Talk:Truncated binary encoding
would be great to have either an exact computation rule for all n or more examples (thus users can extrapolate to their needs). I'm not sure whether this is correct, so please check it and maybe enrich the article
n = 2 (e.g. used in de:Golomb_Code#Beispiele, the German version of Golomb_coding)
| Number | Binary encoding | Truncated binary encoding |
| 0 | 00 | 0 |
| 1 | 01 | 10 (or maybe 01 ??) |
| 2 | 10 | 11 |
| 3 | 11 (UNUSED) | - |
--80.140.106.33 21:23, 20 Feb 2005 (UTC)
Shouldn't the page either
- define n as the size of the alphabet (e.g. for 0 through 4, n = 5, not 4), or
- say that binary codes are optimal when n (as defined currently, the highest number) is one less than a power of two?
--Steve Witham 7 March 2005