| 
November 21st, 2009, 11:12 AM
| | Junior Member | | Join Date: Nov 2008
Posts: 46
Country: Thanks: 4
Thanked 18 Times in 18 Posts
| | Combinatorics & Partitions proof I'm totally lost on this one... Any hints or solutions are welcome! Thanks in advance. For n ≥ 0, let Bn be the number of partitions of a set of size n. For example, B1 = 1, B2 = 2, B3 = 5, B4 = 15. By convention, B0 = 1. Now, let Dn be the number of partitions of a set of size n into subsets of size at least two. Show that <attachment> | 
November 21st, 2009, 05:32 PM
| | Member | | Join Date: Nov 2009
Posts: 231
Country: Thanks: 9
Thanked 77 Times in 73 Posts
| | Confusion over problem If by 'partition' you mean an unordered way to write a number n as a sum of positive integers, the first few partitions are:
P(1) = 1, as
1 can only be written as 1.
P(2) = 2, as
2 can be written as 2 or 1+1.
P(3) = 3, as
3 can be written as 3 or 2+1 or 1+1+1.
P(4) = 5, as
4 can be written as 4 or 3+1 or 2+2, 2+1+1, 1+1+1+1.
which don't match your problem description.
Please clarify your problem. | 
November 22nd, 2009, 02:10 AM
| | Junior Member | | Join Date: Nov 2008
Posts: 46
Country: Thanks: 4
Thanked 18 Times in 18 Posts
| | | 
November 22nd, 2009, 02:37 AM
| | Newbie | | Join Date: Nov 2009
Posts: 5
Country: Thanks: 2
Thanked 0 Times in 0 Posts
| | Partitions He's indeed referring to the Bell numbers, which are the number of partitions of a set (of n integers) into subsets, and not partitions of an integer.
Dn is simply the number of partitions of a set of integers, such that no partition has less than 2 elements (also known as singleton-free partitions, or 2-restricted Bell numbers, see integer sequence A000296).
While I can't prove the entire thing at the moment, I can say that Bi = Di + D(i+1), since if you take away all the singleton-free partitions of Bi, the remaining Bi - Di partitions will (obviously) all have at least one singleton. Group all these singletons together (partitionwise) and add the element i+1, now you have all the D(i+1) partitions.
The proof is very likely to have something to do with the sieve principle. Also, Dn are partition analogues of derangements, if that's any help. | 
November 22nd, 2009, 07:04 AM
| | Super Member | | Join Date: Mar 2008
Posts: 525
Country: Thanks: 23
Thanked 239 Times in 203 Posts
| | Quote:
Originally Posted by r0r0trog I'm totally lost on this one... Any hints or solutions are welcome! Thanks in advance. For n ≥ 0, let Bn be the number of partitions of a set of size n. For example, B1 = 1, B2 = 2, B3 = 5, B4 = 15. By convention, B0 = 1. Now, let Dn be the number of partitions of a set of size n into subsets of size at least two. Show that <attachment> | Hint:  is the number of set partitions that do not include any singleton subsets.
Let's say the original set is  . Say that a set partition has property i if one of the subsets in the partition is  . Now apply the Principle of Inclusion/Exclusion, also known as the Sieve Principle, to find the number of partitions which has none of the properties 1,2, ..., n. | | The following users thank awkward for this useful post: | |  | | Thread Tools | | | | Display Modes | Linear Mode |
Posting Rules
| You may not post new threads You may not post replies You may not post attachments You may not edit your posts HTML code is Off | | | All times are GMT -7. The time now is 06:24 AM. | | |