VMs: Huffman compression (was: VMs: Declaration of WAR against EVA)

[Jacques Guy <jguy@xxxxxxxxxxxxxxxx>]

3/6/03 8:07:06 AM, "Petr Kazil" <kazil@xxxxxxxxxx> wrote:

>In my last mail I stated that there should be a direct algorithm to segment the 
VMS into groups of
>characters in such a way that the number of different groups is minimized and 
mean group-length is
>maximized.

No, that can't be right. Think. The solution is trivial: 

group: the whole message
group length: maximal (the length of the message)
number of different groups: minimal (one)


>I remembered that there exists such an algorithm. It's called "Huffman Coding". 
It's 20 years ago so
>I forgot the details, but the algorithm selects the optimal code for a data-
stream so that all
>redundancy is removed. I think that WinZip uses an "Adaptive Huffman Coding" 
scheme.

Yes. So does LZH. 

>One of you last cited a book on information theory. This coding should be 
described in there. Can
>you look it up and see if it's relevant?

Much better:

The Huffman compression algorithm in Euphoria

http://www.RapidEuphoria.com/huff.zip

Don't worry if you've never heard of Euphoria. If you know a bit
of C, a bit of Pascal (or ALGOL, or Ada, or Oberon or even... BASIC!)
you will have no trouble at all.





______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list