Author: Dieter Buerssner
Date: 13:18:57 04/21/05
Go up one level in this thread
On April 20, 2005 at 19:58:50, Dann Corbit wrote: >On April 20, 2005 at 17:02:13, Dieter Buerssner wrote: > >>On April 20, 2005 at 15:46:00, Dann Corbit wrote: >> >>>I saw an analysis once that said base e was the densest possible storage. >> >>Hmm, I don't understand this. Can you elaborate? > >I will have to search to find the document, which I read many years ago. It >explained that base e is the most compact way to store information and therefore >that base 3 circuits would be somewhat more efficient than base 2 because 3 is >closer to e than 2 is. > >But I do not remember the details. Even if you do not find the article, how would you define "densest possible storage" or "storage density". I would imagine base e encoding as, number = sum(i=0 to n) d_i * e^i. If I want to store the number 5, 2 "e-type digits" will be enough. 2 bits will not be enough. But naturally, the larger the base digit is, the fewer digits you will need. Somehow, one has to define the storage needs for a single digit. At the moment, I cannot imagine how one could define these needs in a way, that yields in a non monotonic "storage need", when we plot storage vs. base. But your sentance above: "base e is the most compact way to store information" seems to imply, that this is the case. Dieter
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.