Re: Efficient way to store an isomorphism?
- From: "Matt Timmermans" <mt0000@xxxxxxxxxxxxxxxxxxxxxxxxxx>
- Date: Thu, 11 Aug 2005 23:10:31 -0400
"Edsko de Vries" <devriese@xxxxxxxxx> wrote in message
news:1123760098.967374.326650@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> [...] can it be shown that such an improvement is not
> possible?
Easily: The hash function you want specifies a chosen one of N!
permutations of the N items in its range. To specify one of N! permutations
takes ceil(log_2(N!)) bits, which is O(N) words, assuming that words are at
least log_2(N) bits wide.
--
Matt
.
- References:
- Efficient way to store an isomorphism?
- From: Edsko de Vries
- Re: Efficient way to store an isomorphism?
- From: Matt Timmermans
- Re: Efficient way to store an isomorphism?
- From: Edsko de Vries
- Efficient way to store an isomorphism?
- Prev by Date: proper deletion algorithm for Patricia trie
- Next by Date: Re: proper deletion algorithm for Patricia trie
- Previous by thread: Re: Efficient way to store an isomorphism?
- Next by thread: Re: Efficient way to store an isomorphism?
- Index(es):