Author: Dieter Buerssner
Date: 10:57:07 09/04/03
Go up one level in this thread
On September 03, 2003 at 22:59:17, Tom Likens wrote: >On September 03, 2003 at 16:28:14, Dieter Buerssner wrote: > >>On September 03, 2003 at 13:43:13, Tom Likens wrote: >> >>>Getting the hamming distance in pseudo-code for N random numbers: >>> >>>1. Generate a random number (index=M) >>> >>>2. Compare it to the 0 ... M-1 valid random numbers alread saved >>> >>> if (popcnt64(new_rand64 ^ array[0..M-1]) >= MIN_HAM) then OK >>> >>>3. If valid, save it into slot M >>> If not valid (hamming distance is too small) goto 1 BTW. This can give an infinite loop. >>>4. Repeat until you have N random numbers >> >>They aren't random anymore, after you filtered them like that. >You're correct of course, the set of numbers generated with the above >algorithm is no longer pseudo-random and I should have been more precise >in my wording. I hope, I didn't sound too pedantic. It was just meant as a side note. I assumed, you were already aware of it. The main point of my posting was, to give a reference to that earlier discussion. I tried different PRNGs (including Mersenne twister) of different quality, and could not see any change (besides noise) in some test suites. Regards, 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.