notes-090924 - MATH 681 1 1.1 Notes Combinatorics and Graph...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
MATH 681 Notes Combinatorics and Graph Theory I 1 Generating functions, continued 1.1 Exponential generating functions and set-partitions At this point, we’ve come up with good generating-function discussions based on 3 of the 4 rows of our twelvefold way. Will our integer-partition tricks work with exponential generating functions to give us rules for putting distinct balls into indistinct boxes (a.k.a. set-partitions)? Recall that for the integer-partition problem our series of sequential procedures was to first select the number of boxes with one ball, then the number of boxes with 2 balls, then the number of boxes with 3 balls, etc. Would that work in this case? Let’s try it: Question 1: Find an exponential generating function for the number of ways to distribute n balls among unlabeled boxes. Answer 1: An exponential generating function for the number of ways to select boxes with 1 ball in is easy: there is only one way to fill zero boxes, one way to fill one box, etc., so the generating function would be: 1 + x + x 2 2! + ··· = X n =0 x n n ! = e x Filling boxes with two balls gets harder, however: there is one way to set up zero boxes with two balls each, one way to set up one box, three ways to set up two boxes, and fifteen ways to set up six, etc. This ends up being ugly pretty fast, and doesn’t lend itself to computation. However, since we already know this statistic (these are the Bell numbers), we shouldn’t have to go home empty-handed: we can actually work out what the generating function is! Recall that B n = n X k =1 S ( n,k ) = n X k =1 1 k ! k X i =0 ( - 1) k - i ± k i ² i n So the generating function we seek is the heinous expression: X n =0 B n x n n ! = X n =0 n X k =1 k X i =0 ( - 1) k - i ( k i ) i n n ! k ! x n This is actually pretty easy! We can find the generating function for S ( n,k ) from the known generating function for k ! S ( n,k ) : X n =0 k ! S ( n,k ) x n n ! = ( e x - 1) k so X n =0 S ( n,k ) x n n ! = ( e x - 1) k k ! and since B n = n k =0 S ( n,k ) , and we can include the zero terms when k > n to get B n = Page 1 of 9 September 24, 2009
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
MATH 681 Notes Combinatorics and Graph Theory I k =0 S ( n,k ) , then: X n =0 B n x n n ! = X n =0 X k =0 S ( n,k ) x n n ! = X k =0 X n =0 S ( n,k ) x n n ! = X k =0 ( e x - 1) k k ! = e ( e x - 1) As far as I know, there is no “from first principles” argument to derive this same elegant and simple form. 1.2 Proving binomial identities with generating functions We can prove binomial identities with generating functions, if we know the GFs of the various binomial expressions. For instance, here’s one from the last problem set: Question 2: Prove that n i =0 i ( n i ) = n 2 n - 1 . Answer 2:
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 01/12/2012 for the course MATH 681 taught by Professor Wildstrom during the Fall '09 term at University of Louisville.

Page1 / 9

notes-090924 - MATH 681 1 1.1 Notes Combinatorics and Graph...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online