# THE ENUMERATION OF ROOTED TREES BY TOTAL HEIGHT

Save this PDF as:

Size: px
Start display at page:

## Transcription

1 THE ENUMERATION OF ROOTED TREES BY TOTAL HEIGHT JOHN RIORDAN and N. J. A. SLOANE (Received 22 July 1968) 1. Introduction The height (as in [3] and [4]) of a point in a rooted tree is the length of the path (that is, the number of lines in the path) from it to the root; the total height of a rooted tree is the sum of the heights of its points. The latter arises naturally in studies of random neural networks made by one of us (N.J.A.S.), where the enumeration of greatest interest is that of trees with all points distinctly labeled. Wiite J vh for the number of rooted trees with p labeled points and total height h, and for the enumerator (= enumerating generating function) of such trees by number of points and by total height. It will be shown that (1) J(x, y) =xexpj(xy, y) which is an analogue of George Polya's familiar formula, [7], R(x) = x exp R(x), with R(x) the enumerator of rooted trees by number of labeled points; indeed R(x) = J(x, 1). Also if with the prime denoting a derivative (W^/R^ is the mean total height of all rooted trees with p labeled points and also the mean height) the enumerator W(x) = 2 W v x v lp\ is found to be (2) W{x) = [xr'{x)f with R'{x) the derivative of R(x). From (2), it follows that (3)

2 [2] The enumeration of rooted trees 279 Since R n = w"-\ it follows that W (4) ^ For completeness, it may be noted that the enumerator H (x, y, z) by number of points, by total height, and by number of points labeled, that is, the generating function having, as coefficient of x v y h z k fk\, H vhk the number of rooted trees with p points, k of which are labeled (distinct labels), and with total height h, is given by 1 ' H(x, y,z)=x ex V [H(xy, y, z)+h(x*y*. 2/ 2 )/2+... with H[x,y, 1) = H(x,y). For oriented trees, the results corresponding to (1) and (5) are (6) L(x, y) = x exp 2L(xy, y) 1 } K(x, y,z)=x exp 2[K(xy, y, z)+k(x*yz, 2/ 2 )/ K(x*y\ y*)lk+...]. 2. Labeled rooted trees by total height The simplest argument leading to (1) is but a slight modification of Polya's [7] derivation of the formula for rooted trees (cf [5], p. 127 et seq). Write J n {y,m) for the enumerator by total height with n labeled points and m lines at the root. Then first (8) 7»fo;i) =»y-\7»-i(y); for there are n possible labels for the root, and for each of these the height of any of the n 1 points of the tree connected to the single line at the root is increased by unity. Next (9) Jn{y, 2) = ~ 2* 1 The factor 2! on the right accounts for the permutations of the two lines at the root; the binomial coefficient in the sum accounts for the assignment of labels to the two trees connected to the two lines, and, as above, each point of a tree connected to a line at the root has its height increased by one. The symbolic version of (9) is (9a) 2\J n (y; 2) = ny^ij+j)^, J* = J k (y), and of course J 0 (y) = 0. It is clear that

3 280 John Riordan and N. J. A. Sloane [3J (10) m\j n {y, m) = ny»- 1 (J+.+J) n '\ P ^ JM, with m terms in the symbolic sum on the right, and since it follows at once that m=l (1) J{x,y) =xexpj(xy,y). Rewriting (1) as ^ J { x. y ) = l r J,1 \ n=i n\ la = exp [xyj 1 (y)+ ^ and noting the equation of definition for the Bell multivariable polynomials Y n (y v y 2,..., y n ) (cf. [5], equation (45) of chapter 2), namely it follows that «2/i+ ~ + I or Jn+ite)l(n+i) - y n Y n {JM> -,Jn{y)) (ii) J n+ i(y) = (»+i)y"y,.(7i(y),,/ (</)) The first few instances of (11) are, omitting functional arguments, J* = Js = Next, (partial) differentiation of (1) with respect to x, results in, using the suffix notation for partial derivatives, (12) xj.(x, y) = J(x, y)[l+xj x (xy, y)} which entails the recurrence (13) nj n (y) = J n (y) + "Z, (*) ky*j k {y)jn-m-

4 [4] The enumeration of rooted trees 281 Similarly (14) J,(x, y) = J(x, y) so that «! k=l = R(x)[xR'(x) = R(x)[l-R(x)]- 1 xr'(x) which is (2). The identity xr'(x) = R(x)[l R(x)]- 1 used in finding the last line follows from differentiating R(x) = x exp R(x). Since it follows from (2) that n=l n - (15) W n = 2 (?) {n-k) n - k k k. From the Cauchy formula (equation (lc) of [6]), associated with Abel's generalization of the binomial formula, (15) is readily reduced to the first equation of (3); the second of (3) follows from Stirling's formula and [8], It is worth noting that the first of (3) implies relations of W n with other graph entities. First, from equation (10) of [1], the number of connected graphs with n labeled points and n labeled lines, C nn = T nnx, is given by Hence C nn = lw!(n-l)! f^. (n-l)\w n = 2C nn. Next, the number of connected graphs with n labeled points is given in [2] by «-i nk Hence W n = n(c n ~R n ). 3. More general enumeration The results given in (5), (6), and (7) are straightforward consequences of the use of George Polya's fundamental theorem in enumerative combi-

5 282 John Riordan and N. J. A. Sloane [5] natorial analysis. The essential modification of the procedure in [5] for enumeration of rooted trees without regard to total height, is that used above: a tree added to a line at the root occasions an increase by unity of the heights of each of its points. References [1] T. L. Austin, R. E. Fagen, W. F. Penney, and John Riordan, 'The number of components in random linear graphs', Ann. Math. Statist. 30 (1959), [2] Leo Katz, 'Probability of indecomposability of a random mapping function', Ann. Math. Statist. 26 (1955), [3] John Riordan, 'The enumeration of trees by height and diameter', IBM Jnl. Res. Dev. 4 (1960), [4] A. Renyi and G. Szekeres, 'On the height of trees', /. Aust. Math. Soc. 7 (1967), [5] John Riordan, An Introduction to Combinatorial Analysis, (Wiley, New York, 1958). [6] H. Salie, 'Uber Abel's Verallgemeinerung der binomischen Formel', Ber. Verh. Sdch. Akad. Wiss. Leipzig Math. - Nat. Kl. 98, No. 4 (1951), [7] G. P61ya, 'Kombinatorische Anzahlbestimmungen fur Gruppen, Graphen, und chemische Verbindungen', Ada Math. 68 (1937), [8] G. N. Watson, 'Theorems stated by Ramanujan (V): Approximations connected with e x, Proc. London Math. Soc. (2), 29 (1929), Rockefeller University and Cornell University

### Department of Information and Computer Scwnce, Unwers~ty of California, lrvine, California 92717

Some Techniques for Solving Recurrences GEORGE S. LUEKER Department of Information and Computer Scwnce, Unwers~ty of California, lrvine, California 92717 Some techmques for solving recurrences are presented,

### LAGUERRE POLYNOMIALS WITH GALOIS GROUP A m FOR EACH m

LAGUERRE POLYNOMIALS WITH GALOIS GROUP A m FOR EACH m Michael Filaseta Mathematics Department University of South Carolina Columbia, SC 29208 Ognian Trifonov Mathematics Department University of South

### Regular Languages are Testable with a Constant Number of Queries

Regular Languages are Testable with a Constant Number of Queries Noga Alon Michael Krivelevich Ilan Newman Mario Szegedy Abstract We continue the study of combinatorial property testing, initiated by Goldreich,

### Random matchings which induce Hamilton cycles, and hamiltonian decompositions of random regular graphs

Random matchings which induce Hamilton cycles, and hamiltonian decompositions of random regular graphs Jeong Han Kim Microsoft Research One Microsoft Way Redmond, WA 9805 USA jehkim@microsoft.com Nicholas

### Revised Version of Chapter 23. We learned long ago how to solve linear congruences. ax c (mod m)

Chapter 23 Squares Modulo p Revised Version of Chapter 23 We learned long ago how to solve linear congruences ax c (mod m) (see Chapter 8). It s now time to take the plunge and move on to quadratic equations.

### = 2 + 1 2 2 = 3 4, Now assume that P (k) is true for some fixed k 2. This means that

Instructions. Answer each of the questions on your own paper, and be sure to show your work so that partial credit can be adequately assessed. Credit will not be given for answers (even correct ones) without

How Bad is Forming Your Own Opinion? David Bindel Jon Kleinberg Sigal Oren August, 0 Abstract A long-standing line of work in economic theory has studied models by which a group of people in a social network,

### Semi-Simple Lie Algebras and Their Representations

i Semi-Simple Lie Algebras and Their Representations Robert N. Cahn Lawrence Berkeley Laboratory University of California Berkeley, California 1984 THE BENJAMIN/CUMMINGS PUBLISHING COMPANY Advanced Book

### ONE-DIMENSIONAL RANDOM WALKS 1. SIMPLE RANDOM WALK

ONE-DIMENSIONAL RANDOM WALKS 1. SIMPLE RANDOM WALK Definition 1. A random walk on the integers with step distribution F and initial state x is a sequence S n of random variables whose increments are independent,

### On line construction of suffix trees 1

(To appear in ALGORITHMICA) On line construction of suffix trees 1 Esko Ukkonen Department of Computer Science, University of Helsinki, P. O. Box 26 (Teollisuuskatu 23), FIN 00014 University of Helsinki,

### WHAT ARE MATHEMATICAL PROOFS AND WHY THEY ARE IMPORTANT?

WHAT ARE MATHEMATICAL PROOFS AND WHY THEY ARE IMPORTANT? introduction Many students seem to have trouble with the notion of a mathematical proof. People that come to a course like Math 216, who certainly

### An Impossibility Theorem for Clustering

An Impossibility Theorem for Clustering Jon Kleinberg Department of Computer Science Cornell University Ithaca NY 14853 Abstract Although the study of clustering is centered around an intuitively compelling

### Orthogonal Bases and the QR Algorithm

Orthogonal Bases and the QR Algorithm Orthogonal Bases by Peter J Olver University of Minnesota Throughout, we work in the Euclidean vector space V = R n, the space of column vectors with n real entries

### =< x, and the logarithm of the least common multiple of all positive integers PRIME NUMBERS APPROXIMATE FORMULAS FOR SOME FUNCTIONS OF

APPROXIMATE FORMULAS FOR SOME FUNCTIONS OF PRIME NUMBERS Dedicated to Hans Rademacher on the occasion of his seventieth birthday BY J. BARKLEY ROSSER AND LOWELL SCHOENFELD 1. Acknowledgments The maior

### Competing in the Dark: An Efficient Algorithm for Bandit Linear Optimization

Competing in the Dark: An Efficient Algorithm for Bandit Linear Optimization Jacob Abernethy Computer Science Division UC Berkeley jake@cs.berkeley.edu Elad Hazan IBM Almaden hazan@us.ibm.com Alexander

### (LMCS, p. 317) V.1. First Order Logic. This is the most powerful, most expressive logic that we will examine.

(LMCS, p. 317) V.1 First Order Logic This is the most powerful, most expressive logic that we will examine. Our version of first-order logic will use the following symbols: variables connectives (,,,,

### THE DIFFERENCE BETWEEN CONSECUTIVE PRIMES, II

THE DIFFERENCE BETWEEN CONSECUTIVE PRIMES, II R. C. BAKER, G. HARMAN and J. PINTZ [Received May 000; revised 5 November 000]. Introduction Beginning with Hoheisel [8], many authors have found shorter and

### ON SOME EXTREMUM PROBLEMS IN ELEMENTARY GEOMETRY

ON SOME EXTREMUM PROBLEMS IN ELEMENTARY GEOMETRY P. ERD& and G. SZEKERES* Mathematical Institute, EBtVBs Lot-and University, Budapest and University Adelaide (Received May 20, 7960) Dedicated to the memory

### The Backpropagation Algorithm

7 The Backpropagation Algorithm 7. Learning as gradient descent We saw in the last chapter that multilayered networks are capable of computing a wider range of Boolean functions than networks with a single

### Some Basic Techniques of Group Theory

Chapter 5 Some Basic Techniques of Group Theory 5.1 Groups Acting on Sets In this chapter we are going to analyze and classify groups, and, if possible, break down complicated groups into simpler components.

### Chapter 9 Unitary Groups and SU(N)

Chapter 9 Unitary Groups and SU(N) The irreducible representations of SO(3) are appropriate for describing the degeneracies of states of quantum mechanical systems which have rotational symmetry in three

### 2 Polynomials over a field

2 Polynomials over a field A polynomial over a field F is a sequence (a 0, a 1, a 2,, a n, ) where a i F i with a i = 0 from some point on a i is called the i th coefficient of f We define three special

### Which graphs are determined by their spectrum?

Linear Algebra and its Applications 373 (2003) 241 272 www.elsevier.com/locate/laa Which graphs are determined by their spectrum? Edwin R. van Dam 1, Willem H. Haemers Department of Econometrics and O.R.,

### Some exactly solvable models of urn process theory

Fourth Colloquium on Mathematics and Computer Science DMTCS proc AG, 26, 59 8 Some exactly solvable models of urn process theory Philippe Flajolet, Philippe Dumas, and Vincent Puyhaubert Algorithms Project,

### Communication Theory of Secrecy Systems

Communication Theory of Secrecy Systems By C. E. SHANNON 1 INTRODUCTION AND SUMMARY The problems of cryptography and secrecy systems furnish an interesting application of communication theory 1. In this

### Math 2001 Homework #10 Solutions

Math 00 Homework #0 Solutions. Section.: ab. For each map below, determine the number of southerly paths from point to point. Solution: We just have to use the same process as we did in building Pascal

### Switching Algebra and Logic Gates

Chapter 2 Switching Algebra and Logic Gates The word algebra in the title of this chapter should alert you that more mathematics is coming. No doubt, some of you are itching to get on with digital design

### Some Optimal Inapproximability Results

Some Optimal Inapproximability Results JOHAN HÅSTAD Royal Institute of Technology, Stockholm, Sweden Abstract. We prove optimal, up to an arbitrary ɛ>0, inapproximability results for Max-Ek-Sat for k 3,