Re: Fibonacci coding

On Oct 25, 10:35 pm, biject <davvid_a_sc...@xxxxxxxxx> wrote:
On Oct 24, 7:59 pm, Fibonacci Code <angli...@xxxxxxxxx> wrote:





On Oct 25, 5:40 am, Jim Leonard <MobyGa...@xxxxxxxxx> wrote:

On Oct 24, 11:35 am, Fibonanci Code <angli...@xxxxxxxxx> wrote:

Hi All,

Just wanna know where does Fibonanci Coding use practically ?
Is there any practical codec, compression program is using fibo code ?

A google search for "Fibonacci compression" will turn up some
interesting uses, but in all cases a better method is found. There
was an audio compression method that used a Fibonacci sequence to
describe the delta step (one reference ishttp://groups.google.com/group/comp.compression/msg/108c97ae8677bea4
and another is documented as the IFF 8SVX chunk in the EA '85 IFF
standard) but other than that I don't have any particular favorites :-)

Thanks a lot Jim,

As I study universal code, really hard to find practical uses
of the fibo encoding,
if one day I found one practicle lossless usage, I will be hitting the
jacpot.

Regards,
Ray

I am not sure what you mean by a lossless usage. Its easy to
write bijective fibonanocci coding. Any type of compression where you
use universal numbers could be done with fibonancci coding since it
can easily be made bijective depending on the distribution of numbers
it could be the oprimal. Just like bijective huffman coding. Some
times
its better than bijective arithmetic coding it just depends on file
distribution.

A trival example would be suppose you have a varible list of integer
numbers
where you excule the null input and output file cases. The input set
being a list
of numbers from one to infinity
if you are not copressing the last number use the full fibonannci
binary string
for that number
11 is 1
011 is 2
0011 is 3
1011 is 4
00011 is 5 and etc

If the list is only one number long you add 2 to the number and encode
all the bits
except the trailing 011 eaxmple
1 goes to 3 encode as 0
2 goes to 4 encode as 1
3 goes to 5 encode as 00 and etc

The one case left is encoding the string for the last
number when there is one or more preceeding numbers
Since the previous number ends in 011 if the last number
is a one just output nothing your done. if the last number
greater than one its sort of like first number but you add 1
and encode and drop the last 011]
example for last number when list longer than one number
1 null nothing
2 goes to 3 encode as 0
3 goes to 4 encode as 1
4 goes to 5 encode as 00 and etc

it this point you have the set of add binary strings as outputs.
you can easily use me code to convert this to files in a bijective
manner

David A. Scott
--
My Crypto codehttp://bijective.dogma.net/crypto/scott19u.ziphttp://www.jim.com/jamesd/Kong/scott19u.zipold version
My Compression codehttp://bijective.dogma.net/
**TO EMAIL ME drop the roman "five" **
Disclaimer:I am in no way responsible for any of the statements
made in the above text. For all I know I might be drugged.
As a famous person once said "any cryptograhic
system is only as strong as its weakest link"- Hide quoted text -

- Show quoted text -

Hi David,

The lossless I mean is, is there any general compressor using
fibonacci code eg. 11, 011, 0011, 1011, etc as the entropy coding ?
I didn't mind it is less efficient, as long as the output is coded as
fibo code
and there is some compression ratio from the existing source data,
I've hit the jackpot.

Many thanks.
Ray


I've been thinking also about your bijective compressor, it is just my
two cents
I am not sure wether it will work or not. You can create as many
bijective algorithm
to handle the block of data and compress / uncompress it. For example
I give 3


Compression
-----------------------------------------------------------
- bijective
arithmatic -> compressed block ari
Source data block goes into > - bijective huffman ->
compressed block huffman
- bijective
rle -> compressed block rle

if compressed block is smallest length, take the compressed block
and write to destination file
if there is more then one same length of the smallest, take the
first one.
if all compressed block is bigger then or equal the source block ->
write source block to destination file (uncompressed)

Decompression
----------------------------------------------------------

Source data block goes into > - bijective airthmatic -> expanded
block ari data block
- bijective
huffman -> expanded huffman data block
- bijective
rle -> expanded rle data block

if expand block is largest in length, take the block and write to
output file
if there is more then one with same length, take the first one.
if all expanded block is smaller then or equal the source block ->
write source block to output file.

Thanks.

.

Relevant Pages

  • Re: RLE question
    ... > encoding the escape code and a pad value. ... > so, this scheme does worse, does it have a plus? ... > compression than is possible with a pcx-like scheme. ...
    (comp.compression)
  • Re: RLE question
    ... encoding the escape code and a pad value. ... compression than is possible with a pcx-like scheme. ... mapped over to low probability symbols, or all the high probability symbols ...
    (comp.compression)
  • Re: OE attachments grow in file size
    ... If you attach a binary file, like a JPEG, the encoded MIME part ... > includes compression but the less compressible is a then the larger it ... > If you are using UUencode to encode your attached file ... > to find the RFCs that describe MIME and read how it encodes attachments. ...
    (microsoft.public.windows.inetexplorer.ie6_outlookexpress)
  • Re: OE attachments grow in file size
    ... > But I have tried the same thing on my corporate email (Outlook with MS Exchange server)> and the files do not grow with this magnitude. ... If you attach a binary file, like a JPEG, the>> encoded MIME part containing the definition of the "attached" file will look very>> different than what you would see in a hex editor. ... So you reduced the file to 50% of its size by compressing it but you only>> gained a 20% for the size of the e-mail (i.e., encoding to plain text with compression>> an already compressed file will actually inflate the size of your e-mail message). ... >> If you are using UUencode to encode your attached file then go Google>> on UUencode/UUdecode to see how encoding works for attaching files using that scheme. ...
    (microsoft.public.windows.inetexplorer.ie6_outlookexpress)
  • Re: Roberts Radio
    ... range compression) that actually makes it easier to encode than pop and ... providing Radio 3 at higher audio quality than all the other radio ... compression and stuff and the difference is immediately obvious to ...
    (alt.radio.digital)