PDF The Mathematics of Finite Networks An Introduction To Operator Graph Theory 1st Edition Michael Rudolph Download

Full download test bank at ebookmeta.
com
The Mathematics of Finite Networks An Introduction

to Operator Graph Theory 1st Edition Michael
Rudolph
For dowload this book click LINK or Button below
https://ebookmeta.com/product/the-mathematics-of-
finite-networks-an-introduction-to-operator-graph-
theory-1st-edition-michael-rudolph/
OR CLICK BUTTON
DOWLOAD EBOOK
Download More ebooks from https://ebookmeta.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Spatial Networks: A Complete Introduction: From Graph

Theory and Statistical Physics to Real-World
Applications Barthelemy
https://ebookmeta.com/product/spatial-networks-a-complete-
introduction-from-graph-theory-and-statistical-physics-to-real-
world-applications-barthelemy/
Discrete Mathematics with Graph Theory 1st Edition

Santosh Kumar Yadav
https://ebookmeta.com/product/discrete-mathematics-with-graph-
theory-1st-edition-santosh-kumar-yadav/
Graph and Network Theory: An Applied Approach using

Mathematica® 1st Edition Michael A. Henning
https://ebookmeta.com/product/graph-and-network-theory-an-
applied-approach-using-mathematica-1st-edition-michael-a-henning/
Introduction to Graph Theory: With Solutions to

Selected Problems 2nd Edition Khee Meng Koh
https://ebookmeta.com/product/introduction-to-graph-theory-with-
solutions-to-selected-problems-2nd-edition-khee-meng-koh/
Guide to Discrete Mathematics: An Accessible
Introduction to the History, Theory, Logic and
Applications 2nd Edition Gerard O'Regan
https://ebookmeta.com/product/guide-to-discrete-mathematics-an-
accessible-introduction-to-the-history-theory-logic-and-
applications-2nd-edition-gerard-oregan/
An Introduction to Chemistry Michael Mosher
https://ebookmeta.com/product/an-introduction-to-chemistry-
michael-mosher/
A Walk Through Combinatorics An Introduction to

Enumeration and Graph Theory and Selected Other Topics
5th Edition Miklós Bóna
https://ebookmeta.com/product/a-walk-through-combinatorics-an-
introduction-to-enumeration-and-graph-theory-and-selected-other-
topics-5th-edition-miklos-bona/
Qualitative Theory of ODEs: An Introduction to

Dynamical Systems Theory 1st Edition Henryk Zoladek
https://ebookmeta.com/product/qualitative-theory-of-odes-an-
introduction-to-dynamical-systems-theory-1st-edition-henryk-
zoladek/
An Undergraduate Introduction to Financial Mathematics

4th Edition Buchanan
https://ebookmeta.com/product/an-undergraduate-introduction-to-
financial-mathematics-4th-edition-buchanan/
The Mathematics of Finite Networks
Since the early eighteenth century, the theory of networks and graphs has matured into
an indispensable tool for describing countless real-world phenomena. However, the
study of large-scale features of a network often requires unrealistic limits, such as
taking the network size to infinity or assuming a continuum. These asymptotic and
analytic approaches can significantly diverge from real or simulated networks when
applied at the finite scales of real-world applications. This book offers an approach to
overcoming these limitations by introducing operator graph theory, an exact,
non-asymptotic set of tools combining graph theory with operator calculus. The book
is intended for mathematicians, physicists, and other scientists interested in discrete
finite systems and their graph-theoretical description, and in delineating the abstract
algebraic structures that characterise such systems. All the necessary background on
graph theory and operator calculus is included for readers to understand the potential
applications of operator graph theory.
m i c h a e l r u d o l p h is a mathematical physicist of the French National Centre

for Scientific Research at the Denis Poisson Institute. His research includes graph
theory and classical number theory, and is directed towards understanding physical
reality from an inherently finite discrete perspective, both mathematically and
philosophically.
The Mathematics of Finite Networks
An Introduction to Operator Graph Theory
M I C H A E L RU D O L P H
CNRS / Institute Denis Poisson, University of Tours
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107134430
DOI: 10.1017/9781316466919
© Michael Rudolph 2022
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2022
A catalogue record for this publication is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Names: Rudolph, Michael, 1969– author.
Title: The mathematics of finite networks : an introduction to operator
graph theory / Michael Rudolph.
Description: Cambridge ; New York, NY : Cambridge University Press, 2022. |
Includes bibliographical references and index.
Identifiers: LCCN 2021058077 (print) | LCCN 2021058078 (ebook) |
ISBN 9781107134430 (hardback) | ISBN 9781316466919 (epub)
Subjects: LCSH: Graph theory. | Operator theory. | System analysis. |
BISAC: COMPUTERS / General
Classification: LCC QA166 .R83 2022 (print) | LCC QA166 (ebook) |
DDC 511/.5–dc23/eng/20220126
LC record available at https://lccn.loc.gov/2021058077
LC ebook record available at https://lccn.loc.gov/2021058078
ISBN 978-1-107-13443-0 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
For Uesat
— in memoriam —
Contents
Preface page ix
1 Introduction 1
PART I OPERATOR GRAPH THEORY

2 Classical Graph Theory: The Mathematical Description
of Networks 13
2.1 Graphs and Networks 13
2.2 Graph Measures 26
2.3 Graph Models 60
2.4 Finite Graph Theory 81
3 Operator Calculus: The Mapping between Vector Spaces 88
3.1 Vector Spaces 88
3.2 Operations on Vector Spaces 93
3.3 Mappings between Vector Spaces 101
3.4 Algebras of Operators 105
3.5 Special Operators 113
4 Operator Graph Theory: The Mathematics of
Finite Networks 117
4.1 From Operations on Graphs to Operator Graphs 117
4.2 Operator Graph Theory 139
4.3 Transforming Operator Graphs 152
4.4 Measuring Operator Graphs 163
4.5 Generating Operator Graphs 169
PART II APPLICATIONS
5 Generating Graphs 181
5.1 Erdős–Rényi Random Graphs 181
5.2 Gilbert Random Graphs 200
vii
viii Contents
5.3 Watts–Strogatz Random Graphs 205

5.4 Barabási–Albert Random Graphs 216
5.5 Random Graphs with Arbitrary Degree Distribution 226
5.6 Randomised Square Grid Graphs 231
6 Measuring Graphs 238
6.1 Total Adjacency of Random Graphs and Digraphs 238
6.2 Asymmetry Index of Random Digraphs 243
6.3 Clustering Coefficient of Random Digraphs 249
6.4 Walks in Gilbert Random Digraphs 264
6.5 Node Degrees of Barabási–Albert Random Graphs 276
6.6 Paths in Randomised Square Grid Graphs 280
6.7 Challenges and Limitations 285
7 Transforming Graphs 291
7.1 Positional Chess Graphs 291
7.2 Generation of Positional Chess Graphs 297
7.3 Transformation of Positional Chess Graphs 321
Afterthought 325
Bibliography 327
Index of Notations 333
Subject Index 336
Preface
Imagine for a moment that you stroll through a beautiful city like London
with its dizzying array of inspiring detail. An overwhelming number of small
streets, unpredictably intersecting with one another, wind around busy alley-
ways which themselves appear to stretch endlessly into all directions. Finding
your way in this arguably chaotic mesh of interconnected paths and road-
ways by foot or by car without resorting to a memorised representation or
detailed map of its layout would require the often time-consuming search
through a vast number of possibilities until, eventually, the desired destination
is reached. Taking a bird’s-eye perspective, however, this confusing maze of
small streets and mews seems to magically disappear and reveal, with concen-
tric rings of expressways framing a star of boulevards which emanate from the
metropole’s centre, a different, much simpler pattern. Although such a coarser
structural view certainly helps to confine your search, it does so at the expense
of abstracting from finer details, – the small, narrow streets which are not only
an integral part of the cultural and historical lifeblood of a city like London but
reside at the very heart of its flair. Naturally, you might then wonder whether it
is possible to hold on to these pertinent details alongside the more crude struc-
tural peculiarities which emerge at larger scales and hold, not unlike a skeleton,
those details together.
When thinking of city maps, we instinctively envision a network of links
along which an ever-changing flow of traffic is carried. Such an idealised
description, however, is not limited to the maps we are all familiar with. From
the interactions between atoms and subatomic particles to the gravitational
forces which act between the billions of galaxies stretching across the known
universe, from the transmission of electrical signals in our brains to the com-
plexity of social interactions between people, most if not all of the phenomena
we encounter, consciously or not, all around us find a natural representation
in the form of networks. Indeed, it can be indubitably argued that the abstract
ix
x Preface
notion of interacting objects resides at the very heart of our conceptual under-
standing of nature as it touches upon the very fabric of physical reality with
its finite and discrete makeup. It is for that reason that we now find the math-
ematical study of interconnected objects, the theory of networks and graphs,
branching deeply through all fields of science and its innumerable applications.
Since its formal inception in the early eighteenth century, graph theory has
constructed a solid foundation on which its conceptual and mathematical roots
now comfortably rest. In doing so, it has amassed a plethora of useful tools
which allow for a thorough qualification and quantification of the sheer
uncountable number of networks we find all around us. Unfortunately, such a
realisation of graph-theoretical concepts does not come without its limitations.
All too often, it is the very nature of a studied system which also can, and does,
throw stones in the way of its graph-theoretical formalisation or viable qual-
ifying generalisations. Worse even, many real-world systems defiantly refuse
an analytical treatment with the toolkit of graph theory and can only be studied
numerically, thus, they are inadvertently yet inescapably tied to the efficiency
of our computational hardware, or the lack thereof.
To bypass these limitations, and to capture the essential structural properties
of real-world networks, the construction of simplified mathematical models
is required. However, as in the example of London’s map, many of the cur-
rent mathematical approaches provide only a bird’s-eye view, at the expense
of the finer details, in order to reveal the large-scale features of the network
that may exist, or to allow certain simplifications to become effective. Indeed,
these approaches regularly consider unrealistic limits, such as taking the net-
work size to infinity or abstracting from a network’s defining discrete makeup
by assuming a continuum. While such ‘asymptotic’ and ‘analytic’ approaches
can certainly provide simple representations of the dominant structural features
and, thus, often succeed at describing network structure in the aggregate, how-
ever, the equations at their core can diverge from real or simulated networks
quite significantly, especially when applied to the finite scales occurring in the
real world. What if the algorithms behind network models – and, for good
models, behind real-world data as well – could be cast in mathematical terms
that would allow for direct manipulations and, in some cases, exact solutions
without the necessity of imposing limits? What if such algorithms could be
cast into a form which would allow one to consider specific graph ensembles
all at once, without the need to computationally construct and assess concrete
individual realisation?
In this book, we aim to introduce, or at least propose, exactly such an
approach to graph theory. The mathematical language used for formalising
this discrete and finite algebraic framework lends itself to operator calculus,
Preface xi
and we have termed this framework operator graph theory. Irrespective of

being tackled from a computational or more mathematical perspective, the
investigation of large yet finite networks and their defining structural peculiar-
ities still remains largely out of reach. Algorithmic approaches often become
lost in a vast set of possibilities, as computational methods cannot cope with
the combinatorial complexities inherent to many graph-theoretical problems;
while asymptotic approaches are often unable to capture the very structure
they are obliged to exclude, especially where real-world data are concerned.
With this in mind, it is our great hope that operator graph theory may pro-
vide a viable language capable of casting such descriptions, by providing an
exact, non-asymptotic set of tools which can be applied to, and provides links
between, real-world data and their graph-theoretical representations.
This book is intended primarily for physicists, mathematicians, biologists
and computer scientists, among others, who are interested in discrete, finite
real-world systems that can be described in graph-theoretical terms, as well
as in delineating the abstract algebraic structures characterising such systems.
Including all the necessary background material on graph theory and opera-
tor calculus, and requiring no specific prerequisites other than linear algebra
and calculus at the undergraduate level, this book is essentially self-contained
and suitable for advanced undergraduates, postgraduates and researchers from
a variety of backgrounds. However, it is important to stress that this book has
to be understood as nothing more than a mere proposition which attempts the
natural, yet arguably novel, fusion of graph theory and operator calculus. With
many details in need of being fleshed out by the unforgiving demands of mathe-
matical rigour still at large, I nevertheless hope to provide sufficient inspiration
and motivation for an exciting explorative journey across uncharted waters. Let
us take this journey together and see where it will lead us, because, after all,
this is what science is all about.
I humbly admit that this book has been some years in the making, during a
time of great professional and personal hardship. For that reason, I must first
and foremost thank my ever-patient editor Lauren Cowles and her editorial
assistant Amy He for their unwavering support and understanding nature. The
same applies to my fellow colleagues at the Institute Denis Poisson in Tours
and Orleans, who gave me in the ‘Year of Crazy’ a new comfortable home and
awoke a memory of still-existing places in which only the sky is the limit, a
cherished memory I thought had long been lost. When thinking about the tran-
quility of a new home, it is impossible not to see Madame Gigi and Daniel
‘Rambo72’ who not only helped me to build, from nothing, a secluded Asgar-
dian paradise deep in the intoxicating beauty of Sarthe, France, but gave me
xii Preface
through their kindness, friendship and unconditional support the strength to

master the uncompromising waves of fate crashing down on me. Writing these
lines would also have been impossible without my ever-growing family of cats
who finally can, like their ‘Old Man’, enjoy a simple life in nature and the
freedom we are born with. Accompanied by the words of the great Charles
Dickens, ‘what a greater gift than the love of a cat’, my beyond-infinite thanks
go to Uesat and Black Mountain, Schmusie and Flusie, as well as the many
cats I could not save from the abyss of human savagery. Last but not least, my
profound and eternal thanks to Morgan and the two ‘Hepburns’, Audrey and
Katherine, and Mark and Andrew, for their selfless support in the dark times I
endured, as well as to Claude for keeping alive the fire of passion which burns
in each true explorer of the mysteries of nature. To all of you I owe my life,
and the candle of hope which, once again, is shining brightly deep inside me.
Finally, thanks go to Lyle E. Muller for his help in preparing the original pro-
posal for this book, as well as to J.A.G. Willow, C.O. Cain, S. Hower and L.S.
Dee for their invaluable guidance, without which the conclusion of this work
would have been impossible.
1
Introduction
When on sunny days in the early years of the eighteenth century people has-
tened through the narrow streets of Königsberg, now Kaliningrad in Russia, it
was not for catching a breeze of fresh air, but for being the lucky first to unlock
the mystery of its now famous bridges. Not a valuable golden treasure was to
be unearthed, but the solution to the simple problem of finding a path over the
seven bridges that connect both banks of the river Pregel with the two islands
around which that beautiful Prussian town was built. The goal, however, was
not to find just any path, but one along which each bridge was crossed only
once. A futile endeavour indeed, and it is unknown how many of Königsberg’s
inhabitants and intrigued travellers from afar succumbed to this formidable
challenge before in 1736 the brilliant mind of a young genius named Leonard
Euler provided the definite answer. Such a path does not exist! [42].
What was Euler’s deep insight which allowed him to solve the problem of
Königsberg’s bridges without setting foot on any of them, the insight which
had eluded all mathematical treasure hunters, professional and recreational
alike, before him? In short, seven bridges connect the four main geographi-
cal areas of Königsberg, but only one or two of these areas can serve as start
and finish of the sought-after path through the city. This means that at least two
areas must be traversed and, in order to do so, must be connected by an even
number of bridges across the river Pregel. However, as a quick look at the map
of Königsberg reveals (Fig. 1.1a), each of its four areas was accessible only
through an odd number of bridges, thus rendering the existence of a valid path
an impossibility.
Although it still took many decades after Euler’s simple yet brilliant reso-
lution of the now famous Königsberg bridge problem before a coherent and
rigorous mathematical framework was established, it undoubtedly marks both
historically and conceptually the beginning of what is now known as network
or graph theory. Indeed, his contribution in this regard cannot be overstated.
1
2 Introduction
a b c
A B C D
C g
A
c d B
e
A D C
D
a b
B f weight
1 2 3
Figure 1.1 The bridges of Königsberg and their graph-theoretical representation.

a: Seven bridges (a through f ) connect both sides of the river Pregel (B and C)
with the large islands Kneiphof and Lomse (A and D). Is it possible to reach all
four parts of the city by crossing each bridge once and only once? b: Euler’s
ingenious solution rests on the abstraction of the problem, replacing A, B, C and
D by nodes, and the bridges by links, or edges, interconnecting these nodes. As
all nodes are connected by an odd number of edges, the problem’s solution is
in the negative [42]. c: Visualisation of the weighted adjacency relations defin-
ing the network at the heart of the Königsberg bridge problem in the form of a
matrix. Rows and columns represent the departure and arrival areas, respectively,
the weights indicate the number of bridges connecting each of these areas with
one another. Panels a and b are modified from Kraitchik [68, Chapter 8.4].
By stripping colourful yet less relevant details off the given original problem,
Euler single-handedly introduced an abstract notion which not just allowed
him to uncover its defining essence, but in fact touches upon the very fabric
of physical reality. Each real-world system is finite and discrete, comprised of
discernible parts which share links to, interact with or assert an action on other
parts of the system. Be it the plethora of elementary particles interacting with
one another through the exchange of virtual bosons, the complex dendritic and
axonal trees which carry electrical signals through a network of synaptically
interlinked neurons in our brains, or the arrangement of galaxies into large
filaments through the action of the gravitational force, all natural phenomena
we discovered so far can be abstractly described by and modelled as networks
of interconnected objects. These objects are called nodes or vertices, their con-
nections edges, and graph theory, in its most general sense, is the mathematical
framework for the study of networks formed by nodes and edges.
Truth be told, it would be an almost futile endeavour to present just a glimpse
of the large body of graph-theoretical literature which emerged in the past three
centuries since Euler’s time. For that reason, we must, and only humbly can,
refer the interested reader to the historically first textbook written by Kőnig
[65], the still-unmatched definite textbook on the subject by Harary [59] and
perhaps the most comprehensive presentation of modern applications of graph
Introduction 3
theory by Newman [86], and by doing so challenge, somewhat self-servingly,

each reader to construct from these initial literature nodes his or her very
own ‘network of exploration’ of the formidable world of graph-theoretical lit-
erature. From this it will become immediately clear that not even a genius
such as Euler could have envisioned the undoubtedly overwhelming success of
graph theory, its underpinning conceptual beauty and far-reaching applications
in the study, modelling and characterisation of real-world systems. Unfortu-
nately, however, even the greatest success story hides a tiny stain, some form
of caveat hinting at a possible limitation or complication, and it is such a stain
on the Michelangelonian masterpiece of graph theory which this book intends
to identify and eventually target.
As we already asserted above, due to the very makeup of physical reality,
most, if not all, real-world phenomena can be described in terms of networks
of interconnected objects. Graph theory, as the mathematical framework for-
malising such a description, has over the past century not just provided us
with a vast yet ever-growing number of concrete examples of networks – rang-
ing in size from a few nodes, such as in the case of the Königsberg bridges,
to hundreds of million of nodes in the case of the World Wide Web – but also
amassed a plethora of tools and methods for analysing and understanding these
networks. Most of these tools and applied methods, however, are of a quantita-
tive nature and, thus, inherently rely on numerical evaluations by hand or, more
contemporarily, the use of computers. Although big advances in computational
hardware over the past few decades have made it possible to characterise and
delineate finer structural details of increasingly larger networks, the question of
principal limitations of computational approaches looms like the sword of des-
tiny over the head of every researcher who dares to venture into the seemingly
endless realms of graph theory. How far can we go, or – more importantly –
how far will we be ever able to go?
To illustrate the more fundamental nature of this question, let us return to
the seven bridges of Königsberg and ignore for a moment Euler’s definite, and
simple, solution of the associated problem in the negative. How many paths
through the city are possible? More precisely, how many Eulerian walks – that
is, paths which traverse each bridge at most once, do exist – and how many
of these walks define Eulerian paths by crossing each bridge exactly once? It
will not take long to draw a tree of all possibilities, a small fraction of which is
shown in Fig. 1.2, and count a total of 90 Eulerian walks starting from either
bank of the river Pregel (node B or C, Fig. 1.1), 6 of which leave three bridges
uncrossed, 12 of which leave two bridges uncrossed, and 72 of which end
with one bridge remaining untraversed. None of these 90 walks – nor, for that
matter, any of the possible walks emanating from the other parts of the city – is,
4 Introduction
Figure 1.2 Solving the problem of Königsberg’s seven bridges through an exhaus-
tive search. Each branch in this tree constructs a possible Eulerian walk, with
bridges being crossed marked in grey and bridges already traversed in black.
Shown are only Eulerian walks starting from one bank of the river Pregel, omit-
ting all walks which begin by crossing the second bridge to the south. Each branch
of this tree of possibilities leaves at least one bridge uncrossed; thus, none of the
constructed walks constitutes a sought-after Eulerian path.
thus, an Eulerian path. Without having ever set foot in Königsberg, we just
solved the problem of its bridges by brute force on a piece of paper. But today,
one can find some 100 bridges and overpasses in Kaliningrad. How long would
it take now to draw a tree of all possible walks through the city?
In short, about 10150 times as long, give or take some orders of magnitude!
A formidable exercise which undoubtedly lies beyond the capacity of even the
most modern computers, let alone a wholeheartedly motivated researcher with
paper and pencil. The longer and more nuanced answer to this question neces-
sitates the introduction of the notion of computational complexity, a measure
Introduction 5
which, in the most general sense, quantifies the scaling behaviour of computa-
tional algorithms with respect to the magnitude of their input variables, in our
case the number of bridges and overpasses. To distinguish further the time it
takes to run a given algorithm from the storage space it requires, both defining
aspects quantifying the efficiency of a computational algorithm in relation to
the stringent constraints imposed by the utilised computational hardware, one
typically differentiates time complexity and space complexity, respectively.
Let us first take a look at the time complexity of our little exercise. Draw-
ing a single walk through the city which traverses each bridge – that is, each
edge in the associated graph – at most once certainly scales linearly with the
total number of bridges E. In mathematical terms, one expresses such a scal-
ing behaviour with the notion O(E). However, our exercise requires us to draw
all Eulerian walks and identify potential Eulerian paths, a rather brute-force
approach which is called exhaustive search. In order to do so, we must start at
every possible point and traverse all the possible paths from that point onwards
(Fig. 1.2), a task with a staggering time complexity of not less than O(E!).
Luckily for us, the original problem absolves us from having to cope with
such an unmanageable complexity, as it requires us only to decide wether an
Eulerian path exists or not. Indeed, following Euler’s inspiring solution, we
only need to calculate the number of edges connected to each node and test
wether any of these numbers is even, a task which can certainly be performed
in O(E) and leaves us off the hook, this time. Unfortunately, however, many
of the quantitative graph-theoretical methods and measures do not allow for
such an elegant approach, and they typically scale polynomially with the size
of the graph, that is, with a time complexity of O(E k ) for some k 1. Undeni-
ably, such a scaling is certainly better than the factorial scaling we encountered
above in a brute-force construction. But even with the development of compu-
tational hardware in the foreseeable future or the conception of novel, highly
efficient computational algorithms, such a polynomial time complexity is still
hard to cope with and likely renders the exact quantification of larger networks
difficult or even impossible for many decades to come.
The situation regarding space complexity does not look much less gloomy
when dealing with graph-theoretical formulations of real-world phenomena.
Although efficient representations for a small subset of special graphs exist,
generally one is left with a matrix representation of a graph’s adjacency rela-
tions (see Fig. 1.1c), an approach with a space complexity of O(N 2 ) in the
number of nodes N. Certainly, neither the seven bridges of Königsberg during
Euler’s time nor the more numerous bridges in current-day Kaliningrad pose a
representational problem here. But what about the graph describing the World
Wide Web with its hundreds of millions of nodes [1, 34]? Even if we assign
6 Introduction
only a Boolean number in the form of a computational bit to indicate the pres-
ence, or absence, of an edge between any given pair of nodes, it would require
a staggering 1000 terabytes to store the complete adjacency matrix of a graph
describing only 100 million linked websites. Luckily for us, in this specific
case, a linear list of source and target nodes would suffice, as each node of
this graph is, at average, connected to only seven or eight other nodes (e.g.,
see [26, 54] but also [77]). But even such a sparse representation would still
require about one gigabyte of memory alone for storing the graph’s defining
structure, without any computational analysis having yet been performed on
it. When considering still larger or less sparsely connected graphs, such as the
network of neurons in the human brain with thousands of connections embed-
ding each of its 100 billion nodes [11, 23], we quickly run out of options as
the memory requirements for their representation alone will breach the storage
capacity made available in modern computers, perhaps even the computational
devices of the next few generations.
The question, then, is, where does this leave us? Of course, a computa-
tional analysis is just one of several available options when facing a graph-
theoretical formulation of a natural phenomenon. Indeed, the very limitation of
a purely numerical approach has led to the development of a whole host of new
and exciting methods and continues to stimulate novel ideas and techniques
beyond the mere optimisation of computational algorithms. Arguably, one of
the conceptually most important and fruitful of these ideas is the distinction of
real-world networks into classes of simplified, heavily idealised graph models.
Examples of such idealised models include perhaps the most-widely known
Erdős–Rényi model [22, 38] which generally serves as the prototype when
referring to random graphs, the small-world graph of Watts and Strogatz [109]
inspired by Stanley Milgram’s infamous ‘six degrees of separation’ [78] and
the scale-free graph of Barabási and Albert [12] which captures the property of
self-similarity found in complex systems, to mention but a few. Most of these
models are conceived by abstracting from many of the finer structural pecu-
liarities we observe in real-world networks, details which are, as we argued
above, increasingly harder to delineate through computational approaches in
cases where larger or densely connected networks are concerned. With these
graph models typically requiring less than a handful of parameters for their
construction and characterisation, the simplified nature of the models then not
only allows for the application of numerical optimisation techniques which
significantly reduce their computational complexity, but eventually opens the
door for a more rigorous mathematical exploration by allowing for an algebraic
formalisation of their algorithmic construction.
Introduction 7
In this book, we will propose and motivate exactly such a rigorous math-
ematical approach. By embarking on this adventurous journey, however, we
will seek to avoid at least one of the major caveats accompanying many of the
idealised graph-theoretical models proposed as descriptions of real-world phe-
nomena since the time of Euler. To briefly illustrate this caveat, let us consider
perhaps the simplest of these abstract models, that of the Erdős–Rényi random
graph. For each point in its two-dimensional parameter space, which is spanned
by the number of nodes and the number of edges, or, alternatively but with sub-
tle conceptual differences, the connection probability between pairs of nodes,
a whole ensemble of possible realisations exists, each of which describes a
distinct connectivity pattern between all possible pairs of nodes. Although the
algebraic formalisation of this model allows, at least in principle, access to this
ensemble, most of the defining characteristics of the Erdős–Rényi model can
still only be assessed numerically, ideally so over a large number of realisations
for each parameter set in order to ensure statistical significance of the obtained
results.
Naturally, just as in the case of graph-theoretical descriptions of real-world
phenomena, such a computational approach is subject to the stringent limita-
tions mentioned earlier, especially when models of larger graphs are consid-
ered. Due to its mathematically well-rooted conception, however, the Erdős–
Rényi random graph model also invites a more rigorous analytical examina-
tion. Some of its characteristics, such as the number and distribution of edges
connected to each node, are indeed governed by exact formulae which, due to
the random makeup of the Erdős–Rényi model, are probabilistic in nature and
deliver the expectation value for those characteristics across the entire param-
eter space and, thus, the whole ensemble of possible realisations. With other
properties – such as the shortest distance between any pair of nodes, measured
in number of edges which must at least be traversed to reach a target node, or
the clustering, quantified by the abundance of triangular cycles – we are unfor-
tunately not so lucky. Indeed, many if not most of the properties allow for an
analytical take only in some parts of the model’s parameter space, that is, under
certain and often stringent conditions imposed on the graph’s makeup.
It was the meticulous work of Paul Erdős and Alfréd Rényi in the early sec-
ond half of the last century which eventually established what is now called
random graph theory, the mathematical study of properties of the probability
space of random graphs, a field which stretches today with applications far
beyond the random graph model named after its original architects (for a com-
prehensive introduction, see [22]). The conceptual core of this theory defines a
set of conditions which allow for a mathematically rigorous quantification of a
8 Introduction
random graph’s properties. But – and herein lies the problem or caveat hinted at
above – these very conditions are rooted in asymptotic considerations, specif-
ically and most importantly the heavy-weighting assumption which demands
that the network size should become infinitely large. Although for the math-
savvy reader such a demand might certainly appear as nothing unusual, as
something being widely employed in mathematics and its many applications –
a trick if you will, to establish some form of analytical framework for describ-
ing a physical phenomenon or, as in the case of random graphs, for formalising
a theoretical model – we must stop here and evaluate its potentially pernicious
repercussions and pitfalls.
The important question to ask before issuing any demands on our approach
is, do we find in nature any system which we know for certain is infinite in
one or more of its properties or, alternatively, is genuinely continuous in an
analytical or, if you like, Cantorian sense? Or is it that our models, describing
these systems, assume such an idealising makeup for reasons of simplicity, or
the sheer lack of a more suitable approach in tune with our study of nature
and, thus, epistemology? We could go as far as posing the heretical question
of whether the fabric of physical reality itself is algebraic, that is finite and
discrete, or analytic, that is infinite and continuous, in nature? Although on
first sight such questions might appear far removed from an graph-theoretical
inquiry of our world – indeed, as being of little or no relevance today as our
analytical formulation of real-world phenomena has undeniably demonstrated
tremendous success – their gravity continues to penetrate philosophical discus-
sions and, arguably, lies at the very bedrock of modern science and its mathe-
matical foundation (a comprehensive review of arguments from both sides can
be found in Hagar [56]).
To illustrate this point, arguably naively so, in the context of this book, let
us take a look at a physical phenomenon which many of us encounter on a
daily basis, namely preparing, and eventually enjoying, a nice cup of coffee.
Hot water is poured on finely grained coffee and finds its way – ‘percolates’ –
through the porous brown powder to finally yield the tasty, longed-for bever-
age. This mundane process is just one example of a whole class of phenomena
in condensed matter physics which finds a natural representation in terms of
random graph theory. However, due to its analytical foundation, the obtained
results are, at best, valid only in the case that the number of nodes in the asso-
ciated graph-theoretical models approaches infinity – that is, we deal with an
impossible-to-imagine amount of coffee in front of us – or if the connection
probability between nodes is small enough that the connectivity pattern of the
model graph resembles that of a tree, a structure that is devoid of cycles or
loops and resembles, to stick with our illustrative example, anything but the
Introduction 9
coffee we are all familiar with. Indeed, the reality of making coffee certainly
looks much different from the model we use to describe it, as we do not have
at our disposal an (almost) infinite quantity of coffee powder for quenching
our desire, nor do we find a resemblance between a sparsely connected tree
and the gaps separating the coffee grains through which the water finds its way
into our cups!
Despite the fact that theoretical models, by their very conceptual defini-
tion, can provide only a simplified, more abstract view of natural phenomena,
the above highlighted differences are certainly, or at least arguably, not even
close to being a viable reflection of the real-world system in question. Thus,
reiterating the above questions, we must ask whether random graph theory,
indeed graph theory in general, with its analytical core is the right mathemati-
cal approach to describe the discrete and finite phenomena, such as percolation,
which we find all around us. After all, every system we can discern in nature,
in lack of a proof to the opposite, is both finite and discrete. An answer to this
question becomes even more pressing when considering systems which, by
their very definition, are bound in size, such as the game of chess, or display
boundaries which crucially shape the system’s defining behavioural charac-
teristics, such as the aromatic powder in a coffee machine. In such cases we
have no choice but to strictly adhere to a finite description, as each simplifying
asymptotic assumption will necessarily push us outside the defining premise
of the studied phenomenon and eventually deliver a model which no longer
observes the phenomenon being considered.
Is it possible to conceive of a mathematical framework which is capable
of more precisely capturing finite and discrete graph-theoretical models, thus
allowing for a more viable and, arguably, more accurate and thus valid descrip-
tion of real-world phenomena? In this book, we will answer this question affir-
matively and motivate a rigorous algebraic approach for the construction, anal-
ysis and characterisation of finite graphs. This approach, which we will term
operator graph theory, comes at the price of abandoning the original definition
of a network, or graph, which resides at the very core of classical graph theory
in favour of a more dynamic, operational viewpoint. A graph will no longer
be a mere static collection of nodes and edges, but instead a construct which
can evolve and change due to the actions of operators. We will argue that by
studying the algebraic properties typically associated with such operators, this
approach opens up a new dimension of qualitative and quantitative insights
into the very properties which define a given finite graph, insights which find
a viable mathematical representation without resorting to the dangerous tool
bag of asymptotic approximations and throttling limitations. In order to further
stress, on epistemological grounds and in full awareness of its heretical nature,
10 Introduction
this very rejection of a continuous, thus analytic, description of real-world phe-

nomena in favour of a finite, thus algebraic, approach, we will restrict, unless
otherwise motivated, to the ring of rational numbers Q instead of the ring of
real numbers R throughout the presentation.
However, before we embark on this adventurous journey, it is of upmost
importance to note that the approach presented in this book must and can only
be viewed as a coarse introduction, primarily focusing on the mere motivation
of a new angle from which to view graph theory and its utilisation as a descrip-
tive tool in the understanding of real-world phenomena. For that reason, the
book is divided into two main parts. In the first part, we will lay the neces-
sary theoretical foundation for operator graph theory. To that end, Chapter 2
will provide a necessarily incomplete introduction into classical graph theory
by focusing on the presentation of concepts, terminology and mathematical
notions needed for the comprehension of the material in the remainder of this
book. A similar approach holds for Chapter 3, in which we will take a brief
look at the essential fundamentals of the vast field of discrete operator calcu-
lus. By fusing the notion of a classical graph with concepts of operator theory
introduced in these two chapters, we will then be ready to define the central
contribution that this book is aiming for, namely that of an operator graph. In
Chapter 4, concluding the first part, not only will this fusion be motivated, but
the core conceptual notions of operator graph theory, along with definitions and
a presentation of its mathematical framework, will be justified and, hopefully,
accessibly illustrated.
The second part of this book focuses on exemplifying the proposed operator
graph-theoretical framework by presenting its applications to a few relatable
systems and well-known conceptual models of real-world phenomena. Specif-
ically, we will formalise the rigorous algebraic generation of the most-widely
used finite random graph models in Chapter 5, and then use this formalisation
in Chapter 6 to obtain exact algebraic expressions for a variety of properties
characterising these graphs, without resorting to asymptotic approximations.
Closing with Chapter 7, we will finish our journey with a brief stroll through
the playful realms of game theory by demonstrating the potential usefulness of
our operator graph-theoretical framework in the construction and analysis of
the game of chess.
We ardently hope that with the chosen examples and applications, the inter-
ested reader will be inspired to partake in the furthering of the yet-to-be-
fleshed-out theoretical foundations of the operator graph-theoretical frame-
work and be thoroughly motivated not just to consider the latter as a potentially
powerful ally in the graph-theoretical description and analysis of real-world
phenomena, but to actively utilise this exciting novel approach in the conquest
of understanding nature.
PART I
OPERATOR GRAPH THEORY

2
Classical Graph Theory:
The Mathematical Description of Networks
What do the bridges of Königsberg, synaptically connected neurons in our

brains, or the galaxies illuminating the dark voids of our universe have in com-
mon? All of these real-world phenomena can be described as collections of
discrete discernible objects which are interlinked to form weblike structures
called networks. This chapter will introduce the mathematical representation
of such networks and familiarise us with the basic concepts, ideas and termi-
nology of a vast and ever-growing research field whose roots date back to the
work of Leonard Euler. Naturally, such an introduction must remain short and
will aim primarily at building an intuition for the basic elements necessary
to arrive at an algebraic formalisation of networks and their structural char-
acteristics. By taking a closer look at a number of concrete network models,
specifically the random graph models which prominently feature as descriptive
vessels for many natural phenomena, and briefly exploring some deep-rooted
conceptual limitations of these models at the end of this chapter, we hope not
only to motivate the need for a rigorous mathematical framework for the study
of networks at finite scales, but also to accentuate the potential advantages of
a more dynamic vantage point from which to view networks and their defining
characteristics in later chapters of this book.
2.1 Graphs and Networks

Let us begin our journey by formalising the core object with which graph the-
ory is concerned. On its most basic conceptual level, a network, or graph in
the mathematical literature, is comprised of a collection of discrete objects,
interchangeably called vertices or nodes, and a collection of edges, also called
links or bonds, which describe relations between these objects. Remaining with
the example of Königsberg as the iconic city whose now historical mystery
13
14 Classical Graph Theory
inspired the mathematical field of network studies, or graph theory, we find

that each of the city’s four river banks and islands constitutes a node, while the
seven bridges connecting these areas establish relations between these nodes
and, thus, can be viewed as comprising a set of edges. The collection of areas
and bridges then forms the graph of Königsberg’s bridges, as illustrated in Fig.
1.1. This brilliant insight of Leonard Euler, by abstracting from colourful yet
less relevant details, not only allowed for a new way to look at and eventu-
ally solve the historical problem associated with Königsberg’s seven bridges
itself, but indeed established a new powerful framework which matured since
Euler’s time into an indispensable tool for studying the governing structural
and functional characteristics of a sheer unlimited range of real-world phenom-
ena and physical systems. In order to fully realise and harness the power of this
framework beyond a mere descriptive vessel, however, we certainly require a
mathematically more rigorous representation of a network and its conceptual
makeup. To that end, we formulate
Definition 2.1 (finite graph) A finite network or graph G is an ordered tuple

G = (N, E) of a set N , ∅ of nodes i with cardinality |N| ∈ N : |N| < ∞,
and a set E of edges with cardinality |E| ∈ N0 comprised of ordered pairs
(i, j) ∈ N × N, each denoting an adjacency relation, or link, between a source
node i and a target node j.
Although this now most commonly cited definition constitutes the very foun-
dation of classical graph theory, it must be cautioned that the defined notion
also remains rather limited in its scope and suffices only as a portrayal of the
principal algebraic building blocks which describe the structural makeup of a
given finite graph. Indeed, as each student of real-world graphs will quickly
realise, an algebraically more precise description and subsequent exploration
of a network’s characteristics necessitates a more suitable mathematical repre-
sentation of the set of nodes and their existing relations. This often includes, or
even requires, a quantification of the strength, or weight, of the relations estab-
lished between the nodes beyond the mere existence, or absence, of adjacency
relations themselves. Such a representation can be achieved by assigning to
each pair of nodes (i, j) with i, j ∈ N a number wi j ∈ Q, such that
wi j , 0 iff (i, j) ∈ E,
(
wi j : N × N → Q with (2.1)
wi j = 0 otherwise.
Referring to a node i by its order in the set N, this mapping is typically rep-
resented as |N| × |N| matrix and called the weight matrix w of the graph G.
Throughout this book – and, arguably, somewhat in defiance of the vast major-
ity of introductory textbooks of graph theory – we will accommodate this more
general stance and consider each of a finite graph’s edges as being weighted.
As we will see below, the various types of networks utilised in the applied
graph-theoretical literature then appear as special cases defined by imposing
constraints on the weight matrix.
Before utilising Definition 2.1 and the weight matrix (2.1) to furnish a classi-
cal graph-theoretical framework, however, let us briefly illustrate the algebraic
representation of a graph in the case of Königsberg’s bridges. The city’s four
areas (Kneiphof, Haberberg or Vorstadt, Löbenicht or Altstadt, and Lomse)
comprise all nodes of this graph and find a suitable representation in form of
the set N = {1, 2, 3, 4} by assigning a number to each area consecutively. With
this enumeration, the seven bridges connecting between the city’s areas can
then be listed as pairs (1, 2) and (2, 1) for the Grüne Brücke and Köttelbrücke,
(1, 3) and (3, 1) for the Krämerbrücke and Schmiedebrücke, (1, 4) and (4, 1)
for the Honigbrücke, (2, 4) and (4, 2) for the Hohebrücke, and (3, 4) and (4, 3)
for the Holzbrücke, and comprise the elements of E. In order to account for the
two bridges connecting Kneiphof and Haberberg, the adjacency relations (1, 2)
and (2, 1) are both assigned a weight of two. The same applies to the adjacency
relation between nodes 1 (Kneiphof) and 3 (Löbenicht), thus leaving us with
the weight matrix
 
 0 2 2 1 
 2 0 0 1 
 
w =  (2.2)
 2 0 0 1 
 
1 1 1 0
as not only an algebraic description of the graph associated with Königsberg’s
bridges (Fig. 2.1a), but also a suitable mathematical representation which will
ease the further analysis of this historically important graph.
2.1.1 Types of Graphs

The set of edges E together with the weight wi j associated with each ele-
ment (i, j) of E provides a complete mathematical description of the adjacency
relations between the nodes of a given graph G. Depending on the specific
properties of the set E and weight matrix w, various types of graphs can
then be distinguished, some of which are illustrated in Fig. 2.1 and listed in
Table 2.1. Definition 2.1 describes what is generally called a directed graph,
or digraph for short, as each (i, j) ∈ E is an ordered pair of nodes describ-
ing a directed edge whose presence in E neither requires the existence of
the associated reciprocal adjacency relations ( j, i) nor imposes conditions on
the weight of an eventually existing reciprocal adjacency relation ( j, i) ∈ E.
3 a b c
1 4
d e
Figure 2.1 Examples of various types of graphs. a: The seven bridges of Königs-
berg form an undirected multigraph, with the weight of the graph’s edges repre-
senting the number of bridges connecting each of the city’s four areas (thin solid:
1 bridge, thick solid: 2 bridges). b: A relational digraph in which pairs of nodes
are connected by directed edges. c: The same graph with undirected edges, form-
ing a simple graph. d: A self-looped undirected multigraph. Although all edges
between pairs of distinct nodes are unweighted, a self-loop is considered as either
two directed links each carrying a weight of 1 or, as shown here, as one undirected
edge of weight 2. e: A self-looped directed multigraph in which each self-loop is
represented by two directed edges (indicated by double-arrows) each carrying a
weight of 1. Graphs a, c and d are examples of undirected graphs, while b and e
are digraphs. Due to the presence of weighted edges and self-loops, graphs dis-
played in a, d and e require a weight matrix (2.3) for their correct representation,
while in the cases displayed in b and c, an adjacency matrix (2.4) suffices.
In terms of the weight matrix, we can thus define a graph as being directed
if ∃ (i, j) ∈ E with i, j ∈ N : wi j , 0 ∧ w ji = 0. On the other hand, the
Königsberg’s bridge graph is a classical example of an undirected graph, as
here for each (i, j) ∈ E also ( j, i) ∈ E such that wi j = w ji for all i, j ∈ N
and each unordered pair (i, j) : wi j , 0 defines an undirected edge (Fig. 2.1a).
On closer inspection, we furthermore notice that in this particular graph none
of its nodes serves both as source and target of an edge. A graph for which
@ i ∈ N : (i, i) ∈ E is called a not self-looped graph, and a self-looped graph if
and only if there exists at least one self-loop, that is, an edge which connects a
node with itself such that wii , 0 for at least one i ∈ N.
Following historically established terminology, the assignment of a weight
wi j ∈ Q to each adjacency relation (i, j) ∈ E of a directed graph with wi j , 1
Table 2.1 Major graph types and their weight matrix representation.
directed ∃i, j ∈ N : wi j , 0 ∧ w ji = 0 undirected ∀i, j ∈ N : wi j = w ji
self-looped ∃i ∈ N : wii , 0 not self-looped ∀i ∈ N : wii = 0
weighted ∃(i, j) ∈ E : wi j , 1 unweighted ∀i, j ∈ N : wi j ∈ B
multidigraph ∀i, j ∈ N: wi j ∈ N0 ∧ ∃i, j ∈ N : wi j , w ji ∧ ∃(i, j) ∈ E : wi j > 1
multigraph ∀i, j ∈ N: wi j ∈ N0 ∧ wi j = w ji ∧ ∃(i, j) ∈ E : wi j > 1
relational digraph ∀i, j ∈ N: wi j ∈ B ∧ ∀i ∈ N : wii = 0
simple graph ∀i, j ∈ N: wi j ∈ B ∧ wi j = w ji ∧ ∀i ∈ N : wii = 0
for at least one (i, j) ∈ E defines a weighted digraph. A first special case of
a weighted digraph is obtained if we set wi j = w ji for each (i, j) ∈ E. In this
case, the graph is undirected and simply called a weighted graph. By further
restricting each non-zero element of the weight matrix to the set of natural
numbers N, we have
wi j ∈ N iff (i, j) ∈ E,
(
wi j : N × N → N0 with (2.3)
wi j = 0 otherwise,
and the corresponding graph is called a directed multigraph, or multidigraph,
if ∃(i, j) ∈ E : wi j > 1, and an undirected multigraph or simply multigraph if,
in addition, its weight matrix is symmetric. In a multigraph or multidigraph,
each adjacency relation (i, j) ∈ E with weight wi j > 1 can be viewed as repre-
senting either wi j edges of weight 1 linking nodes i and j, or a single multiedge
of weight wi j . Recalling the weight matrix (2.2), the graph of Königsberg’s
bridges provides a classical example of such a multigraph (Fig. 2.1a). A final
special type of weighted digraph we will entertain here is realised if each ele-
ment of the weight matrix of a not self-looped digraph maps a pair of nodes
into the Boolean set B := {0, 1}. The weight matrix is then commonly referred
to as adjacency matrix a with elements
1 iff (i, j) ∈ E and i , j,
(
ai j : N × N → B with ai j = (2.4)
0 otherwise,
and defines a relational digraph (Fig. 2.1b) or, in the undirected case with
symmetric adjacency matrix ai j = a ji , a simple graph (Fig. 2.1c).
It is this type of simple – that is, undirected, not self-looped and single-
edged – graph which most prominently features in the contemporary applied
graph-theoretical literature. Indeed, and somewhat unfortunately, the over-
whelming majority of notions and measures introduced into graph theory since
Euler’s time is attuned to utilising the adjacency matrix (2.4), a simplification
which, arguably, only rarely is justified when dealing with graph-theoretical
descriptions of real-world phenomena. For that reason, the presentation in the
remainder of this book will focus, whenever possible, on the most general, i.e.
weighted, directed and self-looped, type of graph formalised in Definition 2.1
together with equation (2.1), and highlight, whenever necessary, results and
subtleties for the specific types of graphs introduced above as special cases.
One of these subtleties concerns the incorporation of nodes with self-loops
in a graph. In order to explore the complication associated with the presence
of self-loops, let us first consider the case of an unweighted undirected graph.
Here, for each ordered pair (i, j) ∈ E of distinct nodes, ( j, i) must also be an
element of E. If the source and target node of an edge are identical and each
edge carries a weight of 1, we obtain for the associated adjacency relations
which suggests that each pair (i, i) must be counted twice in E. Although it is
possible to account for this doubling in the mathematical definition of graph-
theoretical quantities by treating self-loops separately from undirected edges
between distinct nodes, it is often more convenient to represent a self-loop as
one edge and assign to it a weight of two, as illustrated above and exemplified
in Fig. 2.1d. However, such an approach carries the subtle disadvantage that
unweighted undirected graphs with self-loops can no longer be represented by
employing an adjacency matrix (2.4), as in this case aii = 2 < B if (i, i) ∈ E, and
thus limits the commonly used notion of an adjacency matrix to an unweighted
undirected graph without self-loops; that is, simple graphs. In the presence of
self-loops, such graphs are multigraphs and therefore require a weight matrix
with wii = 2 ∀ (i, i) ∈ E for their representation.
Naturally, this convention can be generalised to incorporate undirected multi-
graphs with self-loops and arbitrary edge weights, in which case wii = 2N0 for
all i ∈ N, as well as undirected graphs with edges taking weights in Q. More-
over, and most importantly, the above approach with its associated conditions
imposed on the diagonal elements of the weight matrices (2.1) and (2.3) also
applies to directed graphs if we consider that, by definition, each self-looped
node serves as both source and target of an adjacency relation and, thus, each
self-loop necessarily corresponds to an undirected edge (Fig. 2.1e). Indeed, in
a directed graph, an undirected adjacency relation of weight w between two
distinct nodes i and j can be generated by either adding a reciprocal directed
edge to E (symmetrisation) or discarding the direction of the given edge which,
again, necessitates the addition of a reciprocal directed edge to E (doubling).
In both cases, a self-loop is conceptually consistent with the presence of two
directed edges each carrying a weight of w, or one undirected edge of weight

2w, as illustrated in
w w 2w
i=j
i j ⇒ i ⇒ i
w
⇒ w w
i j ⇒
w w 2w.
i=j
i j ⇒ i ⇒ i
w w
It is important to stress that, unfortunately, in many applications of graph the-

ory this subtle conceptual difference between self-loops and edges between
distinct nodes is largely ignored in order to remain with an adjacency matrix
as a representational vessel. This is typically addressed by discarding self-
loops from graph models – that is, by modifying the very structural makeup
of a given graph – altogether. The convention introduced above and adhered
to throughout this book avoids such a heavy-handed approach, but does so at
the expense of surrendering the commonly used notion of an adjacency matrix
ai j to the consideration of relational digraphs and simple graphs only, that is,
unweighted directed and undirected graphs without self-loops.
2.1.2 Classes of Graphs

In addition to the various types of graphs introduced above and distinguished
by general properties of the set of edges E and the weights assigned to each of
its elements, we can also identify a number of graph classes, some of which
are illustrated in Fig. 2.2 and listed in Table 2.2, by considering more specific
structural characteristics shared between graphs. An important first example of
such a graph class is given by the complete graph, commonly defined as a fully
connected simple graph in which each node establishes edges to all other nodes
such that ai j = 1 for all i, j ∈ N : i , j (Fig. 2.2a). This notion of a complete
graph can certainly be generalised to incorporate relational digraphs, in which
case each pair of distinct nodes is connected by two edges, one in each direc-
tion, to form a complete digraph. Moreover, as the definition of a complete
graph requires only the presence of edges between each pair of distinct nodes
in N, this class can be further expanded to include not self-looped weighted
graphs. In what follows, we will therefore refer to a complete digraph as a
generally weighted graph for which wi j , 0 for all i, j ∈ N : i , j, and a com-
plete graph if, in addition, ∀i, j ∈ N : i , j we have wi j = w ji . We recognise,
a b c d
1
2
e f g h
Figure 2.2 Representative examples of various classes of graphs. Illustrated are

a complete graph (a), a 3-regular graph (b), a disconnected graph (c), a weakly
connected digraph (d), a bipartite graph (also known as a utility graph) (e), the
planar graph of Königsberg’s bridges (f), a complete planar (g) and a complete
non-planar graph (h). With the exception of the weakly connected digraph and the
weighted Königsberg bridge graph, the displayed examples are all simple graphs.
In the case of the weakly connected digraph (d), nodes 1 and 3 are strongly con-
nected (black solid), while nodes 1 and 2 establish only a weak connection (black
dashed). The grey areas in (f) and (g) mark the different faces of the illustrated
planar graphs.
however, that such a generalisation runs into conflict somewhat with the type
of directed graphs introduced in the previous section, a situation which will be
salvaged in Section 2.2.3 with the introduction of the type of weight-directed
graph. An illustrative example of a complete digraph is given by a small collab-
oration network of scientists in which each individual member will contribute,
depending on expertise and relevance, to further the other members’ research.
As a representative example, for a complete graph we find the network of grav-
itational interactions between galaxies, as each galaxy exerts an attractive force
on all others while the strength of this force and, thus, the weight of each edge
in the associated graph will vary between different pairs as a function of the
galaxies’ spatial distance.
Inherently related to the notion of a complete graph is that of a graph com-
plement, also called a graph edge-complement or graph inverse [53]. Given a
relational digraph G = (N, E), the complement of G is typically defined as a
relational digraph Ḡ = (N̄, Ē) with the same set of nodes N̄ = N and a set
of edges Ē such that for all i, j ∈ N : i , j we have (i, j) ∈ Ē if and only if
(i, j) < E. If ai j denotes the adjacency matrix of G, the adjacency matrix of its
complement is thus given by āi j = 1 − ai j for all i, j ∈ N : i , j and āii = 0.
With this, the sum of G and its complement Ḡ then yields a complete digraph
formed by the set of nodes N. Here, the graph sum of two relational digraphs
G1 = (N, E1 ) and G2 = (N, E2 ) sharing the same set of nodes N and edge sets
E1 , E2 : E1 ∩E2 = ∅ is defined as the relational digraph G1 5 G2 := (N, E1 ∪E2 ).
In case two given graphs share at least one common edge or at least one of the
graphs is weighted, their graph sum will necessarily yield a weighted graph.
This generalisation allows naturally extending the notion of a graph comple-
ment to include not self-looped weighted graphs, such that in the remainder of
this book we will refer to a graph complement Ḡ of a not self-looped graph or
digraph G if their sum G 5 Ḡ respectively forms a complete graph or digraph.
Another important graph class is comprised of simple graphs in which each
node establishes the same number of edges to other nodes in the graph, and
is called a regular graph or k-regular graph (Fig. 2.2b). Mathematically, the
requirement defining a k-regular graph can be expressed as
X
∀i ∈ N: ai j = k = const,
j∈N
where the sum runs over all nodes in the graph. As in the case of a complete
graph, this notion also can be generalised to cover relational digraphs, in which
case each node is required to have the same number of outgoing and incoming
directed edges, that is,
X X
∀i ∈ N: ai j = a ji = k = const .
j∈N j∈N
Such a graph is then called a k-regular digraph. Finally, if the edges of a not
self-looped directed graph are weighted, we can formulate
X X
∀i ∈ N: wi j = w ji = w = const
j∈N j∈N
as a necessary condition defining a w-regular digraph and, in the case of sym-

metric wi j , a w-regular graph, although these generalising notions are not yet
employed in the wider graph-theoretical literature. We note that each complete
graph or digraph with node set N and unweighted edges is also a k-regular
graph or digraph with k = |N| − 1, while it is generally not true that a complete
weighted graph is also a w-regular graph.
A third major class of graphs concerns the existence of connections between
possible pairs of nodes. A simple graph in which each node can be reached
from at least one other node by following existing edges is called a connected
graph. Conversely, in a disconnected graph, there exists at least one pair of
nodes which cannot be linked through edges forming the graph. The pres-
ence of an isolated node – that is, a node without associated adjacency rela-
tions – is a trivial example of a disconnected graph; while the disconnected
graph depicted in Fig. 2.2c is comprised of two independent smaller graphs.
Table 2.2 Important graph classes and their weight matrix representation.
Class Typea Representation
d ud w uw
complete • • • • ∀i, j ∈ N, i , j : wi j , 0
k-regular • • • ∀i, j ∈ N P
: wi j ∈ B ∧P ∀i ∈ N : j∈N wi j = j∈N w ji = k
P P
w-regular • • • ∀i ∈ N : j∈N wi j = j∈N w ji = w
connected • • • ∀i, j ∈ N, i , j ∃ jl ∈ N, l ∈ [1, k], j1 = i, jk+1 = j : w jl jl+1 , 0
bipartite • • • • ∃N 0 ⊂ N, N 0 , ∅ such that
∀i, j ∈ N 0 : wi j = 0 ∧ ∀i, j ∈ N \ N 0 : wi j = 0
and ∃i ∈ N 0 , j ∈ N \ N 0 : wi j , 0 ∨ w ji , 0
biclique • • • • ∃N 0 ⊂ N, N 0 , ∅ such that
∀i, j ∈ N 0 : wi j = 0 ∧ ∀i, j ∈ N \ N 0 : wi j = 0
and ∀i ∈ N 0 , j ∈ N \ N 0 : wi j , 0 ∧ w ji , 0
a Graph types (see Table 2.1) to which the given representation applies. Here d, ud, w and uw
denote directed, undirected, weighted and unweighted graphs respectively.
Naturally, the notion of a connected graph can again be expanded to include

undirected weighted graphs, with the graph of Königsberg’s bridges being an
illustrative example of a connected multigraph. In the case of relational or,
more generally, weighted directed graphs, however, the situation is slightly
more involved and requires the notion of strongly and weakly connected nodes.
In the example illustrated in Fig. 2.2d, both nodes 2 and 3 can be reached from
node 1 by following a path of directed edges, but the converse is true only for
node 3. We refer to nodes 1 and 3 as being strongly connected, while nodes
1 and 2 are not. However, node 1 can be reached from node 2 if the directed
edge connecting both nodes is complemented by a directed edge pointing in the
opposite direction, or is simply replaced by an undirected edge, thus render-
ing nodes 1 and 2 weakly connected. If each pair of nodes in a directed graph
is strongly connected, the graph is called a strongly connected digraph; while
in a weakly connected digraph, there exists at least one pair of only weakly
connected nodes. As each strongly connected node pair is also weakly con-
nected, all strongly connected digraphs are necessarily and sufficiently weakly
connected digraphs, while the opposite is not true.
With complete, regular and connected graphs, we certainly have not yet
exhausted the long list of classes which distinguish possible connectivity pat-
terns in graphs. Indeed, by considering patterns of connections between sub-
sets of nodes, many more graph classes can be defined which, in some cases,
find important applications as models in the description of real-world phe-
nomena. As an illustrative example, let us focus here on one of these more
specific classes. If in a given simple graph G = (N, E), two distinct subsets
N1 and N2 of nodes can be identified such that adjacency relations exist only
between nodes which reside in different subsets, that is, if N = N1 ∪ N1 and
for all (i, j) ∈ E : (i ∈ N1 ∧ j ∈ N2 ) ∨ (i ∈ N2 ∧ j ∈ N1 ), the graph is
called a bipartite graph, and a complete bipartite graph, or simply biclique,
in the case all nodes in N1 are connected to all nodes in set N2 (Fig. 2.2e).
This notion applies to both weighted and unweighted as well as directed and
undirected not self-looped graphs. Due to their conceptional simplicity, bipar-
tite graphs find widespread use not only as models of interactions between
groups of objects whose members share common characteristics [10], but also
in the modelling of feedforward networks such as artificial neural networks
[46, 75] and a variety of biological neural networks along with their functional
properties [93].
The final and, historically, arguably most intriguing class of graphs to be
introduced in this abridged presentation of classical graph theory is comprised
of networks which allow for a planar embedding. Such an embedding is defined
as a visual representation of a given graph on a two-dimensional surface such
that edges intersect only at the nodes to which they connect. A simple graph
which allows for such a representation is then called a planar graph, and it
is called a non-planar graph if no such planar embedding exists. The graph of
Königsberg’s bridges (Fig 2.2f) and the complete simple graph with four nodes
depicted in Fig 2.2g are planar, while the bipartite graph shown in Fig 2.2e and
complete simple graph with five nodes in Fig 2.2h are examples of non-planar
graphs. Indeed, the latter two graphs are instrumental in deciding whether a
given simple graph allows for a planar embedding or not, a remarkable insight
formalised in Kuratowski’s theorem [72]. Unfortunately, a just presentation
of this historically significant result along with its many proofs [64, 104] is
beyond the scope of this brief introduction, and we must refer the interested
reader to the excellent textbook by Nishizeki and Chiba [89].
Instead, let us take a closer look at another important contribution which
not only resides at the very heart of the theory of planar graphs but unearthed
one of the most beautiful equations in the theory of finite graphs altogether,
Euler’s formula. Owing to its definition, a planar graph G = (N, E) necessarily
divides the plane into distinct regions or faces, each encircled by a subset of
edges forming a cycle. Let us denote with F the set of these faces, including
the unbounded region outside the graph, such that the cardinality |F | yields the
number of faces. Euler’s formula then states that the number of nodes, given by
|N|, the number of edges, given by |E|, and the number of faces |F | are subject
to the strict identity
|N| − |E| + |F | = 2 (2.5)

for any finite connected planar graph G. In the graph of Königsberg’s bridges
(Fig. 2.2f), for example, we find |N| = 4, |E| = 5 and |F | = 3. For the complete
graph with |N| = 4 depicted in Fig. 2.2g, we have |E| = 6 and |F | = 4. Thus
both examples of planar graphs satisfy (2.5).
Due to the historical significance of Euler’s formula, and its importance for
many real-world applications, we will present here a simple proof demonstrat-
ing its validity. To that end, let us first introduce the notion of a tree graph as
a simple planar graph which has only one face, and show by induction on the
number of nodes that (2.5) holds for this type of graph. If a tree graph contains
only one node and no edge, Euler’s formula clearly applies. We now assume
that (2.5) is true for all tree graphs with |N| ≥ 1 nodes. Following its defini-
tion, each such graph must have |N| − 1 edges. Let T 0 = (N 0 , E0 ) be a tree
graph with |N 0 | = |N| + 1 nodes and |E0 | edges. Starting at any of its nodes, we
then follow an arbitrary path comprised of edges between connected nodes but
without revisiting nodes. As a tree graph has only one face, and thus no cycle,
such a path will not return to the initially chosen node but instead must neces-
sarily terminate at a node which is connected by only one edge. By removing
the path’s final node along with its sole edge, we then construct a new tree
T 00 = (N 00 , E00 ) with |N 00 | = |N 0 | − 1 = |N| nodes and |E00 | = |E0 | − 1 edges. By
the induction hypothesis, as T 00 is a tree graph with |N| nodes, we must have
|E00 | = |N| − 1, from which |E0 | = |N| for T 0 and the validity of Euler’s formula
for trees with |N| + 1 nodes, thus for all tree graphs, follows.
For arbitrary planar graphs, the proof of Euler’s formula (2.5) follows a simi-
lar approach using induction on the number of edges. A connected planar graph
with |N| ≥ 1 nodes must have at least |N| − 1 edges. If the graph has exactly
|N| − 1 edges, it is a tree graph and, as shown above, Euler’s formula cer-
tainly applies. Assuming that (2.5) is true for graphs with |N| ≥ 1 nodes and
|E| ≥ |N| − 1 edges, we have F ≥ 1. Let G0 = (N 0 , E0 ) be a connected simple
graph with |N 0 | = |N| nodes and |E0 | = |E| + 1 edges. Then F > 1, and G must
have at least one cycle. By choosing any of these cycles and removing one of
the chosen cycle’s edges, we then construct a new graph G00 = (N 00 , E00 ) with
|N 00 | = |N 0 | = |N| nodes, |E00 | = |E0 | − 1 = |E| edges and two merged faces,
i.e. |F 00 | = |F 0 | − 1. By the induction hypothesis, as G00 is a connected planar
graph with |N| nodes and |E| edges, we have |N 00 |−|E00 |+|F 00 | = 2, from which
|N 0 | − (|E0 | − 1) + (|F 0 | − 1) = |N 0 | − |E0 | + |F 0 | = 2 for G0 and eventually the
validity of (2.5) for all finite connected planar graphs follows.
The theory of planar graphs undoubtedly forms the root of one of the his-
torically most influential branches of graph theory, from which a whole host
of important real-world applications was spawned. Certainly high among these
applications ranks the four-colour theorem, which states that four colours
Figure 2.3 The Platonic solids (tetrahedron, cube, octahedron, dodecahedron and
icosahedron) and their associated planar graphs (tetrahedral graph, cubical graph,
octahedral graph, dodecahedral graph and icosahedral graph). Using Euler’s for-
mula, it can easily be shown that only five Platonic solids are possible.
suffice to draw the faces of a planar graph in such a way that no adjacent faces
have the same colour [7, 8, 9], and which provides the mathematical founda-
tion for the accessible visual representation of contiguous regions on the many
maps we encounter in our everyday lives. Unfortunately, our necessarily brief
introduction to classical graph theory does not permit a worthy presentation of
this or other important results of planar graph theory. However, we will enter-
tain the intrigued reader with a beautiful yet simple graph-theoretical proof of
another historically significant mathematical theorem which states that there
can be only five Platonic solids: the tetrahedron, cube, octahedron, dodeca-
hedron and icosahedron (Fig. 2.3, top). Platonic solids are defined as convex
polyhedra whose faces are comprised of congruent convex regular polygons.
These iconic geometrical objects not only were the subject of extensive studies
in antiquity, but they played an important role in a variety of fields throughout
history. It was Euclid in his Elements who first showed rigorously that only
five such solids are possible [41].
Here we will present a simple proof utilising Euler’s formula (2.5). To that
end, we first represent each solid as a planar graph G = (N, E), called a
Platonic graph, by associating a node with each corner of the solid, and a
graph edge with each of the solid’s edges (Fig. 2.3). By construction, the
faces of the Platonic graphs correspond to the faces of the Platonic solid,
with each face of a given solid being encircled by the same number of edges
NeF , and each node establishing the same number of adjacency relations NeN .
As furthermore, each edge of our Platonic graphs belongs to exactly two dis-
tinct faces, we have NeF |F | = 2|E|, and as each edge links two distinct nodes,
we find NeN |N| = 2|E|. Employing Euler’s formula and some simple algebraic
Table 2.3 Platonic graphs and their associated five Platonic solids.
Platonic solid Platonic graph NeN NeF |N| |E| |F |
tetrahedron tetrahedral graph 3 3 4 6 4
cube cubical graph 3 4 8 12 6
octahedron octahedral graph 4 3 6 12 8
dodecahedron dodecahedral graph 3 5 20 30 12
icosahedron icosahedral graph 5 3 12 30 20
manipulations, which we leave as an exercise for the reader, we are then left
with the diophantine equation
1 1 1 1
N
+ F = +
Ne Ne |E| 2
as a necessary condition governing the existence of Platonic graphs. By noting
that NeN ≥ 3 and NeF ≥ 3, and that Platonic graphs must have at least one
edge such that |E| ≥ 1, we find that this equation can only have the five integer
solutions for NeN , NeF and |E| listed in Table 2.3, thus demonstrating that only
five Platonic graphs, hence only five Platonic solids, are possible.
With this brief excursion into the realms of our antique forefathers, we will
conclude our brief introduction to the basic notions and conceptual foundation
at the heart of the theory of graphs and networks. These preliminaries should
suffice as a framework for the material presented in the remainder of this book.
In the next section, we will return to general weighted digraphs and introduce,
on more rigorous grounds, some of the concepts and concrete quantities which
are commonly employed throughout the applied graph-theoretical literature to
characterise and measure graphs.
2.2 Graph Measures

Although the classical notion of a graph introduced in Definition 2.1 along with
the small selection of types and classes exemplified in the last section already
allow for a coarse distinction of a graph’s structural makeup, the delineation
and quantification of finer structural characteristics certainly require a more
detailed assessment. Indeed, since Euler’s time, a plethora of such quantities
has emerged, and hundreds of measures are now being commonly employed
throughout the contemporary literature in the comparative analysis of graphs
and real-world networks. A thorough presentation of this ever-growing host
of graph-theoretical measures, however, lies beyond the scope of this book,
2.2 Graph Measures 27
and we must refer the interested reader to the recent textbook of Newman
[86] or the excellent review by Boccaletti and colleagues [20] for a more
comprehensive and detailed overview. Instead, we will focus here on only
the most basic and prominent measures of relevance for presentation in later
chapters, and we will categorise these measures into general measures, degree-
based, weight-based and distance-based measures, as well as measures asso-
ciated with network motifs and graph spectra. With this necessarily abridged
presentation, we nevertheless hope to motivate the serious student to further
explore this vast and fluent branch of graph theory and its countless real-world
applications.
2.2.1 General Measures

In Section 2.1, we already encountered some of the graph-theoretical notions
associated with a general characterisation of a given network, its size, type and
class. To formalise these notions and set the foundation for a mathematically
more rigorous framework, let us begin by introducing a number of measures
with values in the Boolean set B which allow us to distinguish whether a given
graph G = (N, E) is directed,
1 iff ∃ i, j ∈ N, (i, j) ∈ E : wi j , 0 ∧ w ji = 0,
(
λd ∈ B : λd := (2.6)
0 otherwise,
is weighted,
iff ∃ i, j ∈ N, (i, j) ∈ E : wi j , 1,
(
1
λw ∈ B : λw := (2.7)
0 otherwise,
or is self-looped
1 iff ∃ i ∈ N, (i, i) ∈ E : wii , 0,

(
λl ∈ B : λl := (2.8)
0 otherwise.
A simple graph, for example, is then a graph for which λd = 0, λw = 0 and
λl = 0, the undirected multigraph of Königsberg’s bridges (Fig. 2.1a) is a
network with λd = 0, λw = 1 and λl = 0, and the directed graph illustrated in
Fig. 2.1e is one with λd = 1, λw = 1 and λl = 1. These Boolean measures thus
present a more formal means with which to identify the graph types listed in
Table 2.1.
Returning to Definition 2.1, we will henceforth refer to the cardinality of the
set of nodes N as the number of nodes
Nn ∈ N : Nn := |N| (2.9)
Another random document with
no related content on Scribd:
*** END OF THE PROJECT GUTENBERG EBOOK THE MINISTER
HAD TO WAIT ***
Updated editions will replace the previous one—the old editions will
be renamed.
Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States copyright in
these works, so the Foundation (and you!) can copy and distribute it
in the United States without permission and without paying copyright
royalties. Special rules, set forth in the General Terms of Use part of
this license, apply to copying and distributing Project Gutenberg™
electronic works to protect the PROJECT GUTENBERG™ concept
and trademark. Project Gutenberg is a registered trademark, and
may not be used if you charge for an eBook, except by following the
terms of the trademark license, including paying royalties for use of
the Project Gutenberg trademark. If you do not charge anything for
copies of this eBook, complying with the trademark license is very
easy. You may use this eBook for nearly any purpose such as
creation of derivative works, reports, performances and research.
Project Gutenberg eBooks may be modified and printed and given
away—you may do practically ANYTHING in the United States with
eBooks not protected by U.S. copyright law. Redistribution is subject
to the trademark license, especially commercial redistribution.
START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE
PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK
To protect the Project Gutenberg™ mission of promoting the free

distribution of electronic works, by using or distributing this work (or
any other work associated in any way with the phrase “Project
Gutenberg”), you agree to comply with all the terms of the Full
Project Gutenberg™ License available with this file or online at
www.gutenberg.org/license.
Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand, agree
to and accept all the terms of this license and intellectual property
(trademark/copyright) agreement. If you do not agree to abide by all
the terms of this agreement, you must cease using and return or
destroy all copies of Project Gutenberg™ electronic works in your
possession. If you paid a fee for obtaining a copy of or access to a
Project Gutenberg™ electronic work and you do not agree to be
bound by the terms of this agreement, you may obtain a refund from
the person or entity to whom you paid the fee as set forth in
paragraph 1.E.8.
1.B. “Project Gutenberg” is a registered trademark. It may only be

used on or associated in any way with an electronic work by people
who agree to be bound by the terms of this agreement. There are a
few things that you can do with most Project Gutenberg™ electronic
works even without complying with the full terms of this agreement.
See paragraph 1.C below. There are a lot of things you can do with
Project Gutenberg™ electronic works if you follow the terms of this
agreement and help preserve free future access to Project
Gutenberg™ electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright law in
the United States and you are located in the United States, we do
not claim a right to prevent you from copying, distributing,
performing, displaying or creating derivative works based on the
work as long as all references to Project Gutenberg are removed. Of
course, we hope that you will support the Project Gutenberg™
mission of promoting free access to electronic works by freely
sharing Project Gutenberg™ works in compliance with the terms of
this agreement for keeping the Project Gutenberg™ name
associated with the work. You can easily comply with the terms of
this agreement by keeping this work in the same format with its
attached full Project Gutenberg™ License when you share it without
charge with others.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the terms
of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.
1.E. Unless you have removed all references to Project Gutenberg:
1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project Gutenberg™
work (any work on which the phrase “Project Gutenberg” appears, or
with which the phrase “Project Gutenberg” is associated) is
accessed, displayed, performed, viewed, copied or distributed:
This eBook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this eBook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.
1.E.2. If an individual Project Gutenberg™ electronic work is derived

from texts not protected by U.S. copyright law (does not contain a
notice indicating that it is posted with permission of the copyright
holder), the work can be copied and distributed to anyone in the
United States without paying any fees or charges. If you are
redistributing or providing access to a work with the phrase “Project
Gutenberg” associated with or appearing on the work, you must
comply either with the requirements of paragraphs 1.E.1 through
1.E.7 or obtain permission for the use of the work and the Project
Gutenberg™ trademark as set forth in paragraphs 1.E.8 or 1.E.9.
1.E.3. If an individual Project Gutenberg™ electronic work is posted

with the permission of the copyright holder, your use and distribution
must comply with both paragraphs 1.E.1 through 1.E.7 and any
additional terms imposed by the copyright holder. Additional terms
will be linked to the Project Gutenberg™ License for all works posted
with the permission of the copyright holder found at the beginning of
this work.
1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files containing a
part of this work or any other work associated with Project
Gutenberg™.
1.E.5. Do not copy, display, perform, distribute or redistribute this

electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1 with
active links or immediate access to the full terms of the Project
Gutenberg™ License.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if you
provide access to or distribute copies of a Project Gutenberg™ work
in a format other than “Plain Vanilla ASCII” or other format used in
the official version posted on the official Project Gutenberg™ website
(www.gutenberg.org), you must, at no additional cost, fee or expense
to the user, provide a copy, a means of exporting a copy, or a means
of obtaining a copy upon request, of the work in its original “Plain
Vanilla ASCII” or other form. Any alternate format must include the
full Project Gutenberg™ License as specified in paragraph 1.E.1.
1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™ works
unless you comply with paragraph 1.E.8 or 1.E.9.
1.E.8. You may charge a reasonable fee for copies of or providing

access to or distributing Project Gutenberg™ electronic works
provided that:
• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt that
s/he does not agree to the terms of the full Project Gutenberg™
License. You must require such a user to return or destroy all
copies of the works possessed in a physical medium and
discontinue all use of and all access to other copies of Project
Gutenberg™ works.
• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.E.9. If you wish to charge a fee or distribute a Project Gutenberg™

electronic work or group of works on different terms than are set
forth in this agreement, you must obtain permission in writing from
the Project Gutenberg Literary Archive Foundation, the manager of
the Project Gutenberg™ trademark. Contact the Foundation as set
forth in Section 3 below.
1.F.
1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on, transcribe
and proofread works not protected by U.S. copyright law in creating
the Project Gutenberg™ collection. Despite these efforts, Project
Gutenberg™ electronic works, and the medium on which they may
be stored, may contain “Defects,” such as, but not limited to,
incomplete, inaccurate or corrupt data, transcription errors, a
copyright or other intellectual property infringement, a defective or
damaged disk or other medium, a computer virus, or computer
codes that damage or cannot be read by your equipment.
1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except

for the “Right of Replacement or Refund” described in paragraph
1.F.3, the Project Gutenberg Literary Archive Foundation, the owner
of the Project Gutenberg™ trademark, and any other party
distributing a Project Gutenberg™ electronic work under this
agreement, disclaim all liability to you for damages, costs and
expenses, including legal fees. YOU AGREE THAT YOU HAVE NO
REMEDIES FOR NEGLIGENCE, STRICT LIABILITY, BREACH OF
WARRANTY OR BREACH OF CONTRACT EXCEPT THOSE
PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE THAT THE
FOUNDATION, THE TRADEMARK OWNER, AND ANY
DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE LIABLE
TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL,
PUNITIVE OR INCIDENTAL DAMAGES EVEN IF YOU GIVE
NOTICE OF THE POSSIBILITY OF SUCH DAMAGE.
1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

discover a defect in this electronic work within 90 days of receiving it,
you can receive a refund of the money (if any) you paid for it by
sending a written explanation to the person you received the work
from. If you received the work on a physical medium, you must
return the medium with your written explanation. The person or entity
that provided you with the defective work may elect to provide a
replacement copy in lieu of a refund. If you received the work
electronically, the person or entity providing it to you may choose to
give you a second opportunity to receive the work electronically in
lieu of a refund. If the second copy is also defective, you may
demand a refund in writing without further opportunities to fix the
problem.
1.F.4. Except for the limited right of replacement or refund set forth in
paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of damages.
If any disclaimer or limitation set forth in this agreement violates the
law of the state applicable to this agreement, the agreement shall be
interpreted to make the maximum disclaimer or limitation permitted
by the applicable state law. The invalidity or unenforceability of any
provision of this agreement shall not void the remaining provisions.
1.F.6. INDEMNITY - You agree to indemnify and hold the
Foundation, the trademark owner, any agent or employee of the
Foundation, anyone providing copies of Project Gutenberg™
electronic works in accordance with this agreement, and any
volunteers associated with the production, promotion and distribution
of Project Gutenberg™ electronic works, harmless from all liability,
costs and expenses, including legal fees, that arise directly or
indirectly from any of the following which you do or cause to occur:
(a) distribution of this or any Project Gutenberg™ work, (b)
alteration, modification, or additions or deletions to any Project
Gutenberg™ work, and (c) any Defect you cause.
Section 2. Information about the Mission of

Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new computers.
It exists because of the efforts of hundreds of volunteers and
donations from people in all walks of life.
Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project Gutenberg™’s
goals and ensuring that the Project Gutenberg™ collection will
remain freely available for generations to come. In 2001, the Project
Gutenberg Literary Archive Foundation was created to provide a
secure and permanent future for Project Gutenberg™ and future
generations. To learn more about the Project Gutenberg Literary
Archive Foundation and how your efforts and donations can help,
see Sections 3 and 4 and the Foundation information page at
www.gutenberg.org.
Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-profit
501(c)(3) educational corporation organized under the laws of the
state of Mississippi and granted tax exempt status by the Internal
Revenue Service. The Foundation’s EIN or federal tax identification
number is 64-6221541. Contributions to the Project Gutenberg
Literary Archive Foundation are tax deductible to the full extent
permitted by U.S. federal laws and your state’s laws.
The Foundation’s business office is located at 809 North 1500 West,

Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
to date contact information can be found at the Foundation’s website
and official page at www.gutenberg.org/contact
Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission of
increasing the number of public domain and licensed works that can
be freely distributed in machine-readable form accessible by the
widest array of equipment including outdated equipment. Many small
donations ($1 to $5,000) are particularly important to maintaining tax
exempt status with the IRS.
The Foundation is committed to complying with the laws regulating

charities and charitable donations in all 50 states of the United
States. Compliance requirements are not uniform and it takes a
considerable effort, much paperwork and many fees to meet and
keep up with these requirements. We do not solicit donations in
locations where we have not received written confirmation of
compliance. To SEND DONATIONS or determine the status of
compliance for any particular state visit www.gutenberg.org/donate.
While we cannot and do not solicit contributions from states where

we have not met the solicitation requirements, we know of no
prohibition against accepting unsolicited donations from donors in
such states who approach us with offers to donate.
International donations are gratefully accepted, but we cannot make

any statements concerning tax treatment of donations received from
outside the United States. U.S. laws alone swamp our small staff.
Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.
Section 5. General Information About Project

Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could be
freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose network of
volunteer support.
Project Gutenberg™ eBooks are often created from several printed

editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.
This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.

PDF The Mathematics of Finite Networks An Introduction To Operator Graph Theory 1st Edition Michael Rudolph Download

Uploaded by

Copyright:

Available Formats

PDF The Mathematics of Finite Networks An Introduction To Operator Graph Theory 1st Edition Michael Rudolph Download

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PDF The Mathematics of Finite Networks An Introduction To Operator Graph Theory 1st Edition Michael Rudolph Download

Uploaded by

Copyright:

Available Formats

Full download test bank at ebookmeta.

The Mathematics of Finite Networks An Introduction

Download More ebooks from https://ebookmeta.com

Spatial Networks: A Complete Introduction: From Graph

Discrete Mathematics with Graph Theory 1st Edition

Graph and Network Theory: An Applied Approach using

Introduction to Graph Theory: With Solutions to

An Introduction to Chemistry Michael Mosher

A Walk Through Combinatorics An Introduction to

Qualitative Theory of ODEs: An Introduction to

An Undergraduate Introduction to Financial Mathematics

m i c h a e l r u d o l p h is a mathematical physicist of the French National Centre

Cambridge University Press is part of the University of Cambridge.

PART I OPERATOR GRAPH THEORY

5.3 Watts–Strogatz Random Graphs 205

and we have termed this framework operator graph theory. Irrespective of

through their kindness, friendship and unconditional support the strength to

Figure 1.1 The bridges of Königsberg and their graph-theoretical representation.

theory by Newman [86], and by doing so challenge, somewhat self-servingly,

this very rejection of a continuous, thus analytic, description of real-world phe-

OPERATOR GRAPH THEORY

What do the bridges of Königsberg, synaptically connected neurons in our

2.1 Graphs and Networks

inspired the mathematical field of network studies, or graph theory, we find

Definition 2.1 (finite graph) A finite network or graph G is an ordered tuple

2.1.1 Types of Graphs

directed edges each carrying a weight of w, or one undirected edge of weight

It is important to stress that, unfortunately, in many applications of graph the-

2.1.2 Classes of Graphs

Figure 2.2 Representative examples of various classes of graphs. Illustrated are

as a necessary condition defining a w-regular digraph and, in the case of sym-

Naturally, the notion of a connected graph can again be expanded to include

|N| − |E| + |F | = 2 (2.5)

2.2 Graph Measures

2.2.1 General Measures

1 iff ∃ i ∈ N, (i, i) ∈ E : wii , 0,

Creating the works from print editions not protected by U.S.

START: FULL LICENSE

To protect the Project Gutenberg™ mission of promoting the free

Section 1. General Terms of Use and

1.B. “Project Gutenberg” is a registered trademark. It may only be

1.E. Unless you have removed all references to Project Gutenberg:

1.E.1. The following sentence, with active links to, or other

1.E.2. If an individual Project Gutenberg™ electronic work is derived

1.E.3. If an individual Project Gutenberg™ electronic work is posted

1.E.4. Do not unlink or detach or remove the full Project

1.E.5. Do not copy, display, perform, distribute or redistribute this

1.E.7. Do not charge a fee for access to, viewing, displaying,

1.E.8. You may charge a reasonable fee for copies of or providing

• You provide a full refund of any money paid by a user who

• You provide, in accordance with paragraph 1.F.3, a full refund of

1.E.9. If you wish to charge a fee or distribute a Project Gutenberg™

1.F.1. Project Gutenberg volunteers and employees expend

1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except

1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

1.F.5. Some states do not allow disclaimers of certain implied

Section 2. Information about the Mission of

Volunteers and financial support to provide volunteers with the

Section 3. Information about the Project