx\in E and it is integrable on EE and

Efdm=limnEfndm\int_E f \, dm = \lim_{n\to\infty} \int_E f_n \, dm

Proof:

First to show the limit exists almost everywhere. It suffices to show

U={xE:fn(x) is unbounded}\mathcal{U}=\{x\in E: f_n(x) \text{ is unbounded}\}

has measure 0.

Let ϵ>0\epsilon>0 and write

U=n=1EnU=\bigcup_{n=1}^{\infty} E_n

where En={xE:fn(x)ϵ}E_n=\{x\in E: |f_n(x)|\geq \epsilon\}.

Then UUU\subseteq \mathcal{U} and m(U)<ϵm(U)<\epsilon.

CONTINUE NEXT TIME.

QED

Last updated on