Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx

Enumerative Combinatorics: Volume 2:
Second Edition Richard P. Stanley

Visit to download the full and correct content document:
https://ebookmass.com/product/enumerative-combinatorics-volume-2-second-edition-
richard-p-stanley-3/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Enumerative Combinatorics: Volume 2: Second Edition

Richard P. Stanley
https://ebookmass.com/product/enumerative-combinatorics-
volume-2-second-edition-richard-p-stanley-3/
Enumerative Combinatorics: Volume 2: Second Edition

Richard P. Stanley
https://ebookmass.com/product/enumerative-combinatorics-
volume-2-second-edition-richard-p-stanley-2/
Fetal and Neonatal Physiology, 2-Volume Set 6th Edition

Richard A. Polin
https://ebookmass.com/product/fetal-and-neonatal-
physiology-2-volume-set-6th-edition-richard-a-polin/
Principles of Bone Biology (2 Volume Set) 4th Edition

Edition John P. Bilezikian
https://ebookmass.com/product/principles-of-bone-
biology-2-volume-set-4th-edition-edition-john-p-bilezikian/
(eBook PDF) America: The Essential Learning Edition
(Second Edition) (Vol. Volume 2) 2nd Edition
https://ebookmass.com/product/ebook-pdf-america-the-essential-
learning-edition-second-edition-vol-volume-2-2nd-edition/
Gray's basic anatomy Second Edition Richard Lee Drake
https://ebookmass.com/product/grays-basic-anatomy-second-edition-
richard-lee-drake/
Handbook of Applied Hydrology, Second Edition Vijay P.

Singh
https://ebookmass.com/product/handbook-of-applied-hydrology-
second-edition-vijay-p-singh/
Benzel's Spine Surgery, 2-Volume Set: Techniques,

Complication Avoidance and Management 5th Edition
Michael P Steinmetz Md (Editor)
https://ebookmass.com/product/benzels-spine-surgery-2-volume-set-
techniques-complication-avoidance-and-management-5th-edition-
michael-p-steinmetz-md-editor/
Big Chance Cowboy Teri Anne Stanley [Stanley
https://ebookmass.com/product/big-chance-cowboy-teri-anne-
stanley-stanley/
C A M B R I D G E S T U D I E S I N A DVA N C E D M AT H E M AT I C S 2 0 8
Editorial Board
J. B E RTO I N, B. B O L L O B Á S , W. F U LTO N, B. K R A , I . M O E R D I J K ,
C . P R A E G E R , P. S A R NA K , B. S I M O N, B. TOTA RO
ENUMERATIVE COMBINATORICS
Volume 2
Richard Stanley’s two-volume basic introduction to enumerative combinatorics has
become the standard guide to the topic for students and experts alike. This thoroughly
revised second edition of volume two covers the composition of generating functions,
in particular the exponential formula and the Lagrange inversion formula, labelled and
unlabelled trees, algebraic, D-finite, and noncommutative generating functions, and
symmetric functions. The chapter on symmetric functions provides the only availa-
ble treatment of this subject suitable for an introductory graduate course and focusing
on combinatorics, especially the Robinson–Schensted–Knuth algorithm. An appendix
by Sergey Fomin covers some deeper aspects of symmetric functions, including jeu de
taquin and the Littlewood–Richardson rule. The exercises in the book play a vital role in
developing the material, and this second edition features over 400 exercises, including
159 new exercises on symmetric functions, all with solutions or references to solutions.
Richard P. Stanley is Emeritus Professor of Mathematics at the Massachusetts Institute

of Technology and an Arts and Sciences Distinguished Professor at the University of
Miami. He has written over 180 research articles and six books. Among Stanley’s many
distinctions are membership in the National Academy of Sciences (elected in 1995) and
the 2001 Leroy P. Steele Prize for Mathematical Exposition.
C A M B R I D G E S T U D I E S I N A DVA N C E D M AT H E M AT I C S
Editorial Board
J. B E RTO I N, B. B O L L O B Á S , W. F U LTO N, B. K R A , I . M O E R D I J K ,
C . P R A E G E R , P. S A R NA K , B. S I M O N, B. TOTA RO
All the titles listed below can be obtained from good booksellers or from Cambridge University
Press. For a complete series listing, visit www.cambridge.org/mathematics.
Already Published
171 J. Gough & J. Kupsch Quantum Fields and Processes
172 T. Ceccherini-Silberstein, F. Scarabotti & F. Tolli Discrete Harmonic Analysis
173 P. Garrett Modern Analysis of Automorphic Forms by Example, I
174 P. Garrett Modern Analysis of Automorphic Forms by Example, II
175 G. Navarro Character Theory and the McKay Conjecture
176 P. Fleig, H. P. A. Gustafsson, A. Kleinschmidt & D. Persson Eisenstein Series and
Automorphic Representations
177 E. Peterson Formal Geometry and Bordism Operators
178 A. Ogus Lectures on Logarithmic Algebraic Geometry
179 N. Nikolski Hardy Spaces
180 D.-C. Cisinski Higher Categories and Homotopical Algebra
181 A. Agrachev, D. Barilari & U. Boscain A Comprehensive Introduction to Sub-Riemannian
Geometry
182 N. Nikolski Toeplitz Matrices and Operators
183 A. Yekutieli Derived Categories
184 C. Demeter Fourier Restriction, Decoupling and Applications
185 D. Barnes & C. Roitzheim Foundations of Stable Homotopy Theory
186 V. Vasyunin & A. Volberg The Bellman Function Technique in Harmonic Analysis
187 M. Geck & G. Malle The Character Theory of Finite Groups of Lie Type
188 B. Richter Category Theory for Homotopy Theory
189 R. Willett & G. Yu Higher Index Theory
190 A. Bobrowski Generators of Markov Chains
191 D. Cao, S. Peng & S. Yan Singularly Perturbed Methods for Nonlinear Elliptic Problems
192 E. Kowalski An Introduction to Probabilistic Number Theory
193 V. Gorin Lectures on Random Lozenge Tilings
194 E. Riehl & D. Verity Elements of ∞-Category Theory
195 H. Krause Homological Theory of Representations
196 F. Durand & D. Perrin Dimension Groups and Dynamical Systems
197 A. Sheffer Polynomial Methods and Incidence Theory
198 T. Dobson, A. Malnič & D. Marušič Symmetry in Graphs
199 K. S. Kedlaya p-adic Differential Equations
200 R. L. Frank, A. Laptev & T. Weidl Schrödinger Operators:Eigenvalues and Lieb–Thirring
Inequalities
201 J. van Neerven Functional Analysis
202 A. Schmeding An Introduction to Infinite-Dimensional Differential Geometry
203 F. Cabello Sánchez & J. M. F. Castillo Homological Methods in Banach Space Theory
204 G. P. Paternain, M. Salo & G. Uhlmann Geometric Inverse Problems
205 V. Platonov, A. Rapinchuk & I. Rapinchuk Algebraic Groups and Number Theory, I (2nd
Edition)
206 D. Huybrechts The Geometry of Cubic Hypersurfaces
207 F. Maggi Optimal Mass Transport on Euclidean Spaces
208 R. P. Stanley Enumerative Combinatorics, II (2nd edition)
209 M. Kawakita Complex Algebraic Threefolds
210 D. Anderson & W. Fulton Equivariant Cohomology in Algebraic Geometry
“This is one of the great books; readable, deep and full of gems. It brings
algebraic combinatorics to life. I teach out of it and feel that if I can get my
students to ‘touch Stanley’ I have given them a gift for life.”
– Persi Diaconis, Stanford University
“It is wonderful to celebrate the completion of the second edition of Richard

Stanley’s Enumerative Combinatorics, one of the finest mathematical works
of all time. He has added nearly 200 exercises, together with their answers, to
what was already a uniquely masterful summary of a vast and beautiful theory.
When paired with the second edition of Volume 1, his two classic volumes will
surely be a timeless treasure for generations to come.”
– Donald E. Knuth, Stanford University
“An updated classic with a mesmerizing array of interconnected examples.

Through Stanley’s masterful exposition, the current and future generations of
mathematicians will learn the inherent beauty and pleasures of enumeration.”
– June Huh, Princeton University
“I have used Richard Stanley’s books on Enumerative Combinatorics numerous

times for the combinatorics classes I have taught. This new edition contains
many new exercises, which will no doubt be extremely useful for the next
generation of combinatorialists.”
– Anne Schilling, University of California, Davis
“Richard Stanley’s Enumerative Combinatorics, in two volumes, is an essen-

tial reference for researchers and graduate students in the field of enumeration.
Volume 2, newly revised, includes comprehensive coverage of composition
and inversion of generating functions, exponential and algebraic generating
functions, and symmetric functions. The treatment of symmetric functions is
especially noteworthy for its thoroughness and accessibility. Engaging prob-
lems and solutions, and detailed historical notes, add to the value of this book.
It provides an excellent introduction to the subject for beginners while also
offering advanced researchers new insights and perspectives.”
– Ira Gessel, Brandeis University

Enumerative Combinatorics
Volume 2
Second Edition
RICHARD P. STANLEY
Massachusetts Institute of Technology
With an Appendix by
SERGEY FOMIN
University of Michigan
Shaftesbury Road, Cambridge CB2 8EA, 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 Cambridge University Press & Assessment,

a department of the University of Cambridge.
We share the University’s mission to contribute to society through the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781009262491
DOI: 10.1017/9781009262538
© Cambridge University Press 1997, 2012
© Richard P. Stanley 2024
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 & Assessment.
First edition published by Wadsworth 1986
First Cambridge edition 1997
Second edition (Volume 1) published 2012
Printed in the United Kingdom by TJ Books Limited, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
A Cataloging-in-Publication data record for this book is available from the Library of
Congress.
ISBN 978-1-009-26249-1 Hardback
ISBN 978-1-009-26248-4 Paperback
Cambridge University Press & Assessment 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.
to Atsuko, Kenneth, Sharon, Beth, Bennett, and Cole
Contents
Preface to Second Edition page xiii

Preface xv
5 Trees and the Composition of Generating Functions 1
5.1 The Exponential Formula 1
5.2 Applications of the Exponential Formula 11
5.3 Enumeration of trees 23
5.4 The Lagrange inversion formula 37
5.5 Exponential structures 45
5.6 Oriented trees and the Matrix-Tree Theorem 55
Notes 66
Exercises 72
Solutions 109
6 Algebraic, D-finite, and Noncommutative Generating
Functions 174
6.1 Algebraic generating functions 174
6.2 Examples of algebraic series 184
6.3 Diagonals 194
6.4 D-finite generating functions 202
6.5 Noncommutative generating functions 211
6.6 Algebraic formal series 218
6.7 Noncommutative diagonals 225
Notes 228
Exercises 233
Solutions 271
ix
x Contents
7 Symmetric Functions 318

7.1 Symmetric functions in general 318
7.2 Partitions and their orderings 319
7.3 Monomial symmetric functions 321
7.4 Elementary symmetric functions 322
7.5 Complete homogeneous symmetric functions 326
7.6 An involution 328
7.7 Power sum symmetric functions 330
7.8 Specializations 334
7.9 A scalar product 339
7.10 The combinatorial definition of Schur functions 341
7.11 The RSK algorithm 349
7.12 Some consequences of the RSK algorithm 355
7.13 Symmetry of the RSK algorithm 358
7.14 The dual RSK algorithm 366
7.15 The classical definition of Schur functions 369
7.16 The Jacobi–Trudi identity 377
7.17 The Murnaghan–Nakayama rule 380
7.18 The characters of the symmetric group 385
7.19 Quasisymmetric functions 392
7.20 Plane partitions and the RSK algorithm 401
7.21 Plane partitions with bounded part size 407
7.22 Reverse plane partitions and the Hillman–Grassl
correspondence 415
7.23 Applications to permutation enumeration 418
7.24 Enumeration under group action 427
Notes 434
Exercises 444
Solutions 491
Supplementary Problems 574
Supplementary Solutions 626
Appendix 1 Knuth Equivalence, Jeu de Taquin, and the
Littlewood–Richardson Rule 687
A1.1 Knuth equivalence and Greene’s theorem 687
A1.2 Proofs of Theorems A1.1.1, A1.1.4 and A1.1.6 690
A1.3 Jeu de taquin 693
A1.4 The Schützenberger involution 700
A1.5 The Littlewood–Richardson Rule 703
A1.6 Variations of the Littlewood–Richardson rule 709
A1.7 Notes 712
Contents xi
Appendix 2 The Characters of GL(n, C) 714

A2.1 Basic definitions and results 714
A2.2 Two operations on symmetric functions 720
Bibliography 725
Index 743
Preface to Second Edition
The primary difference between this second edition and the first is the addi-
tion of 159 new problems for Chapter 7 (symmetric functions), almost all
with solutions. They appear in the Chapter 7 sections entitled “Supplementary
Exercises” and “Supplementary Solutions.” These new problems are further
evidence of the amazing fecundity of the theory of symmetric functions. Philip
Hall was well aware of the future potential of symmetric functions when he
wrote in a 1955 letter:∗
. . . whenever in mathematics you meet with partitions, you have only to turn over the
stone or lift up the bark, and you will, almost infallibly, find symmetric functions under-
neath. More precisely, if we have a class of mathematical objects which in a natural and
significant way can be placed in one-to-one correspondence with the partitions, we must
expect the internal structure of these objects and their relations to one another to involve
sooner or later . . . the algebra of symmetric functions.
With a handful of exceptions, all the new problems on symmetric functions

can be solved (in principle, at any rate) directly from the material of Chapter 7.
We do not venture into important extensions of the Chapter 7 material such as
Macdonald polynomials, k-Schur functions, LLT polynomials, cylindric Schur
functions, etc.
We have also corrected all known errors in the first edition and changed all
references to material in Volume 1 so that they conform to the second edition
of Volume 1. A few new problems have been added outside Chapter 7, as parts
of other problems (so as not to upset the problem numbering), and references
have been updated and slightly expanded. The separate list of references in
each chapter have been merged into a single list appearing just before the index.
The numbering of theorems, definitions, exercises, etc., has remained the same
∗ See mathshistory.st-andrews.ac.uk/Biographies/Hall.
xiii
xiv Preface to Second Edition
as in the first edition. However, some equation numbers and figure numbers, as
well as almost all the citation numbers, have been changed. The index has been
upgraded; in particular, every reference to a person is a separate subentry.
Innumerable persons have contributed to the new Chapter 7 problems. They
are acknowledged in the problem solutions. I would also like to thank Kaitlin
Leach at Cambridge University Press for her excellent editorial assistance and
to Suresh Kumar for his TEX support.
Preface
This is the second (and final) volume of a graduate-level introduction to enu-

merative combinatorics. To those who have been patiently waiting twelve years
since the publication of Volume 1, I can only say that no one is more pleased
to see Volume 2 finally completed than myself. I have tried to cover what I feel
are the fundamental topics in enumerative combinatotics, and the ones that
are the most useful in applications outside of combinatorics. Though the book
is primarily intended to be a textbook for graduate students and a resource
for professional mathematicians, I hope that undergraduates and even bright
high school students will find something of interest. For instance, many of
the 66 combinatorial interpretations of Catalan numbers provided by Exer-
cise 6.19 should be accessible to undergraduates with a little knowledge of
combinatorics.
Much of the material in this book has never appeared before in textbook
form. This is especially true of the treatment of symmetric functions in Chap-
ter 7. Although the theory of symmetric functions and its connections with
combinatorics is in my opinion one of the most beautiful topics in all of math-
ematics, it is a difficult subject for beginners to learn. The superb book by
Macdonald on symmetric functions is highly algebraic and eschews the funda-
mental combinatorial tool in this subject, viz., the Robinson-Schensted-Knuth
algorithm. I hope that Chapter 7 adequately fills this gap in the mathematical
literature. Chapter 7 should be regarded as only an introduction to the theory
of symmetric functions, and not as a comprehensive treatment.
As in Volume 1 the exercises play a vital role in developing the subject. If in
reading the text the reader is left with the feeling of “what’s it good for?” and is
not satisfied with the applications presented there, then (s)he should turn to the
exercises hopefully to dispel such doubts. Thanks to the wonders of electronic
word-processing I found it much easier than for Volume 1 to assemble a wide
variety of exercises and solutions.
xv
xvi Preface
I am grateful to the many persons who have contributed in a number of

ways to the improvement of this book. Special thanks goes to Sergey Fomin
for his many suggestions related to Chapter 7, and especially for his master-
ful exposition of the difficult material of Appendix 1. Other persons who have
carefully read portions of earlier versions of the book and who have offered
valuable suggestions are Christine Bessenrodt, Francesco Brenti, Persi Diaco-
nis, Wungkum Fong, Phil Hanlon, and Michelle Wachs. Robert Becker typed
most of Chapter 5, and Tom Roby and Bonnie Friedman provided invaluable
TEX assistance. The following additional persons have made at least one sig-
nificant contribution that is not explicitly mentioned in the text, and I regret
if I have inadvertently omitted anyone else who belongs on this list: Chris-
tos Athanasiadis, Anders Björner, Mireille Bousquet-Mélou, Bradley Brock,
David Buchsbaum, Emeric Deutsch, Kimmo Eriksson, Dominique Foata, Ira
Gessel, Curtis Greene, Patricia Hersh, Martin Isaacs, Benjamin Joseph, Martin
Klazar, Donald Knuth, Darla Kremer, Peter Littelmann, Valery Liskovets, Ian
Macdonald, Alexander Mednykh, Thomas Müller, Andrew Odlyzko, Alexan-
der Postnikov, Robert Proctor, Douglas Rogers, Lou Shapiro, Rodica Simion,
Mark Skandera, Louis Solomon, Dennis Stanton, Robert Sulanke, Sheila
Sundaram, Jean-Yves Thibon, and Andrei Zelevinsky.
5
Trees and the Composition of Generating

Functions
5.1 The Exponential Formula

If F(x) and G(x) are formal power series with G(0) = 0, then we have seen
(after Proposition 1.1.9) that the composition F(G(x)) is a well-defined formal
power series. In this chapter we will investigate the combinatorial ramifications
of power series composition. In this section we will be concerned with the case
where F(x) and G(x) are exponential generating functions, and especially the
case F(x) = ex .
Let us first consider the combinatorial significance of the product F(x)G(x)
of two exponential generating functions
X xn
F(x) = f (n) ,
n!
n≥0
X xn
G(x) = g(n) .
n!
n≥0
Throughout this chapter K denotes a field of characteristic 0 (such as C

with some indeterminates adjoined). We also denote by Ef (x) the exponential
generating function of the function f : N → K, that is,
X xn
Ef (x) = f (n) .
n!
n≥0
5.1.1 Proposition. Given functions f , g : N → K, define a new function h :

N → K by the rule
X
h(#X ) = f (#S)g(#T), (5.1)
(S,T)
1
2 Trees and the Composition of Generating Functions
where X is a finite set, and where (S,T) ranges over all weak ordered partitions
of X into two blocks, that is, S ∩ T = ∅ and S ∪ T = X . Then
Eh (x) = Ef (x)Eg (x). (5.2)
Proof. Let #X = n. There are (nk ) pairs (S, T) with #S = k and #T = n − k, so

n
X n
h(n) = f (k)g(n − k).
k
k=0
From this equation (5.2) follows.
One could also prove Proposition 5.1.1 by using Theorem 3.18.41 applied to
the binomial poset B of Example 3.18.3.
We have stated Proposition 5.1.1 in terms of a certain relationship (5.1)
among functions f , g and h, but it is important to understand its combinato-
rial significance. Suppose we have two types of structures, say α and β, which
can be put on a finite set X . We assume that the allowed structures depend only
on the cardinality of X . A new “combined” type of structure, denoted α ∪ β,
can be put on X by placing structures of type α and β on subsets S and T,
respectively, of X such that S ∪ T = X , S ∩ T = ∅. If f (k) (respectively g(k))
are the number of possible structures on a k-set of type α (respectively, β),
then the right-hand side of (5.1) counts the number of structures of type α ∪ β
on X . More generally, we can assign a weight w(0) to any structure 0 of type
α or β. A combined structure of type α ∪ β is defined to have weight equal
to the product of the weights of each part. If f (k) and g(k) denote the sum of
the weights of all structures on a k-set of types α and β, respectively, then the
right-hand side of (5.1) counts the sum of the weights of all structures of type
α ∪ β on X .
5.1.2 Example. Given an n-element set X , let h(n) be the number of ways to
split X into two subsets S and T with S ∪T = X , S ∩T = ∅; and then to linearly
order the elements of S and to choose a subset of T. There are f (k) = k! ways
to linearly order a k-element set, and g(k) = 2k ways to choose a subset of a
k-element set. Hence
  
X xn X xn X xn
n
h(n) =  n!   2 
n! n! n!
n≥0 n≥0 n≥0
e2x
= .
1−x
1 All references to Volume 1 are to the second edition.
Proposition 5.1.1 can be iterated to yield the following result.
5.1.3 Proposition. Fix k ∈ P and functions f1 , f2 , . . . , fk : N → K. Define a

new function h : N → K by
X
h(#S) = f1 (#T1 )f2 (#T2 ) · · · fk (#Tk ),
where (T1 , . . . , Tk ) ranges over all weak ordered partitions of S into k blocks,
that is, T1 , . . . , Tk are subsets of S satisfying: (i) Ti ∩ Tj = ∅ if i 6= j, and (ii)
T1 ∪ · · · ∪ Tk = S. Then
k
Y
Eh (x) = Efi (x).
i=1
We are now able to give the main result of this section, which explains
the combinatorial significance of the composition of exponential generating
functions.
5.1.4 Theorem (the Compositional Formula). Given functions f : P → K

and g : N → K with g(0) = 1, define a new function h : N → K by
X
h(#S) = f (#B1 )f (#B2 ) · · · f (#Bk )g(k), #S > 0,
π={B1 ,...,Bk }∈5(S)
h(0) = 1,
where the sum ranges over all partitions (as defined in Section 1.9) π =
{B1 , . . . , Bk } of the finite set S. Then
Eh (x) = Eg (Ef (x)).

P xn
(Here Ef (x) = n≥1 f (n) n! , since f is only defined on positive integers.)
Proof. Suppose #S = n, and let hk (n) denote the right-hand side of (5.3) for
fixed k. Since B1 , . . . , Bk are nonempty they are all distinct, so there are k! ways
of linearly ordering them. Thus by Proposition 5.1.3,
g(k)
Ehk (x) = Ef (x)k . (5.3)
k!
Summing (5.3) over all k ≥ 1 yields the desired result.
Theorem 5.1.4 has the following combinatorial significance. Many struc-

tures on a set, such as graphs or posets, may be regarded as disjoint unions
of their connected components. In addition, some additional structure may be
placed on the components themselves, for example, the components could be
linearly ordered. If there are f ( j) connected structures on a j-set and g(k) ways
6 3 4 2 8
Figure 5.1 A circular arrangement of lines
to place an additional structure on k components, then h(n) is the total num-

ber of structures on an n-set. There is an obvious generalization to weighted
structures, such as was discussed after Proposition 5.1.1.
The following example should help to elucidate the combinatorial meaning
of Theorem 5.1.4; more substantial applications are given in Section 5.2.
5.1.5 Example. Let h(n) be the number of ways for n persons to form into
nonempty lines, and then to arrange these lines in a circular order. Figure 5.1
shows one such arrangement of nine persons. There are f ( j) = j! ways to
linearly order j persons, and g(k) = (k − 1)! ways to circularly order k ≥ 1
lines. Thus
X xn x
Ef (x) = n! = ,
n! 1−x
n≥1
X xn
Eg (x) = 1 + (n − 1)! = 1 + log(1 − x)−1 ,
n!
n≥1
so
Eh (x) = Eg (Ef (x))
−1
x
= 1 + log 1 −
1−x
= 1 + log(1 − 2x)−1 − log(1 − x)−1
X xn
= 1+ (2n − 1)(n − 1)! ,
n!
n≥1
whence h(n) = (2n− 1)(n − 1)!. Naturally such a simple answer demands a
simple combinatorial proof. Namely, arrange the n persons in a circle in (n−1)!
6 4
3 2
1 8
Figure 5.2 An equivalent form of Figure 5.1
ways. In each of the n spaces between two persons, either do or do not draw a
bar, except that at least one bar must be drawn. There are thus 2n − 1 choices
for the bars. Between two consecutive bars (or a bar and itself if there is only
one bar) read the persons in clockwise order to obtain their order in line. See
Figure 5.2 for this method of representing Figure 5.1.
The most common use of Theorem 5.1.4 is the case where g(k) = 1
for all k. In combinatorial terms, a structure is put together from “con-
nected” components, but no additional structure is placed on the components
themselves.
5.1.6 Corollary (the Exponential Formula). Given a function f : P → K,

define a new function h : N → K by
X
h(#S) = f (#B1 )f (#B2 ) · · · f (#Bk ), #S > 0, (5.4)
π={B1 ,...,Bk }∈5(S)
h(0) = 1.
Then
Eh (x) = exp Ef (x). (5.5)
Let us say a brief word about the computational aspects of equation (5.5).
If the function f (n) is given, then one can use (5.4) to compute h(n). However,
there is a much more efficient way to compute h(n) from f (n) (and conversely).
5.1.7 Proposition. Let f : P → K and h : N → K be related by Eh (x) =

exp Ef (x) (so in particular h(0) = 1). Then we have for n ≥ 0 the recurrences
n
X n
h(n + 1) = h(k)f (n + 1 − k), (5.6)
k
k=0
n
X n
f (n + 1) = h(n + 1) − h(k)f (n + 1 − k). (5.7)
k
k=1
Proof. Differentiate Eh (x) = exp Ef (x) to obtain
Eh0 (x) = Ef0 (x)Eh (x). (5.8)
Now equate coefficients of xn /n! on both sides of (5.8) to obtain (5.6). (It
is also easy to give a combinatorial proof of (5.6).) Equation (5.7) is just a
rearrangement of (5.6).
The compositional and exponential formulas are concerned with structures

on a set S obtained by choosing a partition of S and then imposing some “con-
nected” structure on each block. In some situations it is more natural to choose
a permutation of S and then impose a “connected” structure on each cycle.
These two situations are clearly equivalent, since a permutation is nothing more
than a partition with a cyclic ordering of each block. However, permutations
arise often enough to warrant a separate statement. Recall that S(S) denotes
the set (or group) of all permutations of the set S.
5.1.8 Corollary (the Compositional Formula, permutation version). Given

functions f : P → K and g : N → K with g(0) = 1, define a new function
h : P → K by
X
h(#S) = f (#C1 )f (#C2 ) · · · f (#Ck )g(k), #S > 0, (5.9)
π∈S(S)
h(0) = 1,
where C1 , C2 , . . . , Ck are the cycles in the disjoint cycle decomposition of π.

Then
 
X x n
Eh (x) = Eg  f (n)  .
n
n≥1
Proof. Since there are (j − 1)! ways to cyclically order a j-set, equation (5.9)
may be written
X
h(#S) = [(#B1 − 1)!f (#B1 )] · · · [(#Bk − 1)!f (#Bk )]g(k),
π={B1 ,...,Bk }∈5(S)
so by Theorem 5.1.4,
 
X xn
Eh (x) = Eg  (n − 1)!f (n) 
n!
n≥1
 
. X xn
= Eg  f (n) 
n
n≥1
5.1.9 Corollary (the Exponential Formula, permutation version). Given a

function f : P → K, define a new function h : N → K by
X
h(#S) = f (#C1 )f (#C2 ) · · · f (#Ck ), #S > 0,
π∈S(S)
h(0) = 1,
where the notation is the same as in Corollary 5.1.8. Then

X xn
Eh (x) = exp f (n) .
n
n≥1
In Chapter 3.18 (see Example 3.18.3(b)) we related addition and multiplica-

tion of exponential generating functions to the incidence algebra of the lattice
of finite subsets of N. There is a similar relation between composition of expo-
nential generating functions and the incidence algebra of the lattice 5n of
partitions of [n] (or any n-set). More precisely, we need to consider simulta-
neously all 5n for n ∈ P. Recall from Section 3.10 that if σ ≤ π in 5n , then
we have a natural decomposition
a a
[σ , π] ∼
= 511 × 522 × · · · × 5ann , (5.10)
ai . Let 5 = (51 , 52 , . . . ). For each n ∈ P,

P P
where |σ | = iai and |π| =
let fn ∈ I(5n , K), the incidence algebra of 5n . Suppose that the sequence f =
( f1 , f2 , . . . ) satisfies the following property: There is a function (also denoted
f ) f : P → K such that if σ ≤ π in 5n and [σ , π] satisfies (5.10), then
fn (σ , π) = f (1)a1 f (2)a2 · · · f (n)an . (5.11)
We then call f a multiplicative function on 5.

For instance, if ζn is the zeta function of 5n , then ζ = (ζ1 , ζ2 , . . . ) is mul-
tiplicative with ζ (n) = 1 for all n ∈ P. If µn is the Möbius function of 5n ,
then by Proposition 3.8.2 and equation (3.37) we see that µ = (µ1 , µ2 , . . . ) is
multiplicative with µ(n) = (−1)n−1 (n − 1)!.
Let f = ( f1 , f2 , . . . ) and g = (g1 , g2 , . . . ), where fn , gn ∈ I(5n , K). We can

define the convolution fg = (( fg)1 , ( fg)2 , . . . ) by
( fg)n = fn gn (convolution in I(5n , K)). (5.12)
5.1.10 Lemma. If f and g are multiplicative on 5, then so is fg.
Proof. Let P and Q be locally finite posets, and let u ∈ I(P, K), v ∈ I(Q, K).
Define u × v ∈ I(P × Q, K) by
u × v((x, x0 ), (y, y0 )) = u(x, y)v(x0 , y0 ).
Then a straightforward argument as in the proof of Proposition 3.8.2 shows

that (u × v)(u0 × v0 ) = uu0 × vv0 . Thus from (5.10) we have
( fg)n (σ , π) = f1 g1 (0̂, 1̂)a1 · · · fn gn (0̂, 1̂)an

= fg(1)a1 · · · fg(n)an .
It follows from Lemma 5.1.10 that the set M(5) = M(5, K) of multipli-
cative functions on 5 forms a semigroup under convolution. In fact, M(5)
is even a monoid (= semigroup with identity), since the identity function
δ = (δ1 , δ2 , . . . ) is multiplicative with δ(n) = δ1n . (CAVEAT: M(5) is not
closed under addition!)
5.1.11 Theorem. Define a map φ : M(5) → xK[[x]] (the monoid of power

series with zero constant term under composition) by
X xn
φ( f ) = Ef (x) = f (n) .
n!
n≥1
Then φ is an anti-isomorphism of monoids, that is, φ is a bijection and
φ( fg) = Eg (Ef (x)).
Proof. Clearly φ is a bijection. Since fg is multiplicative by Lemma 5.1.10, it

suffices to show that
X xn
fg(n) = Eg (Ef (x)).
n!
n≥1
By definition of fg(n), we have in I(5n , K)

X
fg(n) = fn (0̂, π)gn (π, 1̂)
π={B1 ,...,Bk }∈5n
X
= f (#B1 ) · · · f (#Bk )g(k). (5.13)
π
Since (5.13) agrees with (5.3), the proof follows from Theorem 5.1.4.
The next result follows from Theorem 5.1.11 in the same way that Prop-
osition 3.18.5 follows from Theorem 3.18.4 (using Proposition 5.4.1), so the
proof is omitted. (A direct proof avoiding Theorem 5.1.11 can also be given.)
If f = ( f1 , f2 , . . . ) where fn ∈ I(5n , K) and each fn−1 exists in I(5n , K), then
we write f −1 = ( f1−1 , f2−1 , . . . ).
5.1.12 Proposition. Suppose f is multiplicative and f −1 exists. Then f −1 is

multiplicative.
5.1.13 Example. Let ζ , δ, µ ∈ M(5) have the same meanings as above, so

ζ µ = µζ = δ. Now
X xn
Eζ (x) = = ex − 1
n!
n≥1
Eδ (x) = x,
so by Theorem 5.1.11
[exp Eµ (x)] − 1 = x
⇒ Eµ (x) = log(1 + x)
X xn
= (−1)n−1 (n − 1)!
n!
n≥1
⇒ µ(n) = (−1)n−1 (n − 1)!.
Thus we have another derivation of the Möbius function of 5n (equa-
tion (3.37)).
5.1.14 Example. Let h(n) be the number of ways to partition the set [n], and
then partition each block into blocks of odd cardinality. We are asking for the
number of chains 0̂ ≤ π ≤ σ ≤ 1̂ in 5n such that all block sizes of π are odd.
Define f ∈ M(5) by

1, n odd
f (n) =
0, n even.
Then clearly h = f ζ 2 , so by Theorem 5.1.11,
Eh (x) = Eζ (Eζ (Ef (x)))

  
X x2n+1 
= exp exp − 1 − 1
(2n + 1)!
n≥0
sinh x
= exp(e − 1) − 1.
We have discussed in this section the combinatorial significance of multiply-

ing and composing exponential generating functions. Three further operations
are important to understand combinatorially: addition, multiplication by x
(really a special case of arbitrary multiplication, but of special significance),
and differentiation.
5.1.15 Proposition. Let S be a finite set. Given functions f , g : N → K, define

new functions h1 , h2 , h3 , and h4 as follows:
h1 (#S) = f (#S) + g(#S) (5.14)

h2 (#S) = (#S)f (#T), where #T = #S − 1 (5.15)
h3 (#S) = f (#T), where #T = #S + 1 (5.16)
h4 (#S) = (#S)f (#S). (5.17)
Then
Eh1 (x) = Ef (x) + Eg (x) (5.18)

Eh2 (x) = xEf (x) (5.19)
Eh3 (x) = Ef0 (x) (5.20)
Eh4 (x) = xEf0 (x). (5.21)
Proof. Easy.
Equation (5.14) corresponds to a choice of two structures to place on S, one

enumerated by f and one by g. In equation (5.15), we “root” a vertex v of S
(i.e., we choose a distinguished vertex v, often called the root) and then place a
structure on the remaining vertices T = S − {x}. Equation (5.16) corresponds
to adjoining an extra element to S and then placing a structure enumerated by
f . Finally in equation (5.17) we are simply placing a structure on S and rooting
a vertex.
As we will see in subsequent sections, many structures have a recursive
nature by which we can obtain from the results of this section functional equa-
tions or differential equations for the corresponding exponential generating
function. Let us illustrate these ideas here by interpreting combinatorially the

formula Eh0 (x) = Ef0 (x)Eh (x) of equation (5.8). The left-hand side corresponds
to the following construction: take a (finite) set S, adjoin a new element t, and
then place on S ∪ {t} a structure enumerated by h (or “h-structure”). The right-
hand side says: choose a subset T of S, adjoin an element t to T, place on T ∪{t}
an f -structure, and place on S − T an h-structure. Clearly if h and f are related
by (5.4) (so that h-structures are unique disjoint unions of f -structures) then
the combinatorial interpretations of Eh0 (x) and Ef0 (x)Eh (x) are equivalent.
5.2 Applications of the Exponential Formula

The most straightforward applications of Corollary 5.1.6 concern structures
which have an obvious decomposition into “connected components.”
5.2.1 Example. The number of graphs (without loops or multiple edges) on
an n-element vertex set S is clearly 2(2) . (Each of the n2 pairs of vertices may
n
or may not be joined by an edge.) Let c(#S) = c(n) be the number of connected
graphs on the vertex set S. Since a graph on S is obtained by choosing a parti-
tion π of S and then placing a connected graph on each block of π, we see that
equation (5.5) holds for h(n) = 2(2) and f (n) = c(n). Hence by Corollary 5.1.6,
n
X n xn
Eh (x) = 2(2)
n!
n≥0
= exp Ec (x)
X xn
= exp c(n) .
n!
n≥1
Equivalently,
xn X n xn
2(2) .
X
c(n) = log (5.22)
n! n!
n≥1 n≥0
Note that the generating functions Eh (x) and Ec (x) have zero radius of
convergence; nevertheless, they still have combinatorial meaning.
Of course there is nothing special about graphs in the above example. If, for
instance, h(n) is the number of posets (or digraphs, topologies, triangle-free
graphs, . . . ) on an n-set and c(n) is the number of connected posets (digraphs,
topologies, triangle-free graphs, . . . ) on an n-set, then the fundamental rela-
tion Eh (x) = exp Ec (x) continues to hold. In some cases (such as graphs
and digraphs) we have an explicit formula for h(n), but this is an incidental
“bonus.”
5.2.2 Example. Suppose we are interested in not just the number of connected
graphs on an n-element vertex set, but rather the number of such graphs with
exactly k components. Let ck (n) denote this number, and define
XX xn
F(x, t) = ck (n)tk . (5.23)
n!
n≥0 k≥0
There are two ways to obtain this generating function from Theorem 5.1.4 and
Corollary 5.1.6. We can either set f (n) = c(n) and g(k) = tk in (3), or set
f (n) = c(n)t in (5.5). In either case we have
X
h(n) = ck (n)tk .
k≥0
Thus
X xn
F(x, t) = exp t c(n)
n!
 n≥1 t
X n xn
=  2(2)  .
n!
n≥0
Again the same reasoning works equally as well for posets, digraphs, topolo-
gies, . . . In general, if Eh (x) is the exponential generating function for the total
number of structures on an n-set (where of course each structure is a unique
disjoint union of connected components), then Eh (x)t also keeps track of the
number of components, as in (5.23). Equivalently, if h(n) is the number of
structures on an n-set and ck (n) the number with k components, then
X
tk Eck (x) = Eh (x)t
k≥0
= exp tEc1 (x)
X Ec (x)k
= tk 1 , (5.24)
k!
k≥0
so
1 1
Eck (x) = Ec1 (x)k = (log Eh (x))k ,
k! k!
where we set ck (0) = δ0k and h(0) = 1. In particular, if h(n) = n! (the number
of permutations of an n-set) then Eh (x) = (1 − x)−1 while ck (n) = c(n, k), the
number of permutations of an n-set with k cycles. In other words, c(n, k) is a
signless Stirling number of the first kind (see Chapter 1.3); and we get
X xn 1 h ik
c(n, k) = log(1 − x)−1 . (5.25)
n! k!
n≥0
Let us give one further “direct” application, concocted for the elegance of the
final answer.
5.2.3 Example. Suppose we have a room of n children. The children gather

into circles by holding hands, and one child stands in the center of each circle.
A circle may consist of as few as one child (clasping his or her hands), but
each circle must contain a child inside it. In how many ways can this be done?
Let this number be h(n). An allowed arrangement of children is obtained by
choosing a partition of the children, choosing a child c from each block B to
be in the center of the circle, and arranging the other children in the block B in
a circle about c. If #B = i ≥ 2, then there are i · (i − 2)! ways to do this, and no
ways if i = 1. Hence (setting h(0) = 1),
X xi
Eh (x) = exp i · (i − 2)!
i!
i≥2
X xi
= exp x
i
i≥1
= exp x log(1 − x)−1
= (1 − x)−x .
Similarly, if ck (n) denotes the number of arrangements of n children with

exactly k circles, then
XX xn
ck (n)tk = (1 − x)−xt .
n!
n≥0 k≥0
We next consider some problems concerned with successively partitioning

the blocks of a partition.
5.2.4 Example. Let B(n) = B1 (n) denote the nth Bell number, that is, B(n) =
#5n (Chapter 1.9). Setting each f (i) = 1 in (5.4), we obtain
X xn
EB (x) = B(n) = exp(ex − 1).
n!
n≥0
(See equation (1.94f).) Now let B2 (n) be the number of ways to partition an n-
set S, and then partition each block. Equivalently, B2 (n) is the number of chains
0̂ ≤ π1 ≤ π2 ≤ 1̂ in 5n . Putting each f (i) = B(i) in (5.4), or equivalently using
Theorem 5.1.11 to compute φ(ζ 3 ), we obtain
X xn
B2 (n) = exp[exp(ex − 1) − 1].
n!
n≥0
123456789
14689 257
3
1
469 8 2 5 7
49
4 9
Figure 5.3 A total partition of [9] represented as a tree
Continuing in this manner, if Bk (n) denotes the number of chains 0̂ ≤ π1 ≤

π2 ≤ · · · ≤ πk ≤ 1̂ in 5n , then
X xn
Bk (n) = 1 + ehk+1i (x),
n!
n≥0
where e(x) = eh1i (x) = ex − 1 and ehk+1i (x) = e(ehki (x)).
5.2.5 Example. The preceding example was quite straightforward. Consider

now the following variation. Begin with an n-set S, and for n ≥ 2 partition S
into at least two blocks. Then partition each non-singleton block into at least
two blocks. Continue partitioning each non-singleton block into at least two
blocks, until only singletons remain. Call such a procedure a total partition
of S. A total partition can be represented in a natural way by an (unordered)
tree, as illustrated in Figure 5.3 for S = [9]. Notice that only the endpoints
(leaves) need to be labelled; the other labels are superfluous. Let t(n) denote
the number of total partitions of S (with t(0) = 0). Thus t(1) = 1, t(2) =
1, t(3) = 4, t(4) = 26.
Consider what happens when we choose a partition π of S and then a total
partition of each block of π. If |π| = 1, then we have done the equivalent
of choosing a total partition of S. On the other hand, partitioning S into at
least two blocks and then choosing a total partition of each block is equivalent
to choosing a total partition of S itself. Thus altogether we obtain each total
partition of S twice, provided #S ≥ 2. If #S = 1, then we obtain the unique
total partition of S only once. If #S = 0 (i.e., S = ∅) then our procedure can
be done in one way (i.e., do nothing), but by our convention there are no total
partitions of S. Hence from Corollary 5.1.6 we obtain
Figure 5.4 A binary total partition represented as a tree
exp Et (x) = 2Et (x) − x + 1. (5.26)
In other words, writing F h−1i (x) for the compositional inverse of F(x) = ax +
bx2 + · · · where a 6 = 0, that is,
F(F h−1i (x)) = F h−1i (F(x)) = x,
we have
Et (x) = (1 + 2x − ex )h−1i . (5.27)
It does not seem possible to obtain a simpler result. In particular, in Section 5.4
we will discuss methods for computing the coefficients of compositional
inverses, but these methods don’t seem to yield anything interesting when
F(x) = 1 + 2x − ex . For some enumeration problems closely related to total
partitions, see Exercises 26 and 40.
5.2.6 Example. Consider the variation of the preceding example where each
non-singleton block must be partitioned into exactly two blocks. Call such a
procedure a binary total partition of S, and denote the number of them by
b(#S). As with total partitions, set b(0) = 0. The tree representing a binary
total partition is a complete (unordered) binary tree, as illustrated in Figure 5.4.
(Thus b(n) is just the number of (unordered) complete binary trees with n
labelled endpoints.) It now follows from Theorem 5.1.4 (with g(k) = δ2k )
or just by (5.3) (with k = 2 and g(2) = 1), in a similar way to how we obtained
(5.26), that
1
Eb (x)2 = Eb (x) − x. (5.28)
2
Figure 5.5 A decreasing labelled tree corresponding to a binary total partition
Solving this quadratic equation yields

√
Eb (x) = 1 − 1 − 2x
X 1/2
= 1− (−2)n xn
n
n≥0
X xn
= 1 · 3 · 5 · · · (2n − 3) ,
n!
n≥1
whence
b(n) = 1 · 3 · 5 · · · (2n − 3).
As usual, when such a simple answer is obtained a direct combinatorial proof

is desired. Now 1 · 3 · 5 · · · (2n − 3) is easily seen to be the number of partitions
of [2n − 2] of type (2n−1 ), that is, with n − 1 2-element blocks. Given a binary
total partition β of [n], we obtain a partition π of [2n − 2] of type (2n−1 ) as
follows. In the tree representing β (such as Figure 5.4), delete all the labels
except the endpoints (leaves). Now iterate the following procedure until all
vertices are labelled except the root. If labels 1, 2, . . . , m have been used, then
label by m + 1 the vertex v satisfying: (a) v is unlabelled and both successors
of v are labelled, and (b) among all unlabelled vertices with both successors
labelled, the vertex having the successor with the least label is v. Figure 5.5
illustrates this procedure carried out for the tree in Figure 5.4. Finally let the
blocks of π consist of the vertex labels of the two successors of a non-endpoint
vertex. Thus from Figure 5.5 we obtain
π = {{1, 4}, {2, 9}, {3, 10}, {5, 7}, {6, 8}, {11, 12}}.
We leave the reader to check (not entirely trivial) that this procedure yields the
desired bijection.
Certain problems involving symmetric matrices are well-suited for use of the
exponential formula. (Analogous results for arbitrary matrices are discussed in
Section 5.5.) The basic idea is that a symmetric matrix A = (auv ) whose rows
and columns are indexed by a set V may be identified with a graph G = GA
on the vertex set V , with an edge uv connecting u and v labelled by auv . (If
auv = 0, then we simply omit the edge uv, rather than labelling it by 0. More
generally, if auv ∈ P then it is often convenient to draw auv (unlabelled) edges
between u and v.) Sometimes the connected components of GA have a simple
structure, so that the exponential formula can be used to enumerate all the
graphs (or matrices).
5.2.7 Example. As in Proposition 4.6.21, let Sn (2) denote the number of n × n

symmetric N-matrices A with every row (and hence every column) sum equal
to two. The graph GA has every vertex of degree two (counting loops once
only). Hence the connected components of GA must be (a) a single vertex with
two loops, (b) a double edge between two vertices, (c) a cycle of length ≥ 3, or
(d) a path of length ≥ 1 with a loop at each end. There are 12 (n−1)! (undirected)
cycles on n ≥ 3 vertices, and 12 n! (undirected) paths on n ≥ 2 vertices with a
loop at each end. Hence by Corollary 5.1.6,
 
X xn x 2 1 X x n 1 X xn
Sn (2) = exp x + + (n − 1)! + n! 
n! 2! 2 n! 2 n!
n≥0 n≥3 n≥2
 
x2 1 X xn 1 X n
= exp  + + x
4 2 n 2
n≥1 n≥1
2
x 1 x
= exp + log(1 − x)−1 +
4 2 2(1 − x)
2
−1/2 x x
= (1 − x) exp + .
4 2(1 − x)
Using the technique of Exercise 24(c), we obtain the recurrence (writing Sm =

Sm (2))
1
Sn+1 = (2n + 1)Sn − (n)2 Sn−1 − (n)2 Sn−2 + (n)3 Sn−3 , n ≥ 0.
2
5.2.8 Example. Suppose that in the previous example A must be a 0–1 matrix
(i.e., the entry two is not allowed). Now the components of GA of type (a) or
(b) above are not allowed. If we let Sn∗ (2) denote the number of such matrices,
it follows that
X xn x2 X xn
Sn∗ (2) = e−x− 2 Sn (2)
n! n!
n≥0 n≥0
x2

−1/2 x
= (1 − x) exp −x − + .
4 2(1 − x)
As a further variation, suppose we again allow two as an entry, but that tr A = 0
(i.e., all main diagonal entries are zero). Now the components of GA cannot
have loops, so are of type (b) or (c). Hence, letting Tn (2) be the number of such
matrices, we have
 
X xn x2 1 X x n
Tn (2) = exp  + (n − 1)! 
n! 2! 2 n!
n≥0 n≥3
x x2

= (1 − x) −1/2 exp − + .
2 4
Similarly, if Tn∗ (2) denotes the number of traceless symmetric n×n 0-1 matrices
with line sum 2, then
 
X x n 1 X x n
Tn∗ (2) = exp  (n − 1)! 
n! 2 n!
n≥0 n≥3
x x2

−1/2
= (1 − x) exp − − . (5.29)
2 4
The recurrence relations for Sn∗ (2), Tn (2), and Tn∗ (2) turn out to be (using the
technique of Exercise 24(c))
∗ (2) = 2nS ∗ (2) − n(n − 2) S ∗ (2) − 1 (n) S ∗ (2)

Sn+1 n n−1 3 n−3
2
Tn+1 (2) = nTn (2) + nTn−1 (2) − n2 Tn−2 (2)

∗ (2) = nT ∗ (2) + n T ∗ (2),

Tn+1 n 2 n−2
all valid for n ≥ 0.
The next example is an interesting variation of the preceding two, where it

is not a priori evident that the exponential formula is relevant.
5.2.9 Example. Let Xn = (xij ) be an n × n symmetric matrix whose entries

xij are independent indeterminates (over R, say), except that xij = xji . Let L(n)
be the number of terms (i.e., distinct monomials) in the expansion of det Xn ,

where we use only the variables xij for i ≤ j. For instance,
det X3 = x11 x22 x33 + 2x12 x23 x13 − x213 x22 − x11 x223 − x212 x33 .
Hence L(3) = 5, since the above sum has five terms. One might ask whether
we should count a monomial which does arise in the expansion of det Xn but
whose coefficient because of cancellation turns out to be zero. But we will soon
see that no cancellation is possible; all occurrences of a given monomial have
the same sign. Suppose now that π ∈ Sn . Define
Mπ = x1,π(1) x2,π(2) · · · xn,π(n) ,
where we set xji = xij if j > i. Thus Mπ is the monomial corresponding

to π in the expansion of det Xn . Define a graph Gπ whose vertex set is [n],
and with an (undirected) edge between i and π(i) for 1 ≤ i ≤ n. Thus the
components of Gπ are cycles of length ≥ 1. (A cycle of length 1 is a loop,
and of length 2 is a double edge.) The crucial observation, whose easy proof
we omit, is that Mπ = Mσ if and only if Gπ = Gσ . Since a permutation π
is even (respectively, odd) if and only if Gπ has an even number (respectively,
odd number) of cycles of even length, it follows that if Mπ = Mσ then Mπ
and Mσ occur in the expansion of det Xn with the same sign. In other words,
there is no cancellation in the expansion of det Xn . Also note that a graph G on
[n], every component of which is a cycle, is equal to Gπ for some π ∈ Sn . (In
fact, the number of such π is exactly 2d(π ) , where π has d(π ) cycles of length
≥ 3.) It follows that L(n) is simply the number of graphs on [n] for which every
component is a cycle (including loops and double edges). Hence
 
X xn x 2 1 X xn
L(n) = exp x + + (n − 1)! 
n! 2! 2 n!
n≥0 n≥3
x x2

= (1 − x)−1/2 exp + .
2 4
Note also that if Pπ is the permutation matrix corresponding to π ∈ Sn , then

Gπ = Gσ if and only if Pπ + Pπ−1 = Pσ + Pσ−1 . Hence L(n) is the number
of distinct matrices of the form Pπ + Pπ−1 , where π ∈ Sn . Equivalently, if we
define π, σ ∈ Sn to be equivalent if every cycle of π is a cycle or inverse of a
cycle of σ , then L(n) is the number of equivalence classes.
We conclude this section with some examples where it is more natural to use
the permutation version of the exponential formula (Corollary 5.1.9).
5.2.10 Example. Let π ∈ Sn be a permutation. Suppose that π has ci = ci (π )

P
cycles of length i, so ici = n. Form a monomial
Z(π) = Z(π , t) = t1c1 t2c2 · · · tncn
in the variables t = (t1 , t2 , . . . ). We call Z(π) the cycle index (or cycle indicator
or cycle monomial) of π. Define the cycle index or cycle index polynomial (or
cycle indicator, etc.) Z(Sn ) (also denoted ZSn (t), PSn (t), Cyc(Sn , t), etc.) of
Sn by
1 X
Z(Sn ) = Z(Sn , t) = Z(π ).
n!
π∈Sn
(In Chapter 7.7.24 we will consider the cycle index of other permutation
groups.) It is sometimes more convenient to work with the augmented cycle
index
X
Z̃(Sn ) = n! Z(Sn ) = Z(π).
π ∈Sn
For instance
Z̃(S1 ) = t1
Z̃(S2 ) = t12 + t2
Z̃(S3 ) = t13 + 3t1 t2 + 2t3
Z̃(S4 ) = t14 + 6t12 t2 + 8t1 t3 + 3t22 + 6t4 .
Clearly if we define f : P → K by f (n) = tn , then

X
Z̃(Sn ) = f (#C1 )f (#C2 ) · · · f (#Ck ),
π∈Sn
where C1 , C2 , · · · , Ck are the cycles of π. Hence by Corollary 5.1.9,

X xn X xi
Z̃(Sn ) = exp ti . (5.30)
n! i
n≥0 i≥1
There are many interesting specializations of (5.30) related to enumerative

properties of Sn . For instance, fix r ∈ P and let er (n) be the number of π ∈ Sn
satisfying π r = id. A permutation π satisfies π r = id if and only if cd (π ) = 0
whenever d6 |r. Hence
er (n) = Z̃(Sn )
1, d|r
td =
0, d6 |r.
It follows that
 
X xn X xd
er (n) = exp  td  
 1, d|r
n! d t =
n≥0 d≥1 d 
0, d6 |r
 
X xd
= exp  . (5.31)
d
d|r
In particular, the number e2 (n) of involutions in Sn satisfies

xn x2
X
e2 (n) = exp x + . (5.32)
n! 2
n≥0
This is the same generating function encountered way back in equation (1.11).
Now we are able to understand its combinatorial significance more clearly.
There is a surprising connection between (a) Corollary 5.1.9, (b) the rela-
tionship between linear and circular words obtained in Proposition 4.7.13, and
(c) the bijection π 7 → π̂ discussed in Chapter 1.3 between permutations writ-
ten as products of cycles and as words. Basically, such a connection arises from
a formula of the type
X xn X xn
n! f (n) = exp (n − 1)! g(n)
n! n!
n≥0 n≥1
because n! is the number of linear words (permutations) on [n], while (n − 1)!

is the number of circular words (cycles). The next example may be regarded as
the archetype for this line of thought.
5.2.11 Example. Let

Y
F(x) = (1 − tk x)−1 , (5.33)
k
where k ranges over some index set, say k ∈ P. Thus

X
log F(x) = log(1 − tk x)−1
k
XX xn
= tkn
n
k n≥1
X xn
= pn (t) , (5.34)
n
n≥1
X
where pn (t) = tkn . On the other hand, it is clear from (5.33) that
k
X
F(x) = hn (t)xn , (5.35)
n≥0
where hn (t) is the sum of all monomials of degree n in t = (t1 , t2 , · · · ), that is,
t1a1 t2a2 · · · .
X
hn (t) =
a1 +a2 +···=n
ai ∈N
(In Chapter 7 we will analyze the symmetric functions pn (t) and hn (t), as well
as many others, in much greater depth.) From (5.34) and (5.35) we conclude
X xn X xn
n! hn (t) = exp pn (t) . (5.36)
n! n
n≥0 n≥1
We wish to give a direct combinatorial proof. By Corollary 5.1.9, the right-

hand side is the exponential generating function for the following structure:
Choose a permutation π ∈ Sn , and weight each cycle C of π by a monomial
tk#C for some k. Define the total weight of π to be the product of the weights of
each cycle. For instance the list of structures of weight u2 v (where u = t1 and
v = t2 , say) is given by
(1) (2) (3) (12) (3)

u u v uu v
(1) (2) (3) (13) (2)
(5.37)
u v u uu v
(1) (2) (3) (1) (23)

v u u v uu
Moreover, the left-hand side of (5.36) is clearly the exponential generating
function for pairs (π , ta ), where π ∈ Sn and ta is a monomial of degree n in t.
Thus the structures of weight u2 v are given by
123, u2 v 231, u2 v
132, u2 v 312, u2 v (5.38)
2 2
213, u v 321, u v.
In both (5.37) and (5.38) there are six items.
In general, in order to prove (5.36) bijectively, we need to do the following.
Given a monomial ta of degree n, find a bijection φ : C a → Sn , where Ca
is the set of all permutations π in Sn , with each cycle C weighted by tj#C for
some j = j(C), such that the total weight 5C tj(C)
#C is equal to ta . To describe φ,
first impose some linear ordering on the tj ’s, say t1 < t2 < · · · . For fixed j, take
all the cycles C of π with weight tj#C and write their standard representation (in
the sense of Proposition 1.3.1), that is, the largest element of each cycle is writ-
ten first in the cycle, and the cycles are written left-to-right in increasing order
of their largest elements. Remove the parentheses from this standard represen-
tation, obtaining a word wj . Finally set φ(π ) = (w, ta ), where w = w1 w2 · · ·
(juxtaposition of words). For instance, suppose π is the weighted permutation
(19) (82) (3) (547) (6)
π=
vv uu v uuu u
where u < v. The cycles weighted by u’s and v’s, respectively, have standard
form
(6)(754)(82) (weight u6 )
(3)(91) (weight v3 ).
Hence
w1 = 675482, w2 = 391,
φ(π) = (675482391, u6 v3 ).
It is easy to check that φ is a bijection. Given (w, ta ), the monomial ta deter-
|w |
mines the words wj with their weights tj j . Each word wj then corresponds to
|C|
a collection of cycles C (with weight tj ) using the inverse of the bijection
π 7 → π̂ of Chapter 1.3.
A similar argument leads to a direct combinatorial proof of equation (4.41);
see Exercise 21.
5.3 Enumeration of trees

Trees have a recursive structure which makes them highly amenable to the
methods of this chapter. We will develop in this section some basic properties
of trees as a prelude to the Lagrange inversion formula of the next section.
Trees are also fascinating objects of study for their own sake, so we will cover
some topics not strictly germane to the composition of generating functions.
For the basic definitions and terminology concerning trees, see Appendix
A of Volume 1. We also define a planted forest (also called a rooted forest or
forest of rooted trees) to be a graph for which every connected component is
a (rooted) tree. We begin with an investigation of the total number pk (n) of
planted forests with k components on the vertex set [n]. Note that p1 (n) is just
the number r(n) of rooted trees on [n]. If S ⊆ [n] and #S = k, then define pS (n)
Figure 5.6 A plane tree built from the subtrees of its root
n forests on [n] with k components, whose set of

to be the number of planted
roots is S. Thus pk (n) = pS (n), since clearly pS (n) = pT (n) if #S = #T.
k
5.3.1 Proposition. Let
X xn
y = R(x) = r(n) ,
n!
n≥1
where r(n) as above is the number of rooted trees on the vertex set [n] (with
r(0) = 0). Then y = xey , or equivalently (since x = ye−y ),
y = (xe−x )h−1i . (5.39)
Moreover, for k ∈ P we have

1 k X xn
y = pk (n) . (5.40)
k! n!
n≥1
Proof. By Corollary 5.1.6, ey is the exponential generating function for

planted forests on the vertex set [n]. By equation (5.19), xey is the expo-
nential generating function for the following structure on [n]. Choose a root
vertex i, and place a planted forest F on the remaining vertices [n] − {i}. But
this structure is equivalent to a tree with root i, whose subtrees of the root
are the components of F. (See Figure 5.6.) Thus xey is just the exponential
generating function for trees, so y = xey . Equation (5.40) then follows from
Proposition 5.1.3.
In the functional equation y = xey of Proposition 5.3.1, substitute xey for the
occurrence of y on the right-hand side to obtain
y
xe
y = xe
Again making the same substitution yields
y
xe
xe
y = xe .
Iterating this procedure yields the “formula”
·
·
x·
xe
xe
y = xe . (5.41)
The precise meaning of (5.41) is the following. Define y0 = x and for k ≥ 0,

yk+1 = xeyk . Then limk→∞ yk = y, where the limit exists in the formal power
series sense of Chapter 1.1. Moreover, by Corollary 5.1.6 and the second case
of Proposition 5.1.15, we see that
X xn
yk = rk (n) ,
n!
n≥1
where rk (n) in the number of rooted trees on [n] of length ≤ k. For instance,
X xn
y1 = xex = n ,
n!
n≥1
so r1 (n) = n (as is obvious from the definition of r1 (n)).

The following quantities are closely related to the number r(n) of rooted
trees on the vertex set [n]:
t(n) = number of free trees on [n]
f (n) = number of free forests (i.e., disjoint unions of free trees) on [n]
p(n) = number of planted forests on [n].
We set t(0) = 0, f (0) = 1, p(0) = 1. Also write T(x) = Et (x), F(x) = Ef (x),
and P(x) = Ep (x). It is easy to verify the following relations:
r(n) = np(n − 1) = nt(n), p(n) = t(n + 1)
F(x) = eT(x) , P(x) = eR(x) (5.42)
P(x) = T 0 (x), R(x) = xP(x).

Figure 5.7 Constructing the Prüfer sequence of a labelled forest
[n]
5.3.2 Proposition. We have pS (n) = knn−k−1 for any S ∈ k . Thus

n n−k−1 n − 1 n−k
pk (n) = k n = n
k k−1
n−1
r(n) = n
t(n) = nn−2
p(n) = (n + 1)n−1 .
First proof. The case n = k is trivial, so assume n ≥ k + 1. The number of

sequences s = (s1 , . . . , sn−k ) with si ∈ [n] for 1 ≤ i ≤ n − k − 1 and sn−k ∈ S
is equal to knn−k−1 . Hence we seek a bijection γ : Tn,S → [n]n−k−1 ×S, where
Tn,S is the set of planted forests on [n] with root set S. Given a forest σ ∈ Tn,S ,
define a sequence σ1 , σ2 , . . . , σn−k+1 of subforests (all with root set S) of σ as
follows: Set σ1 = σ . If i < n − k + 1 and σi has been defined, then define
σi+1 to be the forest obtained from σi by removing its largest nonroot endpoint
vi (and the edge incident to vi ). Then define si to be the unique vertex of σi
adjacent to vi , and let γ (σ ) = (s1 , s2 , . . . , sn−k ). The sequence γ (σ ) is called
the Prüfer sequence or Prüfer code of the planted forest σ . Figure 5.7 illustrates
this construction with a forest σ = σ1 ∈ T11,{2,7} and the subforests σi , with
vertex vi circled. Hence for this example γ (σ ) = (5, 11, 5, 2, 9, 2, 7, 5, 7).
We claim that the map γ : Tn,S → [n]n−k−1 × S is a bijection. The cru-
cial fact is that the largest element of [n] − S missing from the sequence
(s1 , . . . , sn−k ) must be v1 [why?]. Since v1 and s1 are adjacent, we are reduced
to computing σ2 . But γ (σ2 ) = (s2 , s3 , . . . , sn−k ) (keeping in mind that the ver-
tex set of σ2 is [n] − {v1 }, and not [n − 1]). Hence by induction on n (the case
n = k+1 being easy) we can recover σ uniquely from any (s1 , . . . , sn−k ), so the
proof is complete.
5.3.3 Example. Let S = {2, 7} and (s1 , . . . , s9 ) = (5, 11, 5, 2, 9, 2, 7, 5, 7), so

n − k = 9 and n = 11. The largest element of [11] missing from (s1 , . . . , s9 )
is 10. Hence 10 is an endpoint of σ adjacent to s1 = 5. The largest element of

[11] − {10} missing from (s2 , . . . , s9 ) = (11, 5, 2, 9, 2, 7, 5, 7) is 8. Hence 8 is
an endpoint of σ2 adjacent to s2 = 11. The largest element of [11] − {8, 10}
missing from (s3 , . . . , s9 ) = (5, 2, 9, 2, 7, 5, 7) is 11. Hence 11 is an endpoint
of σ3 adjacent to s3 = 5. Continuing in this manner, we obtain the sequence
of endpoints 10, 8, 11, 6, 4, 9, 3, 1, 5. By beginning with the roots 2 and 7, and
successively adding the endpoints in reverse order to the vertices (s9 , . . . , s1 ) =
(7, 5, 7, 2, 9, 2, 5, 11, 5), we obtain the forest σ = σ1 of Figure 5.7.
Second proof of Proposition 5.3.2. We will show by a suitable bijection that

n
npk (n) = k nn−k . (5.43)
k
The underlying idea of the bijection is that a permutation can be represented
both as a word and as a disjoint union of cycles. The bijection can be simplified
somewhat for the case of rooted trees (k = 1), so we will present this special
case first. Given a rooted tree τ on the vertex set [n], circle a vertex s ∈ [n].
Let w = w1 w2 · · · wk be the sequence (or word) of vertices in the unique path
P in τ from the root r to s. Regard w as a permutation of its elements written in
increasing order. For instance, if w = 57283, then w represents the permutation
given by w(2) = 5, w(3) = 7, w(5) = 2, w(7) = 8, w(8) = 3, which in
cycle notation is (2, 5)(3, 7, 8). Let Dw be the directed graph with vertex set
A = {w1 , . . . , wk }, and with an edge from j to w(j) for all j ∈ A. Thus Dw is a
disjoint union of (directed) cycles. When we remove from τ the edges of the
path from r to s, we obtain a disjoint union of trees. Attach these trees to Dw by
identifying vertices with the same label, and direct all the edges of these trees
toward Dw . We obtain a digraph D(τ , s) for which all vertices have outdegree
one. Moreover, the rooted tree τ , together with the distinguished vertex s, can
be uniquely recovered from D(τ , s) by reversing the above steps. Since there
are n choices for the vertex s, it follows that nr(n) is equal to the number of
digraphs on the vertex set [n] for which every vertex has outdegree one. But
such a digraph is just the digraph Df of a function f : [n] → [n] (i.e., for each
j ∈ [n] draw an edge from j to f ( j)). Since there are nn such functions, we get
nr(n) = nn , so r(n) = nn−1 as desired.
If we try the same idea for arbitrary planted forests σ , we end up at the
end needing to count functions f from some subset B of [n] to [n] such that
the digraph Df with vertex set [n] and edges j → f ( j) is nonacyclic (i.e.,
has at least one directed cycle). Since there is no obvious way to count such
functions, some modification of the above bijection is needed. Note that the
nonacyclicity condition is irrelevant when k = 1, since when B = [n] the
digraph Df is always nonacyclic.
Another random document with
no related content on Scribd:
They didn't have to wait long. From the edge of the hole, there
suddenly appeared a moving shape. It was a manlike figure clad in a
brilliant crimson spacesuit. The helmet was a dark purple, and it was
difficult to see the head within.
"Looks like a man," said Blake.

"Not quite," MacIlheny said. "Look at the joints in the arms and legs.
He's got two knees and two elbows."
"What's that he's holding cradled in his arms?" Blake wondered.
The major grunted. "Weapon of some sort. Look how he's pointing it
straight at us."
For a full minute, the figure stood there, for all the world as though he
were on the surface of a planet instead of on the outer hull of a space
station. Then, slowly, it lowered the thing in its hands. When nothing
happened, the figure put the weapon down on the steel hull at its feet
and held its oddly double-jointed arms out from its body.
"Wild Bill Hickok," breathed Blake softly.
"Huh?" said the major.
"Hickok used to say: 'I'm a peaceable man.' I guess that's what this
guy's trying to say."
"Looks like it," agreed MacIlheny. "I wish there were some way of
signaling him."
"We've got the spotlights," suggested the major.
MacIlheny shook his head. "Leave 'em alone. We couldn't make any
sense with them, and our friend out there might think they were
weapons of some kind. I don't know what that thing he laid down will
do, but I don't want to find out just yet."
The alien, his hands still out from his sides, walked slowly toward the
RJ-37, his legs moving with a strange, loose suppleness. He came
right up to the forward window and peered inside—at least, the
attitude of his head suggested peering; within the dark purple helmet,
the features could not be distinguished clearly.
At last, the figure stepped back and started making wigwag signs with
his arms.
"Smart boy," said MacIlheny. "He recognizes that the ship is remote
controlled. Wonder what he's trying to say."
The alien waved his hands and made gestures, but there was no
recognizable pattern. None of the hand-signals meant anything to the
Earthmen.
Blake leaned over and whispered into MacIlheny's ear. "Hadn't we
better call the President, Mac? He'll want to know."
MacIlheny considered for a moment, then nodded. "Give him a direct
beam on what's coming over this screen. Then give me a pair of
earphones connected to his office. I want to be able to hear what he
says, but I don't want him countermanding my orders to Major
Hamacher."
The alien was still making his meaningless signals when Blake
brought in a pair of earphones and clamped them on MacIlheny's
head. A throat mike around his neck completed the communication
circuit. "Can you hear me, Mr. President?" MacIlheny asked.
"Yes. Your man Blake explained everything to me."
"Got any advice?"
"Not yet. Let's see what happens. By the way, I've given the
impression to the rest of the world that it was through your efforts that
Number Four avoided crashing; I don't think we'd better let this leak
out just yet."
"Right. Meantime, I'm going to try to capture that lad."
"How?" asked the President.
"Invite him into the ship and bring him back with it."
"All right," said the President, "but be careful."
"He's given up," said Blake, gesturing toward the screen.
The alien had given up his incomprehensible gesticulating and stood
with his odd arms folded in an uncomfortable-looking knot.
"Major," said MacIlheny, "open the cargo hold."
The officer looked puzzled, but did as he was told. After all, the
President himself had ordered him to obey MacIlheny. He touched a
button on one side of the control panel. After four or five seconds, a
light came on above it, indicating that the cargo hold of the RJ-37 was
open. The alien evidently saw the door swing inward; he hesitated for
a moment, then went around to the side of the ship, out of range of
the TV camera.
But he didn't go inside immediately. MacIlheny hadn't expected him
to; the alien couldn't be that stupid. After perhaps half a minute, the
alien figure reappeared and strode deliberately back to his own ship.
He opened a port in the side and disappeared within.
Then, quite suddenly, the screen went blank.
"What happened?" snapped MacIlheny.
Blake, who had been watching the beam control instruments, said: "I
don't know how he's done it, but he's managed to jam our radio
beam! We're not getting any signal through!"
The President's voice crackled in MacIlheny's ears.
"Fitz! Detonate that bomb! We can't take any chances!"
MacIlheny half grinned. "Major," he said, "set off the H-bomb."
The major pressed a red button on the control panel.
Twenty minutes later, the screen came on again, showing the same
scene as before. No one was surprised. By then, reports had come in
that the satellite was still visible, still in its orbit. The H-bomb had
failed to go off; the signal had never reached the detonation device.
The alien was standing in front of the camera, holding a large piece of
mechanism in his hands. On Earth, the thing would have been almost
too heavy to lift, but the gravitational pull of Satellite Number Four
was almost negligible.
"He's got the H-bomb!"
MacIlheny recognized the President's voice in his ears.
The alien bowed toward the camera, then straightened and went
back to his own ship. He clambered up the side of it with magnetic
soles as easily as he had walked on the hull of the space station.
Near the end of his ship, he opened a small door in the hull. Within
was utter blackness.
Working slowly and deliberately, he pushed the H-bomb into the

blackness. It wasn't just ordinary darkness; it seemed to be an actual,
solid wall, painted deep black. As the bomb went in, it looked as
though it were cut off abruptly at the black wall. Finally, there was
nothing outside except the two detonating wires, which had been
clipped off from inside the Earth ship. The alien took the wires in his
hands.
"My God!" said the major. "He's going to blow up his own ship! Is he
crazy?"
"I don't think so," said MacIlheny slowly. "Let's see what happens."
As the two wires came in contact, the black wall inside the small door
became lighter, a pearly gray in color. There was no other result.
"Well I'll be damned," said Blake in a low, shocked voice.
The alien closed the door in the side of his ship and came back down
to the camera. He bowed again. Then he pointed to the weapon that
he had been carrying and waved his hands. He picked it up and
brought it around the RJ-37. From the microphones inside the ship
came a faint scraping sound. Then the alien reappeared in front of
the ship. His arms were empty; he had put the weapon inside the
open cargo hold.
"A fair trade is no robbery," Blake said softly.
The alien bowed once more, then turned on his heel and walked back
to his ship. This time, he got inside and closed the door. Then the
blue ship moved.
Slowly, like a car backing out of a garage, it pulled out of the hole in
the satellite. Nowhere on its surface was there a mark or a scratch.
When it was finally free of the satellite, it turned a little, its nose
pointing off into space. A pale, rose-colored glow appeared at the tail
of the ship, and the cigar of blue metal leaped forward. To all intents
and purposes, it simply vanished.
"That," said the major in awe, "is what I call acceleration."
"Here's the way I see it, Mr. President," said MacIlheny several hours
later. "When he cracked up by accidentally plowing into Number Four,
something happened to his energy supply. Maybe he was already
low, I don't know. Anyway, he was out of fuel."
"What do you think he used for fuel?"
"The most efficient there is," said MacIlheny. "Pure energy. Imagine
some sort of force field that will let energy in, but won't let it out. It
would be dead black on the outside, just like that whatever-it-was in
the alien's ship. He just set off the H-bomb inside that field; what little
radiation did get out made the field look gray—and that's a damned
small loss in comparison with the total energy of that bomb."
"You know, Fitz, I'm going to have a hell of a job explaining where that
bomb went," said the President.
"Yeah, but we've got his gun or whatever in exchange."
"But how do you know our technicians will be able to figure it out?"
"I think they will," MacIlheny said. "Their technology must be similar to
ours or he wouldn't have been able to figure out how to fire the jets on
the satellite or how to set off that bomb. He wouldn't have even
known what the bomb was unless he was familiar with something
similar. And he wouldn't have been able to blank out our controls
unless he had a good idea of how they operated. They may be a little
ahead of us, but not too much, and I'll bet we have some things they
haven't."
"The trouble is," the President said worriedly, "that we don't know
where he came from. He knows where we are, but we don't have any
idea where his home planet is."
"That's true. On the other hand, we know something about his
physical characteristics, while he doesn't know anything about ours.
For instance, I doubt if he'd be happy here on Earth; judging by the
helmet he wore, he can't stand too much light. He had it polarized
almost black. Probably comes from a planet with a dim, red sun."
"Well, Fitz, when they do come, I hope it's for trade and not for war."
MacIlheny grinned. "It won't be war. Don't you remember? We've
started trading already!"
*** END OF THE PROJECT GUTENBERG EBOOK BEYOND OUR
CONTROL ***
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.

Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx

Uploaded by

Copyright:

Available Formats

You might also like

Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx

Uploaded by

Document Information

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

Enumerative Combinatorics Volume 2 Second Edition Richard P Stanley 3 full chapter pdf docx

Uploaded by

Copyright:

Available Formats

Enumerative Combinatorics: Volume 2:

Second Edition Richard P. Stanley

Enumerative Combinatorics: Volume 2: Second Edition

Enumerative Combinatorics: Volume 2: Second Edition

Fetal and Neonatal Physiology, 2-Volume Set 6th Edition

Principles of Bone Biology (2 Volume Set) 4th Edition

Gray's basic anatomy Second Edition Richard Lee Drake

Handbook of Applied Hydrology, Second Edition Vijay P.

Benzel's Spine Surgery, 2-Volume Set: Techniques,

Big Chance Cowboy Teri Anne Stanley [Stanley

Richard P. Stanley is Emeritus Professor of Mathematics at the Massachusetts Institute

– Persi Diaconis, Stanford University

“It is wonderful to celebrate the completion of the second edition of Richard

– Donald E. Knuth, Stanford University

“An updated classic with a mesmerizing array of interconnected examples.

– June Huh, Princeton University

“I have used Richard Stanley’s books on Enumerative Combinatorics numerous

– Anne Schilling, University of California, Davis

“Richard Stanley’s Enumerative Combinatorics, in two volumes, is an essen-

– Ira Gessel, Brandeis University

Cambridge University Press is part of Cambridge University Press & Assessment,

Preface to Second Edition page xiii

7 Symmetric Functions 318

Appendix 2 The Characters of GL(n, C) 714

With a handful of exceptions, all the new problems on symmetric functions

This is the second (and final) volume of a graduate-level introduction to enu-

I am grateful to the many persons who have contributed in a number of

Trees and the Composition of Generating

5.1 The Exponential Formula

Throughout this chapter K denotes a field of characteristic 0 (such as C

5.1.1 Proposition. Given functions f , g : N → K, define a new function h :

Proof. Let #X = n. There are (nk ) pairs (S, T) with #S = k and #T = n − k, so

Proposition 5.1.1 can be iterated to yield the following result.

5.1.3 Proposition. Fix k ∈ P and functions f1 , f2 , . . . , fk : N → K. Define a

5.1.4 Theorem (the Compositional Formula). Given functions f : P → K

Eh (x) = Eg (Ef (x)).

Theorem 5.1.4 has the following combinatorial significance. Many struc-

Figure 5.1 A circular arrangement of lines

to place an additional structure on k components, then h(n) is the total num-

= 1 + log(1 − 2x)−1 − log(1 − x)−1

Figure 5.2 An equivalent form of Figure 5.1

5.1.6 Corollary (the Exponential Formula). Given a function f : P → K,

5.1.7 Proposition. Let f : P → K and h : N → K be related by Eh (x) =

Proof. Differentiate Eh (x) = exp Ef (x) to obtain

Eh0 (x) = Ef0 (x)Eh (x). (5.8)

The compositional and exponential formulas are concerned with structures

5.1.8 Corollary (the Compositional Formula, permutation version). Given

where C1 , C2 , . . . , Ck are the cycles in the disjoint cycle decomposition of π.

5.1.9 Corollary (the Exponential Formula, permutation version). Given a

where the notation is the same as in Corollary 5.1.8. Then

In Chapter 3.18 (see Example 3.18.3(b)) we related addition and multiplica-

ai . Let 5 = (51 , 52 , . . . ). For each n ∈ P,

fn (σ , π) = f (1)a1 f (2)a2 · · · f (n)an . (5.11)

We then call f a multiplicative function on 5.

Let f = ( f1 , f2 , . . . ) and g = (g1 , g2 , . . . ), where fn , gn ∈ I(5n , K). We can

( fg)n = fn gn (convolution in I(5n , K)). (5.12)

5.1.10 Lemma. If f and g are multiplicative on 5, then so is fg.

u × v((x, x0 ), (y, y0 )) = u(x, y)v(x0 , y0 ).

Then a straightforward argument as in the proof of Proposition 3.8.2 shows

n forests on [n] with k components, whose set of