Birkhäuser Advanced Texts Basler Lehrbücher

Series editors
Steven G. Krantz, Washington University, St. Louis, USA
Shrawan Kumar, University of North Carolina at Chapel Hill, Chapel Hill, USA
Jan Nekovář, Université Pierre et Marie Curie, Paris, France

More information about this series at http://www.springer.com/series/4842

Emmanuele DiBenedetto

Real Analysis
Second Edition
Emmanuele DiBenedetto
Nashville, TN
USA

ISSN 1019-6242 ISSN 2296-4894 (electronic)

Birkhäuser Advanced Texts Basler Lehrbücher
ISBN 978-1-4939-4003-5 ISBN 978-1-4939-4005-9 (eBook)
DOI 10.1007/978-1-4939-4005-9

Library of Congress Control Number: 2016933667

Mathematics Subject Classification (2010): 26A21, 26A45, 26A46, 26A48, 26A51, 26B05, 26B30,
28A33, 28C07, 31A15, 39B72, 46B25, 46B26, 46B50, 46C05, 46C15, 46E10, 46E25, 46E27, 46E30,
46E35, 46F05, 46F10

© Springer Science+Business Media New York 2002, 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This book is published under the trade name Birkhäuser

The registered company is Springer Science+Business Media LLC New York
(www.birkhauser-science.com)
Preface to the Second Edition

This is a revised and expanded version of my 2002 book on real analysis. Some
topics and chapters have been rewritten (i.e., Chaps. 7–10) and others have been
expanded in several directions by including new topics and, most importantly,
considerably more practice problems. Noteworthy is the collection of problems in
calculus with distributions at the end of Chap. 8. These exercises show how to solve
algebraic equations and differential equations in the sense of distributions, and how
to compute limits and series in D 0 . Distributional calculations in most texts are
limited to computing the fundamental solution of some linear partial differential
equations. We have sought to give an array of problems to show the wide appli-
cability of calculus in D 0 . I must thank U. Gianazza and V. Vespri for providing me
with most of these problems, taken from their own class notes. Chapter 9 has been
expanded to include a proof of the Riesz convolution rearrangement inequality in N-
dimensions. This is preceded by the topics on Steiner symmetrization as a supporting
background. Chapter 11 is new, and it goes more deeply in the local fine properties
of weakly differentiable functions by using the notion of p-capacity of sets in RN . It
clarifies various aspects of Sobolev embedding by means of the isoperimetric
inequality and the co-area formula (for smooth functions). It also links to measure
theory in Chaps. 3 and 4, as the p-capacity separates the role of measures versus
outer measure. In particular, while Borel sets are p-capacitable, Borel sets of positive
and finite capacity are not measurable with respect to the measure generated by the
outer measure of p-capacity. Thus, it also provides an example of nonmetric outer
measures and non-Borel measure. As it stands, this book provides a background to
more specialized fields of analysis, such as probability, harmonic analysis, functions
of bounded variation in several dimensions, partial differential equations, and
functional analysis. A brief connection to BV functions in several variables is offered
in Sect. 7.2c of the Complements of Chap. 5.
The numbering of the sections of the Problems and Complements of each
chapter follow the numbering of the section in that chapter. Exceptions are Chaps. 6
and 8. Most of the Problems and Complements of Chap. 8 are devoted to calculus
with distributions, not directly related to the sections of that chapter.

v
vi Preface to the Second Edition

Sections 20c–23c of the Complements of Chap. 6 are devoted to present the

Vitali–Saks–Hahn theorem. The relevance of the theorem is in that it gives suffi-
cient conditions on a set of integrable functions to be uniformly integrable. This in
turn it permits one to connect the notions of weak and strong convergence to
convergence in measure. In particular, as a consequence it gives necessary and
sufficient conditions for a weakly convergent sequence in L1 to be strongly con-
vergent in L1 . As an application, in ‘1 , weak and strong convergence coincide
(Sects. 22c–23c of Chap. 6).
Over the years, I have benefited from comments and suggestions from several
collaborators and colleagues including U. Gianazza, V. Vespri, U. Abdullah,
Olivier Guibé, A. Devinatzy , J. Serriny , J. Manfredi, and several current and former
students, including Naian Liao, Colin Klaus Stockdale, Jordan Nikkel, and Zach
Gaslowitz Special thanks go to Ugo Gianazza and Olivier Guibé for having read in
detail large parts of the manuscript and for pointing out imprecise statements and
providing valuable suggestions. To all of them goes my deep gratitude.
This work was partially supported by NSF grant DMS-1265548.
Preface to the First Edition

This book is a self-contained introduction to real analysis assuming only basic

notions on limits of sequences in RN , manipulations of series, their convergence
criteria, advanced differential calculus, and basic algebra of sets.
The passage from the setting in RN to abstract spaces and their topologies is
gradual. Continuous reference is made to the RN setting where most of the basic
concepts originated.
The first eight chapters contain material forming the backbone of a basic training
in real analysis. The remaining three chapters are more topical, relating to maximal
functions, functions of bounded mean oscillation, rearrangements, potential theory
and the theory of Sobolev functions. Even though the layout of the book is theo-
retical, the entire book and the last chapters in particular have in mind applications
of mathematical analysis to models of physical phenomena through partial differ-
ential equations.
The preliminaries contain a review of the notions of countable sets and related
examples. We introduce some special sets, such as the Cantor set and its variants
and examine their structure. These sets will be a reference point for a number of
examples and counterexamples in measure theory (Chapter 3) and in the Lebesgue
differentiability theory of absolute continuous functions (Chapter 5). This initial
Chapter contains a brief collection of the various notions of ordering, the Hausdorff
maximal principle, Zorn’s Lemma, the well-ordering principle, and their funda-
mental connections.
These facts keep appearing in measure theory (Vitali’s construction of a
Lebesgue non-measurable set), topological facts (Tychonov’s Theorem on the
compactness of the product of compact spaces; existence of Hamel bases) and
functional analysis (Hahn-Banach Theorem; existence of maximal orthonormal
bases in Hilbert spaces).
Chapter 2 is an introduction to those basic topological issues that hinge upon
analysis or that are, one way or another, intertwined with it. Examples include
Uhryson’s Lemma and the Tietze Extension Theorem, characterization of com-
pactness and its relation to the Bolzano-Weierstrass property, structure of the

vii
viii Preface to the First Edition

compact sets in RN , and various properties of semi-continuous functions defined on

compact sets. This analysis of compactness has in mind the structure of the compact
subsets of the space of continuous functions (Chapter 5) and the characterizations
of the compact subsets of the spaces Lp ðEÞ for all 1  p\1 (Chapter 6).
The Tychonov Theorem is proved with its application in mind in the proof of the
Alaoglu Theorem on the weak compactness of closed balls in a linear, normed space.
We introduce the notion of linear, topological vector spaces and that of linear
maps and functionals and their relation to boundedness and continuity.
The discussion turns quickly to metric spaces, their topology, and their structure.
Examples are drawn mostly from spaces of continuous or continuously differen-
tiable functions or integrable functions. The notions and characterizations of
compactness are rephrased in the context of metric spaces. This is preparatory to
characterizing the structure of compact subsets of Lp ðEÞ.
The structure of complete metric spaces is analyzed through Baire’s Category
Theorem. This plays a role in subsequent topics, such as an indirect proof of the
existence of nowhere differentiable functions (Chapter 5), in the structure of Banach
spaces (Chapter 6), and in questions of completeness and non-completeness of
various topologies on Co1 ðEÞ (Chapter 8).
Chapter 3 is a modern account of measure theory. The discussion starts from the
structure of open sets in RN as sequential coverings to construct measures and a
brief introduction to the algebra of sets. Measures are constructed from outer
measure by the Charathéodory process. The process is implemented in specific
examples such as the Lebesgue-Stiltjes measures in R and the Hausdorff measure.
The latter seldom appears in introductory textbooks in Real Analysis. We have
chosen to present it in some detail because it has become, in the past two decades,
an essential tool in studying the fine properties of solutions of partial differential
equations and systems. The Lebesgue measure in RN is introduced directly starting
from the Euclidean measure of cubes rather than regarding it, more or less
abstractly, as the N-product of the Lebesgue measure on R. In RN , we distinguish
between Borel sets and Lebesgue measurable sets, by cardinality arguments and by
concrete counterexamples.
For general measures, emphasis is put on necessary and sufficient criteria of
measurability in terms of . In this, we have in mind the operation of
measuring a set as an approximation process. From the applications point of view,
one would like to approximate the measure of a set by the measure of measurable
sets containing it and measurable sets contained into it. The notion is further
expanded in the theory of Radon measures and their regularity properties.
It is also further expanded into the covering theorems, even though these rep-
resent an independent topic in their own right. The Vitali Covering Theorem is
presented by the proof due to Banach. The Besicovitch covering is presented by
emphasizing its value for general Radon measures in RN . For both, we stress the
measure-theoretical nature of the covering as opposed to the notion of covering a
set by inclusion.
Preface to the First Edition ix

Coverings have made possible an understanding of the local properties of

solutions of partial differential equations, chiefly the Harnack inequality for
non-negative solutions of elliptic and parabolic equations. For this reason, in the
Complements of this chapter, we have included various versions of the Vitali and
Besicovitch covering theorems.
Chapter 4 introduces the Lebesgue integral. The theory is preceded by the
notions of measurable functions, convergence in measure, Egorov’s Theorem on
selecting almost everywhere convergent subsequences from sequences convergent
in measure, and Lusin’s Theorem characterizing measurability in terms of
quasi-continuity. This theorem is given relevance as it relates measurability and
local behavior of measurable functions. It is also a concrete application of the
necessary and sufficient criteria of measurability of the previous chapter.
The integral is constructed starting from non-negative simple functions by the
Lebesgue procedure. Emphasis is placed on convergence theorems and the Vitali’s
Theorem on the absolute continuity of the integral. The Peano-Jordan and Riemann
integrals are compared to the Lebesgue integral by pointing out differences and
analogies.
The theory of product measures and the related integral is developed in the
framework of the Charathéodory construction by starting from measurable rect-
angles. This construction provides a natural setting for the Fubini-Tonelli
Theorem on multiple integrals.
Applications are provided ranging from the notion of convolution, the conver-
gence of the Marcinkiewicz integral, to the interpretation of an integral in terms
of the distribution function of its integrand.
The theory of measures is completed in this chapter by introducing the notion of
signed measure and by proving Hahn’s Decomposition Theorem. This leads to
other natural notions of decompositions such as the Jordan and Lebesgue
Decomposition Theorems. It also suggests naturally other notions of comparing two
measures, such as the absolute continuity of a measure m with respect to another
measure l. It also suggests representing m, roughly speaking, as the integral of l by
the Radon-Nykodým Theorem.
Relating two measures finds application in the Besicovitch-Lebesgue Theorem,
presented in the next chapter, and connecting integrability of a function to some of
its local properties.
Chapter 5 is a collection of applications of measure theory to issues that are at
the root of modern analysis. What does it mean for a function of one real variable to
be differentiable? When can one compute an integral by the Fundamental
Theorem of Calculus? What does it mean to take the derivative on an integral?
These issues motivated a new way of measuring sets and the need for a new notion
of integral.
The discussion starts from functions of bounded variation in an interval and their
Jordan’s characterization as the difference of two monotone functions. The notion
of differentiability follows naturally from the definition of the four Dini’s numbers.
For a function of bounded variation, its Dini numbers, regarded as functions, are
measurable. This is a remarkable fact due to Banach.
x Preface to the First Edition

Functions of bounded variations are almost everywhere differentiable. This is a

celebrated theorem of Lebesgue. It uses, in an essential way, Vitali’s Covering
Theorem of Chapter 3.
We introduce the notion of absolutely continuous functions and discuss simi-
larities and differences with respect to functions of bounded variation. The
Lebesgue theory of differentiating an integral is developed in this context. A natural
related issue is that of the density of a Lebesgue measurable subset of an interval.
Almost every point of a measurable set is a density point for that set. The proof uses
a remarkable theorem of Fubini on differentiating term by term a series of monotone
functions.
Similar issues for functions of N real variables are far more delicate. We present
the theory of differentiating a measure m with respect to another l by identifying
precisely such a derivative in terms of the singular part and the absolutely con-
tinuous part of l with respect to m. The various decompositions of measures of
Chapter 4 find here their natural application, along with the Radon-Nykodým
Theorem.
The pivotal point of the theory is the Besicovitch-Lebesgue Theorem asserting
that the limit of the integral of a measurable function f when the domain of inte-
gration shrinks to a point x actually exists for almost all x and equals the value of f
at x. The shrinking procedure is achieved by using balls centered at x, and the
measure can be any Radon measure. This is the strength of the Besicovitch covering
theorem. We discuss the possibility of replacing balls with domains that are,
roughly speaking, comparable to a ball. As a consequence, almost every point of an
N-dimensional Lebesgue-measurable set is a density point for that set.
The final part of the chapter contains an array of facts of common use in real
analysis. These include basic facts on convex functions of one variable and their
almost everywhere double differentiability. A similar fact for convex functions of
several real variables (known as the Alexandrov Theorem) is beyond the scope
of these notes. In the Complements, we introduce the Legendre transform and
indicate the main properties and features.
We present the Ascoli-Arzelá Theorem, keeping in mind a description of
compact subsets of spaces of continuous functions.
We also include a theorem of Kirzbraun and Pucci extending bounded, con-
tinuous functions in a domain into bounded, continuous functions in the whole RN
with the same upper bound and the same concave modulus of continuity. This
theorem does not seem to be widely known.
The final part of the chapter contains a detailed discussion of the
Stone-Weierstrass Theorem. We present first the Weierstrass Theorem (in N
dimensions) as a pure fact of Approximation Theory. The polynomials approxi-
mating a continuous function f in the sup-norm over a compact set are constructed
explicitly by means of the Bernstein polynomials. The Stone Theorem is then
presented as a way of identifying the structure of a class of functions that can be
approximated by polynomials.
Preface to the First Edition xi

Chapter 6 introduces the theory of Lp spaces for 1  p  1. The basic

inequalities of Hölder and Minkowski are introduced and used to characterize the
norm and the related topology of these spaces. A discussion is provided to identify
elements of Lp ðEÞ as equivalence classes.
We introduce also the Lp ðEÞ spaces for 0\p\1 and the related topology. We
establish that there are not convex open sets except Lp ðEÞ itself and the empty set.
We then turn to questions of convergence in the sense of Lp ðEÞ and their
completeness (Riesz-Fisher Theorem) as well as issues of separating such spaces by
simple functions. The latter serves as a tool in the notion of weak convergence of
sequences of functions in Lp ðEÞ. Strong and weak convergence are compared and
basic facts relating weak convergence and convergence of norms are stated and
proved.
The Complements contain an extensive discussion comparing the various
notions of convergence.
We introduce the notion of functional in Lp ðEÞ and its boundedness and conti-
nuity and prove the Riesz representation Theorem, characterizing the form of all the
bounded linear functionals in Lp ðEÞ for 1  p  1. This proof is based on the
Radon-Nykodým Theorem and as such is measure theoretical in nature.
We present a second proof of the same theorem based on the topology of Lp . The
open balls that generate the topology of Lp ðEÞ are strictly convex for 1\p\1.
This fact is proved by means of the Hanner and Clarkson’s inequality, which while
technical, are of interest in their own right.
The Riesz Representation Theorem permits one to prove that if E is a
Lebesgue-measurable set in RN , then Lp ðEÞ for 1  p\1, are separable. It also
permits one to select weakly convergent subsequences from bounded ones. This
fact holds in general, reflexive, separable Banach spaces (Chapter 7). We have
chosen to present it independently as part of the Lp theory. It is our point of view
that a good part of functional analysis draws some of its key facts from concrete
spaces, such as spaces of continuous functions, the Lp , space and the spaces ‘p .
The remainder of the chapter presents some technical tools regarding Lp ðEÞ for
E, a Lebesgue-measurable set in RN , to be used in various parts of the later
chapters. These include the continuity of the translation in the topology of Lp ðEÞ,
the Friedrichs mollifyiers, and the approximation of functions in Lp ðEÞ with C 1 ðEÞ
functions. It includes also a characterization of the compact subsets in Lp ðEÞ.
Chapter 7 is an introduction to those aspects of functional analysis closely related
to the Euclidean spaces RN , the spaces of continuous functions defined on some
open set E  RN , and the spaces Lp ðEÞ. These naturally suggest the notion of finite
dimensional and infinite dimensional normed spaces. The difference between the
two is best characterized in terms of the compactness of their closed unit ball. This
is a consequence of a beautiful counterexample of Riesz.
The notions of maps and functionals is rephrased in terms of the norm topology.
In RN , one thinks of a linear functional as an affine functions whose level sets are
hyperplanes through the origin. Much of this analogy holds in general normed
spaces with the proper rephrasing.
xii Preface to the First Edition

Families of pointwise equi-bounded maps are proven to be uniformly

equi-bounded as an application of Baire’s Category Theorem.
We also briefly consider special maps such as those generated by Riesz potential
(estimates of these potentials are provided in Chapter 9), and related Fredholm
integral equations.
A proof of the classical Open Mapping Theorem and Closed Graph Theorem are
presented as a way of inverting continuous maps to identify isomorphisms out of
continuous linear maps.
The Hahn-Banach Theorem is viewed in its geometrical aspects of separating
closed convex sets in a normed space and of “drawing” tangent planes to a convex
set.
These facts all play a role in the notion of weak topology and its properties.
Mazur’s Theorem on weak and strong closure of convex sets in a normed space is
related to the weak topology of the Lp ðEÞ spaces. These provide the main examples,
as convexity is explicit through Clarkson’s inequalities.
The last part of the chapter gives an introduction to Hilbert spaces and their
geometrical aspects through the parallelogram identity. We present the Riesz
Representation Theorem of functionals through the inner product. The notion of
basis is introduced and its cardinality is related to the separability of a Hilbert space.
We introduce orthonormal systems and indicate the main properties (Bessel’s
inequality) and some construction procedures (Gram-Schmidt). The existence of a
complete system is a consequence of the Hausdorff maximum principle. We also
discuss various equivalent notions of completeness.
Chapter 8 is about spaces of real-valued continuous functions, differentiable
functions, infinitely differentiable functions with compact support in some open set
E  RN , and weakly differentiable functions.
Together with the Lp ðEÞ spaces, these are among the backbone spaces of real
analysis.
We prove the Riesz Representation Theorem for continuous functions of com-
pact support in RN . The discussion starts from positive functionals and their rep-
resentation. Radon measures are related to positive functionals and bounded, signed
Radon measures are related to bounded linear functionals. Analogous facts hold for
the space of continuous functions with compact support in some open set E  RN .
We then turn to making precise the notion of a topology for Co1 ðEÞ.
Completeness and non-completeness are related to metric topologies in a con-
structive way. We introduce the Schwartz topology and the notion of continuous
maps and functionals with respect to such a topology. This leads to the theory of
distributions and its related calculus (derivatives, convolutions etc. of distributions).
Their relation to partial differential equations is indicated through the notion of
fundamental solution. We compute the fundamental solution for the Laplace
operator also in view of its applications to potential theory (Chapter 9) and to
Sobolev inequalities (Chapter 10).
The notion of weak derivative in some open set E  RN is introduced as an
aspect of the theory distributions. We outline their main properties and state and
Preface to the First Edition xiii

prove the by now classical Meyers-Serrin Theorem. Extension theorems and

approximation by smooth functions defined in domains larger than E are provided.
This leads naturally to a discussion of the smoothness properties of @E for these
approximations and/or extensions to take place (cone property, segment property,
etc.).
We present some calculus aspects of weak derivatives (chair rule, approxima-
tions by difference quotients, etc.) and turn to a discussion of W 1;1 ðEÞ and its
relation to Lipschitz functions. For the latter, we conclude the chapter by stating and
proving the Rademaker Theorem.
Chapter 9 is a collection of topics of common use in real analysis and its
applications. First is the Wiener version of the Vitali Covering Theorem (commonly
referred to as the “simple version” of Vitali’s Theorem). This is applied to the
notion of maximal function, its properties, and its related strong type Lp estimates
for 1\p\1. Weak estimates are also proved and used in the Marcinkiewicz
Interpolation Theorem. We prove the by now classical Calderón-Zygmund
Decomposition Theorem and its applications to the space functions of bounded
mean oscillation (BMO) and the Stein-Fefferman Lp estimate for the sharp maximal
function.
The space of BMO is given some emphasis. We give the proof of the
John-Nirenberg estimate and provide its counterexample. We have in mind here the
limiting case of some potential estimates (later in the chapter) and the limiting
Sobolev embedding estimates (Chapter 10).
We introduce the notion of rearranging the values of functions and provide their
properties and the related notion of equi-measurable function. The discussion is for
functions of one real variable. Extensions to functions of N real variables are
indicated in the Complements.
The goal is to prove the Riesz convolution inequality by rearrangements. The
several proofs existing (Riesz, Zygmund, Hardy-Littlewood-Polya) all use, one way
or another, the symmetric rearrangement of an integrable function.
We have reproduced here the proof of Hardy-Littlewood-Polya as appearing in
their monograph [70]. In the process, we need to establish Hardy’s inequality, of
interest in its own right.
The Riesz convolution inequality is presented in several of its variants, leading
to an N-dimensional version of it through an application of the continuous version
of the Minkowski inequality.
Besides its intrinsic interest of these inequalities, what we have in mind here is to
recover some limiting cases of potential estimates an their related Sobolev
embedding inequalities.
The final part of the chapter introduces the Riesz potentials and their related Lp
estimates, including some limiting cases. These are on one hand based on the
previous Riesz convolution inequality, and on the other hand to Trudinger’s version
of the BMO estimates for particular functions arising as potentials.
Chapter 10 provides an array of embedding theorems for functions in Sobolev
spaces. Their importance to analysis and partial differential equations cannot be
xiv Preface to the First Edition

underscored. Although good monographs exist ([1, 104]), I have found it laborious
to extract the main facts, listed in a clean manner and ready for applications.
We start from the classical Gagliardo-Nirenberg inequalitites and proceed to
Sobolev inequalities. We have made an effort to trace, in the various embedding
inequalities, how the smoothness of the boundary enters in the estimates. For
example, whenever the cone condition is required, we trace back in the various
constant the dependence on the height and the angle of the cone. We present the
Poincaré inequalities for bounded, convex domains E, and trace the dependence
of the various constants on the “modulus of convexity” of the domain through the
ratio of the radius of the smallest ball containing E and the largest ball contained in
E. The limiting case p ¼ N of the Sobolev inequality builds of the limiting
inequalities for the Riesz potentials, and is preceded by an introduction to Morrey
spaces and their connection to BMO.
The characterization of the compact subsets of Lp ðEÞ (Chapter 6) is used to
prove Reillich’s Theorem on compact Sobolev inequalities.
We introduce the notion of trace of function in W 1;p ðRN  R þ Þ on the hyper-
plane xN þ 1 ¼ 0. Through a partition of unity and a local covering, this provides the
notion of trace of functions in W 1;p ðEÞ on the boundary @E, provided such a
boundary is sufficiently smooth. Sharp inequalities relating functions in W 1;p ðEÞ
with the integrability and regularity of their traces on @E are established in terms of
fractional Sobolev spaces. Such inequalities are first established for E being a
half-space and @E an hyperplane, and then extended to general domains E with
sufficiently smooth boundary @E. In the Complements we characterize functions f
defined and integrable on @E as traces on @E of functions in some Sobolev spaces
W 1;p ðEÞ. The relation between p and the order of integrability of f on @E is shown
to be sharp. For special geometries, such as a ball, the inequality relating the
integral of the traces and the Sobolev norm can be made explicit. This is indicated
in the Complements.
The last part of the chapter contains a newly established multiplicative Sobolev
embedding for functions in W 1;p ðEÞ that do not necessarily vanish on @E. The open
set E is required to be convex. Its value is in its applicability to the asymptotic
behavior of solutions to Neumann problems related to parabolic partial differential
equations.
Acknowledgments

These notes have grown out of the courses and topics in real analysis I have taught
over the years at Indiana University, Bloomington; Northwestern University; the
University of Rome Tor Vergata, Italy; and Vanderbilt University.
My thanks go to the numerous students who have pointed out misprints and
imprecise statements. Among them are John Renze, Ethan Pribble, Lan Yueheng,
Kamlesh Parwani, Ronnie Sadka, Marco Battaglini, Donato Gerardi, Zsolt
Macskasi, Tianhong Li, Todd Fisher, Liming Feng, Mikhail Perepelitsa, Lucas
Bergman, Derek Bruff, David Peterson, and Yulya Babenko.
Special thanks go to Michael O’Leary, David Diller, Giuseppe Tommassetti, and
Gianluca Bonuglia. The material of the last three chapters results from topical
seminars organized, over the years, with these former students.
I am indebted to Allen Devinatzy , Edward Nussbaum, Ethan Devinatz, Haskell
Rhosental, and Juan Manfredi for providing me with some counterexamples and for
helping me to make precise some facts related to unbounded linear functionals in
normed spaces.
I would like also to thank Henghui Zou, James Serriny , Avner Friedman, Craig
Evans, Robert Glassey, Herbert Amann, Enrico Magenesy , Giorgio Talenti, Gieri
Simonett, Vincenzo Vespri, and Mike Mihalik for reading, at various stages, por-
tions of the manuscript and for providing valuable critical comments.
The input of Daniele Andreucci has been crucial and it needs to be singled out.
He has read the entire manuscript and has made critical remarks and suggestions.
He has also worked out in detail a large number of the problems and suggested
some of his own. I am very much indebted to him.
These notes were conceived as a book in 1994, while teaching topics in real
analysis at the School of Engineering of the University of Rome Tor Vergata, Italy.
Special thanks go to Franco Maceri, former Dean of the School of Engineering of
that university, for his vision of mathematics as the natural language of applied
sciences and for fostering that vision.

xv
xvi Acknowledgments

I learned real analysis from Carlo Pucciy at the University of Florence, Italy in
1974–1975. My view of analysis and these notes are influenced by Pucci’s
teaching. According to his way of thinking, every theorem had to be motivated and
had to go along with examples and counterexamples, i.e., had to withstand a
scientific “critique.”
Contents

1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1 Countable Sets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 The Cantor Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3 Cardinality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.1 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Cardinality of Some Infinite Cartesian Products . . . . . . . . . . . . 5
5 Orderings, the Maximal Principle, and the Axiom of Choice . . . 7
6 Well Ordering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
6.1 The First Uncountable . . . . . . . . . . . . . . . . . . . . . . . . 9
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1c Countable Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2c The Cantor Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1c A Generalized Cantor Set of Positive Measure . . . . . . 11
2.2c A Generalized Cantor Set of Measure Zero . . . . . . . . . 12
2.3c Perfect Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3c Cardinality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Topologies and Metric Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1 Topological Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.1 Hausdorff and Normal Spaces . . . . . . . . . . . . . . . . . . 18
2 Urysohn’s Lemma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 The Tietze Extension Theorem . . . . . . . . . . . . . . . . . . . . . . . 20
4 Bases, Axioms of Countability and Product Topologies . . . . . . 22
4.1 Product Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5 Compact Topological Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1 Sequentially Compact Topological Spaces . . . . . . . . . . 25
6 Compact Subsets of RN . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
7 Continuous Functions on Countably Compact Spaces . . . . . . . . 28
8 Products of Compact Spaces . . . . . . . . . . . . . . . . . . . . . . . . . 28

xvii
xviii Contents

9 Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
9.1 Convex Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
9.2 Linear Maps and Isomorphisms . . . . . . . . . . . . . . . . . 32
10 Topological Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 32
10.1 Boundedness and Continuity . . . . . . . . . . . . . . . . . . . 34
11 Linear Functionals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
12 Finite Dimensional Topological Vector Spaces. . . . . . . . . . . . . 36
12.1 Locally Compact Spaces . . . . . . . . . . . . . . . . . . . . . . 36
13 Metric Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
13.1 Separation and Axioms of Countability . . . . . . . . . . . . 38
13.2 Equivalent Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . 39
13.3 Pseudo Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
14 Metric Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
14.1 Maps Between Metric Spaces . . . . . . . . . . . . . . . . . . . 41
15 Spaces of Continuous Functions. . . . . . . . . . . . . . . . . . . . . . . 42
15.1 Spaces of Continuously Differentiable Functions. . . . . . 43
15.2 Spaces of Hölder and Lipschitz Continuous
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
16 On the Structure of a Complete Metric Space . . . . . . . . . . . . . 44
16.1 The Uniform Boundedness Principle . . . . . . . . . . . . . . 45
17 Compact and Totally Bounded Metric Spaces . . . . . . . . . . . . . 46
17.1 Pre-Compact Subsets of X . . . . . . . . . . . . . . . . . . . . . 48
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1c Topological Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.12c Connected Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 49
1.19c Separation Properties of Topological Spaces . . . . . . . . 49
4c Bases, Axioms of Countability and Product Topologies . . . . . . 50
4.10c The Box Topology . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5c Compact Topological Spaces . . . . . . . . . . . . . . . . . . . . . . . . 51
5.8c The Alexandrov One-Point Compactification
of fX; Ug ([3]) . . . . . . . . . . . . . . . . . . . . . . . . .... 52
7c Continuous Functions on Countably Compact Spaces . . . .... 53
7.1c Upper-Lower Semi-continuous Functions . . . . . . .... 53
7.2c Characterizing Lower-Semi Continuous Functions
in RN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7.3c On the Weierstrass-Baire Theorem . . . . . . . . . . . . . . . 54
7.4c On the Assumptions of Dini’s Theorem . . . . . . . . . . . 56
9c Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
9.3c Hamel Bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
9.6c On the Dimension of a Vector Space . . . . . . . . . . . . . 58
10c Topological Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 58
13c Metric Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
13.10c The Hausdorff Distance of Sets . . . . . . . . . . . . . . . . . 60
13.11c Countable Products of Metric Spaces . . . . . . . . . . . . . 61
Contents xix

14c Metric Vector Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 62

15c Spaces of Continuous Functions . . . . . . . . . . . . . . . . . . . . .. 63
15.1c Spaces of Hölder and Lipschitz Continuous
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
16c On the Structure of a Complete Metric Space . . . . . . . . . . . . . 63
16.3c Completion of a Metric Space . . . . . . . . . . . . . . . . . . 64
16.4c Some Consequences of the Baire Category Theorem . . 65
17c Compact and Totally Bounded Metric Spaces . . . . . . . . . . . . . 65
17.1c An Application of the Lebesgue Number Lemma . . . . . 66
3 Measuring Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1 Partitioning Open Subsets of RN . . . . . . . . . . . . . . . . . . . . . . 67
2 Limits of Sets, Characteristic Functions, and r-Algebras . . . . . . 68
3 Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.1 Finite, r-Finite, and Complete Measures . . . . . . . . . . . 72
3.2 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4 Outer Measures and Sequential Coverings . . . . . . . . . . . . . . . . 73
4.1 The Lebesgue Outer Measure in RN . . . . . . . . . . . . . . 74
4.2 The Lebesgue–Stieltjes Outer Measure [89, 154]. . . . . . 75
5 The Hausdorff Outer Measure in RN [71] . . . . . . . . . . . . . . . . 75
5.1 Metric Outer Measures . . . . . . . . . . . . . . . . . . . . . . . 77
6 Constructing Measures from Outer Measures [26] . . . . . . . . . . 78
7 The Lebesgue–Stieltjes Measure on R . . . . . . . . . . . . . . . . . . 80
7.1 Borel Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
8 The Hausdorff Measure on RN . . . . . . . . . . . . . . . . . . . . . . . 81
9 Extending Measures from Semi-algebras to r-Algebras . . . . . . . 83
9.1 On the Lebesgue–Stieltjes and Hausdorff Measures . . . . 85
10 Necessary and Sufficient Conditions for Measurability . . . . . . . 85
11 More on Extensions from Semi-algebras to r-Algebras . . . . . . . 86
12 The Lebesgue Measure of Sets in RN . . . . . . . . . . . . . . . . . . . 87
12.1 A Necessary and Sufficient Condition
of Measurability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
13 Vitali’s Nonmeasurable Set [168] . . . . . . . . . . . . . . . . . . . . . . 90
14 Borel Sets, Measurable Sets, and Incomplete Measures . . . . . . . 91
14.1 A Continuous Increasing Function f : ½0; 1 ! ½0; 1 . . . 91
14.2 On the Preimage of a Measurable Set . . . . . . . . . . . . . 93
14.3 Proof of Propositions 14.1 and 14.2 . . . . . . . . . . . . . . 94
15 Borel Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
16 Borel, Regular, and Radon Measures . . . . . . . . . . . . . . . . . . . 96
16.1 Regular Borel Measures. . . . . . . . . . . . . . . . . . . . . . . 97
16.2 Radon Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
17 Vitali Coverings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
18 The Besicovitch Covering Theorem . . . . . . . . . . . . . . . . . . . . 101
19 Proof of Proposition 18.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
20 The Besicovitch Measure-Theoretical Covering Theorem . . . . . 106
xx Contents

Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

1c Partitioning Open Subsets of RN . . . . . . . . . . . . . . . . . . . . . . 108
2c Limits of Sets, Characteristic Functions and r-Algebras . . . . . . 108
3c Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
3.1c Completion of a Measure Space . . . . . . . . . . . . . . . . 111
4c Outer Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
5c The Hausdorff Outer Measure in RN . . . . . . . . . . . . . . . . . . . 112
5.1c The Hausdorff Dimension of a Set E  RN . . . . . . . . . 112
5.2c The Hausdorff Dimension of the Cantor Set
is ln 2/ln 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 113
8c The Hausdorff Measure in RN . . . . . . . . . . . . . . . . . . . . . .. 114
8.1c Hausdorff Outer Measure of the Lipschitz Image
of a Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 114
8.2c Hausdorff Dimension of Graphs of Lipschitz
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 115
9c Extending Measures from Semi-algebras to r-Algebras . . . . .. 115
9.1c Inner and Outer Continuity of k
on Some Algebra Q . . . . . . . . . . . . . . . . . . . . . . . .. 115
10c More on Extensions from Semi-algebras to r-Algebras . . . . .. 116
10.1c Self-extensions of Measures . . . . . . . . . . . . . . . . . .. 116
10.2c Nonunique Extensions of Measures k on
Semi-algebras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
12c The Lebesgue Measure of Sets in RN . . . . . . . . . . . . . . . . . . 118
12.1c Inner Measure and Measurability . . . . . . . . . . . . . . . . 120
12.2c The Peano–Jordan Measure of Bounded Sets in RN . . . 120
12.3c Lipschitz Functions and Measurability . . . . . . . . . . . . 121
13c Vitali’s Nonmeasurable Set . . . . . . . . . . . . . . . . . . . . . . . . . 122
14c Borel Sets, Measurable Sets and Incomplete Measures . . . . . . . 123
16c Borel, Regular and Radon Measures . . . . . . . . . . . . . . . . . . . 124
16.1c Regular Borel Measures . . . . . . . . . . . . . . . . . . . . . . 124
16.2c Regular Outer Measures . . . . . . . . . . . . . . . . . . . . . . 124
17c Vitali Coverings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
17.1c Pointwise and Measure-Theoretical Vitali Coverings . . 125
18c The Besicovitch Covering Theorem . . . . . . . . . . . . . . . . . . . . 126
18.1c The Besicovitch Theorem for Unbounded E . . . . . . . . 126
18.2c The Besicovitch Measure-Theoretical Inner Covering
of Open Sets E  RN . . . . . . . . . . . . . . . . . . . . . . .. 127
18.3c A Simpler Form of the Besicovitch Theorem . . . . . .. 127
18.4c Another Besicovitch-Type Covering . . . . . . . . . . . . .. 130
4 The Lebesgue Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
1 Measurable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
2 The Egorov–Severini Theorem [39, 145]. . . . . . . . . . . . . . . . . 135
2.1 The Egorov–Severini Theorem in RN . . . . . . . . . . . . . 137
3 Approximating Measurable Functions by Simple Functions . . . . 137
Contents xxi

4 Convergence in Measure (Riesz [125], Fisher [46]) . . . . . . . . . 139

5 Quasicontinuous Functions and Lusin’s Theorem . . . . . . . . . . . 141
6 Integral of Simple Functions ([87]). . . . . . . . . . . . . . . . . . . . . 143
7 The Lebesgue Integral of Nonnegative Functions . . . . . . . . . . . 144
8 Fatou’s Lemma and the Monotone Convergence Theorem. . . . . 145
9 More on the Lebesgue Integral . . . . . . . . . . . . . . . . . . . . . . . 147
10 Convergence Theorems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
11 Absolute Continuity of the Integral. . . . . . . . . . . . . . . . . . . . . 150
12 Product of Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
13 On the Structure of ðA  BÞ . . . . . . . . . . . . . . . . . . . . . . . . . 152
14 The Theorem of Fubini–Tonelli . . . . . . . . . . . . . . . . . . . . . . . 155
14.1 The Tonelli Version of the Fubini Theorem . . . . . . . . . 156
15 Some Applications of the Fubini–Tonelli Theorem . . . . . . . . . . 157
15.1 Integrals in Terms of Distribution Functions. . . . . . . . . 157
15.2 Convolution Integrals . . . . . . . . . . . . . . . . . . . . . . . . 158
15.3 The Marcinkiewicz Integral ([101, 102]) . . . . . . . . . . . 160
16 Signed Measures and the Hahn Decomposition . . . . . . . . . . . . 161
17 The Radon-Nikodým Theorem. . . . . . . . . . . . . . . . . . . . . . . . 163
17.1 Sublevel Sets of a Measurable Function. . . . . . . . . . . . 164
17.2 Proof of the Radon-Nikodým Theorem . . . . . . . . . . . . 165
18 Decomposing Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
18.1 The Jordan Decomposition. . . . . . . . . . . . . . . . . . . . . 168
18.2 The Lebesgue Decomposition . . . . . . . . . . . . . . . . . . . 168
18.3 A General Version of the Radon-Nikodým Theorem . . . 169
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
1c Measurable Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
1.1c Sublevel Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
2c The Egorov–Severini Theorem . . . . . . . . . . . . . . . . . . . . . . . 173
3c Approximating Measurable Functions by Simple Functions . . . 173
4c Convergence in Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
7c The Lebesgue Integral of Nonnegative Measurable
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 175
7.1c Comparing the Lebesgue Integral
with the Peano-Jordan Integral . . . . . . . . . . . . . . . . . 175
7.2c On the Definition of the Lebesgue Integral . . . . . . . . . 176
9c More on the Lebesgue Integral . . . . . . . . . . . . . . . . . . . . . . . 177
10c Convergence Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
10.1c Another Version of Dominated Convergence . . . . . . . . 178
11c Absolute Continuity of the Integral . . . . . . . . . . . . . . . . . . . . 181
12c Product of Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
12.1c Product of a Finite Sequence of Measure Spaces . . . . . 182
13c On the Structure of ðA  BÞ . . . . . . . . . . . . . . . . . . . . . . . . 183
13.1c Sections and Their Measure . . . . . . . . . . . . . . . . . . . 184
14c The Theorem of Fubini–Tonelli . . . . . . . . . . . . . . . . . . . . . . 185
xxii Contents

15c Some Applications of the Fubini–Tonelli Theorem . . ....... 186

15.1c Integral of a Function as the “Area
Under the Graph” . . . . . . . . . . . . . . . . . . . ....... 186
15.2c Distribution Functions . . . . . . . . . . . . . . . . ....... 186
17c The Radon-Nikodým Theorem . . . . . . . . . . . . . . . . ....... 187
18c A Proof of the Radon-Nikodým Theorem When Both
l and m Are r-Finite . . . . . . . . . . . . . . . . . . . . . . . ....... 189
5 Topics on Measurable Functions of Real Variables. . . . . . . . . . . . 193
1 Functions of Bounded Variation ([78]) . . . . . . . . . . . . . . . . . . 193
2 Dini Derivatives ([37]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
3 Differentiating Functions of Bounded Variation . . . . . . . . . . . . 197
4 Differentiating Series of Monotone Functions . . . . . . . . . . . . . 198
5 Absolutely Continuous Functions ([91, 169]) . . . . . . . . . . . . . . 199
6 Density of a Measurable Set . . . . . . . . . . . . . . . . . . . . . . . . . 201
7 Derivatives of Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
8 Differentiating Radon Measures . . . . . . . . . . . . . . . . . . . . . . . 204
9 Existence and Measurability of Dl m . . . . . . . . . . . . . . . . . . . . 206
9.1 Proof of Proposition 9.2 . . . . . . . . . . . . . . . . . . . . . . 208
10 Representing Dl m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
10.1 Representing Dl m for m  l. . . . . . . . . . . . . . . . . . . . 208
10.2 Representing Dl m for m ? l . . . . . . . . . . . . . . . . . . . . 210
11 The Lebesgue-Besicovitch Differentiation Theorem . . . . . . . . . 210
11.1 Points of Density . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
11.2 Lebesgue Points of an Integrable Function . . . . . . . . . . 211
12 Regular Families . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
13 Convex Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
14 The Jensen’s Inequality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
15 Extending Continuous Functions . . . . . . . . . . . . . . . . . . . . . . 216
15.1 The Concave Modulus of Continuity of f . . . . . . . . . . 216
16 The Weierstrass Approximation Theorem . . . . . . . . . . . . . . . . 218
17 The Stone-Weierstrass Theorem . . . . . . . . . . . . . . . . . . . . . . . 219
18 Proof of the Stone-Weierstrass Theorem . . . . . . . . . . . . . . . . . 220
18.1 Proof of Stone’s Theorem . . . . . . . . . . . . . . . . . . . . . 221
19 The Ascoli-Arzelà Theorem. . . . . . . . . . . . . . . . . . . . . . . . . . 222
19.1 Pre-compact Subsets of CðEÞ  ............. . . . . . . 223
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
1c Functions of Bounded Variations . . . . . . . . . . . . . . . . . . . . . 224
1.1c The Function of The Jumps . . . . . . . . . . . . . . . . . . . 225
1.2c The Space BV½a; b . . . . . . . . . . . . . . . . . . . . . . . . . 225
2c Dini Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
2.1c A Continuous, Nowhere Differentiable
Function ([167]) . . . . . . . . . . . . . . . . . . . . . ...... 228
2.2c An Application of the Baire Category Theorem ...... 229
Contents xxiii

4c Differentiating Series of Monotone Functions . . . . ......... 229

5c Absolutely Continuous Functions . . . . . . . . . . . . ......... 229
5.1c The Cantor Ternary Function ([23]) . . . . . ......... 230
5.2c A Continuous Strictly Monotone Function
with a.e. Zero Derivative . . . . . . . . . . . . ......... 231
5.3c Absolute Continuity of the Distribution
Function of a Measurable Function . . . . . ......... 233
7c Derivatives of Integrals . . . . . . . . . . . . . . . . . . . ......... 234
7.1c Characterizing BV½a; b Functions . . . . . . ......... 236
7.2c Functions of Bounded Variation
in N Dimensions [55] . . . . . . . . . . . . . . . . . . . . . . . . 238
13c Convex Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
13.8c Convex Functions in RN . . . . . . . . . . . . . . . . . . . . . . 240
13.14c The Legendre Transform ([92]) . . . . . . . . . . . . . . . . . 241
13.15c Finiteness and Coercivity . . . . . . . . . . . . . . . . . . . . . 241
13.16c The Young’s Inequality . . . . . . . . . . . . . . . . . . . . . . 242
14c Jensen’s Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
14.1c The Inequality of the Geometric
and Arithmetic Mean . . . . . . . . . . . . . . . . . . . . . . . . 243
14.2c Integrals and Their Reciprocals . . . . . . . . . . . . . . . . . 243
15c Extending Continuous Functions . . . . . . . . . . . . . . . . . . . . . . 244
16c The Weierstrass Approximation Theorem . . . . . . . . . . . . . . . . 244
17c The Stone-Weierstrass Theorem . . . . . . . . . . . . . . . . . . . . . . 244
19c A General Version of the Ascoli-Arzelà Theorem . . . . . . . . . . 245
6 The LP Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
1 Functions in Lp ðEÞ and Their Norm . . . . . . . . . . . . . . . . . . . . 247
2 The Hölder and Minkowski Inequalities . . . . . . . . . . . . . . . . . 248
3 More on the Spaces Lp and Their Norm . . . . . . . . . . . . . . . . . 250
3.1 Characterizing the Norm kf kp for 1  p\1 . . . . . . . . 250
3.2 The Norm k  k1 for E of Finite Measure . . . . . . . . . . 250
3.3 The Continuous Version of the Minkowski
Inequality. . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 251
4 Lp ðEÞ for 1  p  1 as Normed Spaces of Equivalence
Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 252
4.1 Lp ðEÞ for 1  p  1 as a Metric Topological
Vector Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
5 Convergence in Lp ðEÞ and Completeness . . . . . . . . . . . . . . . . 253
6 Separating Lp ðEÞ by Simple Functions . . . . . . . . . . . . . . . . . . 255
7 Weak Convergence in Lp ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . . 257
7.1 Counterexample . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
8 Weak Lower Semi-continuity of the Norm in Lp ðEÞ . . . . . . . . . 258
9 Weak Convergence and Norm Convergence . . . . . . . . . . . . . . 259
9.1 Proof of Proposition 9.1 for p 2 . . . . . . . . . . . . . . . . 260
9.2 Proof of Proposition 9.1 for 1\p\2 . . . . . . . . . . . . . 261
xxiv Contents

10 Linear Functionals in Lp ðEÞ. . . . . . . . . . . . . . . . . . . . . . . . .. 261

11 The Riesz Representation Theorem. . . . . . . . . . . . . . . . . . . .. 263
11.1 Proof of Theorem 11.1: The Case
of fX; A; lg Finite . . . . . . . . . . . . . . . . . . . . . . . . .. 263
11.2 Proof of Theorem 11.1:
The Case of fX; A; lg r-Finite. . . . . . . . . . . . . . . . . . 264
11.3 Proof of Theorem 11.1: The Case 1\p\1 . . . . . . . . 265
12 The Hanner and Clarkson Inequalities. . . . . . . . . . . . . . . . . . . 266
12.1 Proof of Hanner’s Inequalities . . . . . . . . . . . . . . . . . . 268
12.2 Proof of Clarkson’s Inequalities . . . . . . . . . . . . . . . . . 268
13 Uniform Convexity of Lp ðEÞ for 1\p\1 . . . . . . . . . . . . . . . 269
14 The Riesz Representation Theorem By Uniform Convexity . . . . 271
14.1 Proof of Theorem 14.1. The Case 1\p\1 . . . . . . . . 271
14.2 The Case p ¼ 1 and E of Finite Measure . . . . . . . . . . . 272
14.3 The Case p ¼ 1 and fX; A; lg r-Finite . . . . . . . . . . . . 273
15 If E  RN and p 2 ½1; 1Þ, then Lp ðEÞ Is Separable . . . . . . . . . 273
15.1 L1 ðEÞ Is Not Separable. . . . . . . . . . . . . . . . . . . . . . . 276
16 Selecting Weakly Convergent Subsequences . . . . . . . . . . . . . . 276
17 Continuity of the Translation in Lp ðEÞ for 1  p\1 . . . . . . . . 277
17.1 Continuity of the Convolution . . . . . . . . . . . . . . . . . . 280
18 Approximating Functions in Lp ðEÞ with Functions in C 1 ðEÞ. . . 280
19 Characterizing Pre-compact Sets in Lp ðEÞ . . . . . . . . . . . . . . . . 283
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
1c Functions in Lp ðEÞ and Their Norm . . . . . . . . . . . . . . . . . . . 285
1.1c The Spaces Lp for 0\p\1 . . . . . . . . . . . . . . . . . . . . 285
1.2c The Spaces Lq for q\0 . . . . . . . . . . . . . . . . . . . . . . 285
1.3c The Spaces ‘p for 1  p  1 . . . . . . . . . . . . . . . . . . . 285
2c The Inequalities of Hölder and Minkowski . . . . . . . . . . . . . . . 286
2.1c Variants of the Hölder and Minkowski Inequalities . . . 286
2.2c Some Auxiliary Inequalities . . . . . . . . . . . . . . . . . . . 287
2.3c An Application to Convolution Integrals . . . . . . . . . . . 287
2.4c The Reverse Hölder and Minkowski Inequalities . . . . . 288
2.5c Lp ðEÞ-Norms and Their Reciprocals . . . . . . . . . . . . . . 288
3c More on the Spaces Lp and Their Norm . . . . . . . . . . . . . . . . . 288
3.4c A Metric Topology for Lp ðEÞ when 0\p\1 . . . . . . . 289
3.5c Open Convex Subsets of Lp ðEÞ for 0\p\1 . . . . . . . . 289
5c Convergence in Lp ðEÞ and Completeness . . . . . . . . . . . . . . . . 290
5.1c The Measure Space fX; A; lg and the Metric
Space fA; dg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
6c Separating Lp ðEÞ by Simple Functions . . . . . . . . . . . . . . . . . . 291
7c Weak Convergence in Lp ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . 292
7.3c Comparing the Various Notions of Convergence . . . . . 293
7.5c Weak Convergence in ‘p . . . . . . . . . . . . . . . . . . . . . 294
Contents xxv

9c Weak Convergence and Norm Convergence . . . . . . . . . . . . . . 295

9.1c Proof of Lemmas 9.1 and 9.2 . . . . . . . . . . . . . . . . . . 295
11c The Riesz Representation Theorem . . . . . . . . . . . . . . . . . . . . 296
11.1c Weakly Cauchy Sequences in Lp ðXÞ for 1\p  1 . . . 296
11.2c Weakly Cauchy Sequences in Lp ðXÞ for p ¼ 1 . . . . . . 297
11.3c The Riesz Representation Theorem in ‘p . . . . . . . . . . . 297
14c The Riesz Representation Theorem By Uniform Convexity . . . 297
14.1c Bounded Linear Functional in Lp ðEÞ for 0\p\1 . . . . 297
14.2c An Alternate Proof of Proposition 14.1c . . . . . . . . . . . 298
15c If E  RN and p 2 ½1; 1Þ, then Lp ðEÞ Is Separable . . . . . . . . . 299
18c Approximating Functions in Lp ðEÞ with Functions in C 1 ðEÞ . . 299
18.1c Caloric Extensions of Functions in Lp ðRN Þ . . . . . . . . . 299
18.2c Harmonic Extensions of Functions in Lp ðRN Þ . . . . . . . 301
18.3c Characterizing Hölder Continuous Functions . . . . . . . . 303
19c Characterizing Pre-compact Sets in Lp ðEÞ . . . . . . . . . . . . . . . 304
19.1c The Helly’s Selection Principle . . . . . . . . . . . . . . . . . 304
20c The Vitali-Saks-Hahn Theorem [59, 138, 170] . . . . . . . . . . . . 305
21c Uniformly Integrable Sequences of Functions . . . . . . . . . . . . . 307
22c Relating Weak and Strong Convergence and Convergence
in Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 309
23c An Independent Proof of Corollary 22.1c . . . . . . . . . . . . . . .. 311
7 Banach Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
1 Normed Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
1.1 Semi-norms and Quotients . . . . . . . . . . . . . . . . . . . . . 314
2 Finite and Infinite Dimensional Normed Spaces . . . . . . . . . . . . 315
2.1 A Counterexample . . . . . . . . . . . . . . . . . . . . . . . . . . 315
2.2 The Riesz Lemma. . . . . . . . . . . . . . . . . . . . . . . . . . . 316
2.3 Finite Dimensional Spaces . . . . . . . . . . . . . . . . . . . . . 317
3 Linear Maps and Functionals . . . . . . . . . . . . . . . . . . . . . . . . . 318
4 Examples of Maps and Functionals . . . . . . . . . . . . . . . . . . . . 320
4.1 Functionals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
4.2 Linear Functionals on CðEÞ  ................. . . . 321
a 
4.3 Linear Functionals on C ðEÞ for Some a 2 ð0; 1Þ . . . . . 321
5 Kernels of Maps ad Functionals . . . . . . . . . . . . . . . . . . . . . . . 321
6 Equibounded Families of Linear Maps . . . . . . . . . . . . . . . . . . 322
6.1 Another Proof of Proposition 6.1 . . . . . . . . . . . . . . . . 323
7 Contraction Mappings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
7.1 Applications to Some Fredholm Integral Equations . . . . 325
8 The Open Mapping Theorem. . . . . . . . . . . . . . . . . . . . . . . . . 325
8.1 Some Applications . . . . . . . . . . . . . . . . . . . . . . . . . . 327
8.2 The Closed Graph Theorem . . . . . . . . . . . . . . . . . . . . 327
9 The Hahn–Banach Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 328
xxvi Contents

10 Some Consequences of the Hahn–Banach Theorem . . . . . .... 330

10.1 Tangent Planes . . . . . . . . . . . . . . . . . . . . . . . . . .... 332
11 Separating Convex Subsets of a Hausdorff, Topological
Vector Space fX; Ug . . . . . . . . . . . . . . . . . . . . . . . . . . .... 332
11.1 Separation in Locally Convex, Hausdorff,
Topological Vector Spaces fX; Ug . . . . . . . . . . . . . . . 334
12 Weak Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
12.1 Weak Boundedness . . . . . . . . . . . . . . . . . . . . . . . . . . 336
12.2 Weakly and Strongly Closed Convex Sets . . . . . . . . . . 337
13 Reflexive Banach Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
14 Weak Compactness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
14.1 Weak Sequential Compactness . . . . . . . . . . . . . . . . . . 340
15 The Weak Topology of X  . . . . . . . . . . . . . . . . . . . . . . . . . . 342
16 The Alaoglu Theorem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
17 Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
17.1 The Schwarz Inequality . . . . . . . . . . . . . . . . . . . . . . . 345
17.2 The Parallelogram Identity . . . . . . . . . . . . . . . . . . . . . 346
18 Orthogonal Sets, Representations and Functionals . . . . . . . . . . 346
18.1 Bounded Linear Functionals on H. . . . . . . . . . . . . . . . 348
19 Orthonormal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
19.1 The Bessel Inequality . . . . . . . . . . . . . . . . . . . . . . . . 349
19.2 Separable Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . 350
20 Complete Orthonormal Systems . . . . . . . . . . . . . . . . . . . . . . . 351
20.1 Equivalent Notions of Complete Systems. . . . . . . . . . . 352
20.2 Maximal and Complete Orthonormal Systems . . . . . . . 352
20.3 The Gram–Schmidt Orthonormalization
Process ([142]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
20.4 On the Dimension of a Separable Hilbert Space . . . . . . 353
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
1c Normed Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
1.1c Semi-Norms and Quotients . . . . . . . . . . . . . . . . . . . . 354
2c Finite and Infinite Dimensional Normed Spaces . . . . . . . . . . . 355
3c Linear Maps and Functionals . . . . . . . . . . . . . . . . . . . . . . . . 356
6c Equibounded Families of Linear Maps . . . . . . . . . . . . . . . . . . 357
8c The Open Mapping Theorem . . . . . . . . . . . . . . . . . . . . . . . . 358
9c The Hahn–Banach Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 358
9.1c The Complex Hahn–Banach Theorem . . . . . . . . . . . . 359
9.2c Linear Functionals in L1 ðEÞ . . . . . . . . . . . . . . . . . . . 359
11c Separating Convex Subsets of X . . . . . . . . . . . . . . . . . . . . . . 360
11.1c A Counterexample of Tukey [164] . . . . . . . . . . . . . . . 360
11.2c A Counterexample of Goffman and Pedrick [56] . . . . . 361
11.3c Extreme Points of a Convex Set . . . . . . . . . . . . . . . . 361
11.4c A General Version of the Krein–Milman Theorem . . . . 363
Contents xxvii

12c Weak Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363

12.1c Infinite Dimensional Normed Spaces . . . . . . . . . . . . . 364
12.2c About Corollary 12.5 . . . . . . . . . . . . . . . . . . . . . . . . 365
12.3c Weak Closure and Weak Sequential Closure . . . . . . . . 365
14c Weak Compactness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
14.1c Linear Functionals on Subspaces of CðEÞ  ........ . . 368
14.2c Weak Compactness and Boundedness . . . . . . . . . . . . 369
15c The Weak Topology of X  . . . . . . . . . . . . . . . . . . . . . . . . . 369
15.1c Total Sets of X . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
15.2c Metrization Properties of Weak Compact Subsets
of X  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
16c The Alaoglu Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
16.1c The Weak Topology of X  . . . . . . . . . . . . . . . . . . . 372
16.2c Characterizing Reflexive Banach Spaces . . . . . . . . . . . 374
16.3c Metrization Properties of the Weak Topology
of the Closed Unit Ball of a Banach Space . . . . . . . . . 374
16.4c Separating Closed Sets in a Reflexive Banach Space . . 375
17c Hilbert Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
17.1c On the Parallelogram Identity . . . . . . . . . . . . . . . . . . 376
18c Orthogonal Sets, Representations and Functionals . . . . . . . . . . 376
19c Orthonormal Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
8 Spaces of Continuous Functions, Distributions,
and Weak Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
1 Bounded Linear Functionals on Co ðRN Þ . . . . . . . . . . . . . . . . . 379
1.1 Positive Linear Functionals on Co ðRN Þ . . . . . . . . . . . . 380
1.2 The Riesz Representation Theorem . . . . . . . . . . . . . . . 380
2 Partition of Unity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
3 Proof of Theorem 1.1. Constructing l. . . . . . . . . . . . . . . . . . . 381
4 An Auxiliary Positive Linear Functional on Co ðRN Þ þ . . . . . . . 383
4.1 Measuring Compact Sets by T þ . . . . . . . . . . . . . . . . . 384
5 Representing T þ on Co ðRN Þ þ as in (1.1) for a Unique lB . . . . 385
6 Proof of Theorem 1.1. Representing T on Co ðRN Þ
as in (1.3) for a Unique l-Measurable w . . . . . . . . . . . . . . . . . 386
7 A Topology for Co1 ðEÞ for an Open Set E  RN . . . . . . . . . . . 387
8 A Metric Topology for Co1 ðEÞ . . . . . . . . . . . . . . . . . . . . . . . 389
8.1 Equivalence of These Topologies . . . . . . . . . . . . . . . . 389
8.2 DðEÞ Is Not Complete. . . . . . . . . . . . . . . . . . . . . . . . 390
9 A Topology for Co1 ðKÞ for a Compact Set K  E . . . . . . . . . . 391
9.1 DðKÞ Is Complete. . . . . . . . . . . . . . . . . . . . . . . . . . . 391
9.2 Relating the Topology of DðEÞ to the Topology
of DðKÞ. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 392
10 The Schwartz Topology of DðEÞ . . . . . . . . . . . . . . . . . . .... 392
xxviii Contents

11 DðEÞ Is Complete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

11.1 Cauchy Sequences in DðEÞ and Completeness . . . . . . . 394
11.2 The Topology of DðEÞ Is Not Metrizable . . . . . . . . . . 394
12 Continuous Maps and Functionals . . . . . . . . . . . . . . . . . . . . . 395
12.1 Distributions in E . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
12.2 Continuous Linear Maps T : DðEÞ ! DðEÞ . . . . . . . . . 396
13 Distributional Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
14 Fundamental Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
14.1 The Fundamental Solution of the Wave Operator
in R2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
14.2 The Fundamental Solution of the Laplace Operator . . . . 400
15 Weak Derivatives and Main Properties . . . . . . . . . . . . . . . . . . 401
16 Domains and Their Boundaries . . . . . . . . . . . . . . . . . . . . . . . 403
16.1 @E of Class C1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
16.2 Positive Geometric Density and @E Piecewise
Smooth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
16.3 The Segment Property . . . . . . . . . . . . . . . . . . . . . . . . 404
16.4 The Cone Property . . . . . . . . . . . . . . . . . . . . . . . . . . 405
16.5 On the Various Properties of @E . . . . . . . . . . . . . . . . . 405
17 More on Smooth Approximations. . . . . . . . . . . . . . . . . . . . . . 405
18 Extensions into RN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
19 The Chain Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
20 Steklov Averagings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
20.1 Characterizing W 1;p ðEÞ for 1\p\1 . . . . . . . . . . . . . 412
20.2 Remarks on W 1;1 ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . 413
21 The Rademacher’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 413
Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
1c Bounded Linear Functionals on Co ðRN ; Rm Þ . . . . . . . . . . . . . . 415
2c Convergence of Measures . . . . . . . . . . . . . . . . . . . . . . . . . . 416
3c Calculus with Distributions . . . . . . . . . . . . . . . . . . . . . . . . . 418
4c Limits in D0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
5c Algebraic Equations in D0 . . . . . . . . . . . . . . . . . . . . . . . . . . 424
6c Differential Equations in D0 . . . . . . . . . . . . . . . . . . . . . . . . . 426
7c Miscellaneous Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
9 Topics on Integrable Functions of Real Variables. . . . . . . ...... 431
1 A Vitali-Type Covering . . . . . . . . . . . . . . . . . . . . . . ...... 431
2 The Maximal Function (Hardy–Littlewood [69]
and Wiener [175]) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
3 Strong Lp Estimates for the Maximal Function. . . . . . . . . . . . . 435
3.1 Estimates of Weak and Strong Type . . . . . . . . . . . . . . 436
4 The Calderón–Zygmund Decomposition Theorem [20] . . . . . . . 437
Contents xxix

5 Functions of Bounded Mean Oscillation . . . . . . . . . . . . . .... 438

5.1 Some Consequences of the John–Nirenberg
Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
6 Proof of the John–Nirenberg Theorem 5.1. . . . . . . . . . . . . . . . 441
7 The Sharp Maximal Function . . . . . . . . . . . . . . . . . . . . . . . . 444
8 Proof of the Fefferman–Stein Theorem . . . . . . . . . . . . . . . . . . 445
9 The Marcinkiewicz Interpolation Theorem. . . . . . . . . . . . . . . . 447
9.1 Quasi-linear Maps and Interpolation . . . . . . . . . . . . . . 448
10 Proof of the Marcinkiewicz Theorem . . . . . . . . . . . . . . . . . . . 449
11 Rearranging the Values of a Function . . . . . . . . . . . . . . . . . . . 451
12 Some Integral Inequalities for Rearrangements . . . . . . . . . . . . . 453
12.1 Contracting Properties of Symmetric
Rearrangements . . . . . . . . . . . . . . . . . . . . . . . . .... 454
12.2 Testing for Measurable Sets E Such that ¼ E a.e.
in RN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 455
13 The Riesz Rearrangement Inequality. . . . . . . . . . . . . . . . .... 457
13.1 Reduction to Characteristic Functions
of Bounded Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
14 Proof of (13.1) for N ¼ 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
14.1 Reduction to Finite Union of Intervals. . . . . . . . . . . . . 458
14.2 Proof of (13.1) for N ¼ 1. The Case T þ S  R. . . . . . . 460
14.3 Proof of (13.1) for N ¼ 1. The Case S þ T [ R . . . . . . 461
14.4 Proof of the Lemma 14.1. . . . . . . . . . . . . . . . . . . . . . 463
15 The Hardy’s Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463
16 The Hardy–Littlewood–Sobolev Inequality for N ¼ 1 . . . . . . . . 465
16.1 Some Reductions . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
17 Proof of Theorem 16.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
18 The Hardy–Littlewood–Sobolev Inequality for N 1 . . . . . . . . 468
18.1 Proof of Theorem 18.1 . . . . . . . . . . . . . . . . . . . . . . . 468
19 Potential Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
20 Lp Estimates of Riesz Potentials. . . . . . . . . . . . . . . . . . . . . . . 470
20.1 Motivating Lp Estimates of Riesz Potentials
as Embeddings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
21 Lp Estimates of Riesz Potentials for p ¼ 1 and p [ N. . . . . . . . 472
22 The Limiting Case p ¼ N . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
23 Steiner Symmetrization of a Set E  RN . . . . . . . . . . . . . . . . . 475
24 Some Consequences of Steiner’s Symmetrization . . . . . . . . . . . 477
24.1 Symmetrizing a Set About the Origin . . . . . . . . . . . . . 477
24.2 The Isodiametric Inequality . . . . . . . . . . . . . . . . . . . . 478
24.3 Steiner Rearrangement of a Function . . . . . . . . . . . . . . 479
25 Proof of the Riesz Rearrangement Inequality for N ¼ 2 . . . . . . 479
25.1 The Limit of fFn g . . . . . . . . . . . . . . . . . . . . . . . . . . 481
25.2 The Set F Is the Disc F  . . . . . . . . . . . . . . . . . . . . . 483
26 Proof of the Riesz Rearrangement Inequality for N [ 2 . . . . . . 484
xxx Contents

Problems and Complements. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . 485

11c Rearranging the Values of a Function . . . . . . . . . . .. . . . . . . 485
12c Some Integral Inequalities for Rearrangements . . . . .. . . . . . . 486
20c Lp Estimates of Riesz Potentials . . . . . . . . . . . . . . .. . . . . . . 486
21c Lp Estimates of Riesz Potentials for p ¼ 1 and p N . . . . . . . 487
22c The Limiting Case p ¼ Na . . . . . . . . . . . . . . . . . . . .. . . . . . . 487
23c Some Consequences of Steiner’s Symmetrization . . .. . . . . . . 488
23.1c Applications of the Isodiametric Inequality . .. . . . . . . 488
10 Embeddings of W1,p(E) into Lq(E) . . . . . . . . . . . . . . . . . . . . . . . . 491
1 Multiplicative Embeddings of Wo1; p ðEÞ . . . . . . . . . . . . . . . . . . 491
1.1 Proof of Theorem 1.1 . . . . . . . . . . . . . . . . . . . . . . . . 493
2 Proof of Theorem 1.1 for N ¼ 1 . . . . . . . . . . . . . . . . . . . . . . 493
3 Proof of Theorem 1.1 for 1  p\N . . . . . . . . . . . . . . . . . . . . 494
4 Proof of Theorem 1.1 for 1  p\N Concluded . . . . . . . . . . . . 496
5 Proof of Theorem 1.1 for p N [ 1. . . . . . . . . . . . . . . . . . . . 496
5.1 Estimate of I1 ðx; RÞ . . . . . . . . . . . . . . . . . . . . . . . . . . 497
5.2 Estimate of I2 ðx; RÞ . . . . . . . . . . . . . . . . . . . . . . . . . . 498
6 Proof of Theorem 1.1 for p N [ 1 Concluded. . . . . . . . . . . . 498
7 On the Limiting Case p ¼ N . . . . . . . . . . . . . . . . . . . . . . . . . 499
8 Embeddings of W 1; p ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
9 Proof of Theorem 8.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
10 Poincaré Inequalities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503
10.1 The Poincaré Inequality . . . . . . . . . . . . . . . . . . . . . . . 504
10.2 Multiplicative Poincaré Inequalities . . . . . . . . . . . . . . . 505
10.3 Extensions of ðu uE Þ for Convex E . . . . . . . . . . . . . 506
11 Level Sets Inequalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
12 Morrey Spaces [110] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
12.1 Embeddings for Functions in the Morrey Spaces. . . . . . 509
13 Limiting Embedding of W 1; N ðEÞ . . . . . . . . . . . . . . . . . . . . . . 510
14 Compact Embeddings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
15 Fractional Sobolev Spaces in RN . . . . . . . . . . . . . . . . . . . . . . 514
16 Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
17 Traces and Fractional Sobolev Spaces. . . . . . . . . . . . . . . . . . . 518
18 Traces on @E of Functions in W 1; p ðEÞ . . . . . . . . . . . . . . . . . . 519
18.1 Traces and Fractional Sobolev Spaces . . . . . . . . . . . . . 521
19 Multiplicative Embeddings of W 1; p ðEÞ . . . . . . . . . . . . . . . . . . 522
20 Proof of Theorem 19.1. A Special Case . . . . . . . . . . . . . . . . . 524
21 Constructing a Map Between E and Q. Part I . . . . . . . . . . . . . 526
21.1 Case 1. Rn Intersects Bq . . . . . . . . . . . . . . . . . . . . . . 527
21.2 Case 2. Rn Does not Intersect Bq . . . . . . . . . . . . . . . . 528
22 Constructing a Map Between E and Q. Part II . . . . . . . . . . . . . 528
23 Proof of Theorem 19.1 Concluded . . . . . . . . . . . . . . . . . . . . . 530
24 The Spaces Wp1; p ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
Contents xxxi

Problems and Complements. . . . . . . . . . . . . . . . . . . . . . . . . . .... 534

1c Multiplicative Embeddings of Wo1;p ðEÞ . . . . . . . . . . . . . . .... 534
8c Embeddings of W 1;p ðEÞ . . . . . . . . . . . . . . . . . . . . . . . . .... 534
8.1c Differentiability of Functions in W 1;p ðEÞ for p N ... 535
14c Compact Embeddings . . . . . . . . . . . . . . . . . . . . . . . . . .... 536
17c Traces and Fractional Sobolev Spaces . 1. . . . . . . . . . . . . .... 536
17.1c Characterizing Functions in W 1 p;p ðRN Þ as Traces .... 536
18c Traces on @E of Functions in W 1;p ðEÞ . . . . . . . . . . . . . . .... 538
18.1c Traces on a Sphere . . . . . . . . . . . . . . . . . . . . . .... 538
11 Topics on Weakly Differentiable Functions . . . . . . . . . . . . . . . .. 541
1 Sard’s Lemma [140]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 541
2 The Co-area Formula for Smooth Functions . . . . . . . . . . . . .. 544
3 The Isoperimetric Inequality for Bounded Sets E
with Smooth Boundary @E . . . . . . . . . . . . . . . . . . . . . . . . .. 545
3.1 Embeddings of Wo1; p ðEÞ Versus the Isoperimetric
Inequality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 546
4 The p-Capacity of a Compact Set K  RN , for 1  P\N . . . .. 547
4.1 Enlarging the Class of Competing Functions . . . . . . .. 548
5 A Characterization of the p-Capacity of a Compact
Set K  RN , for 1  P\N . . . . . . . . . . . . . . . . . . . . . . . . .. 550
6 Lower Estimates of cp ðKÞ for 1  P\N . . . . . . . . . . . . . . . .. 552
6.1 A Simpler Proof of Lemma 6.1 with a Coarser
Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 553
6.2 p-Capacity of a Closed Ball B  q  RN , for 1  p\N . .. 554
6.3 
cp ðBq Þ ¼ cp ð@Bq Þ . . . . . . . . . . . . . . . . . . . . . . . . . .. 555
7 The Norm kDukp , for 1  P\N, in Terms of the p-Capacity
Distribution Function of u 2 Co1 ðRN Þ. . . . . . . . . . . . . . . . . .. 555
7.1 Some Auxiliary Estimates for 1\p\N . . . . . . . . . . .. 555
7.2 Proof of Theorem 7.1 . . . . . . . . . . . . . . . . . . . . . . .. 557
8 Relating Gagliardo Embeddings, Capacities,
and the Isoperimetric Inequality . . . . . . . . . . . . . . . . . . . . . . . 558
9 Relating HN p ðKÞ to cp ðKÞ for 1\p\N . . . . . . . . . . . . . . . . 559
9.1 An Auxiliary Proposition . . . . . . . . . . . . . . . . . . . . . . 559
9.2 Proof of Theorem 9.1 . . . . . . . . . . . . . . . . . . . . . . . . 561
10 Relating cp ðKÞ to HN p þ e ðKÞ for 1  p\N . . . . . . . . . . . . . . 562
11 The p-Capacity of a Set E  RN for 1  P\N . . . . . . . . . . . . 563
12 Limits of Sets and Their Outer p-Capacities . . . . . . . . . . . . . . 566
13 Capacitable Sets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
14 Capacities Revisited and p-Capacitability of Borel Sets . . . . . . . 570
14.1 The Borel Sets in RN Are p-Capacitable . . . . . . . . . . . 571
14.2 Generating Measures by p-Capacities . . . . . . . . . . . . . 572
15 Precise Representatives of Functions in L1loc ðRN Þ . . . . . . . . . . . 572
xxxii Contents

16 Estimating the p-Capacity of ½u [ t for t [ 0 . . . . . . . . . . . . . 574

1;p
17 Precise Representatives of Functions in Wloc ðRN Þ . . . . . . . . . . 576
17.1 Quasi-Continuous Representatives of Functions
1;p
u 2 Wloc ðRN Þ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585