ECJ-INTEREST-L Archives

July 2013

ECJ-INTEREST-L@LISTSERV.GMU.EDU
Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
Re: Random number generator
From:
Sean Luke <[log in to unmask]>
Reply To:
ECJ Evolutionary Computation Toolkit <[log in to unmask]>
Date:
Mon, 15 Jul 2013 16:17:41 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (17 lines) 
On Jul 15, 2013, at 10:03 AM, SUBSCRIBE ECJ-INTEREST-L Anonymous wrote:

> I created 12 threads and specified 12 seeds for each, ran it twice but didn't get 
> replicable results.
> 
> Anybody have any ideas?

The latest version of ECJ introduced a new SimpleEvaluator which could have caused a bug, but it seems more likely that you might not be practicing good random number generator hygene in your problem.

	- Always use state.random[threadnum], unless you *know* you are in a single-threaded environment and don't have a threadnum, in which case use state.random[0]

	- Never use java.util.Random, Math.random(), etc.

Could you give an example of ECJ not producing replicable results on any of the demo problems provided?  Please use the SVN repository version.

Sean

ATOM RSS1 RSS2