Register | Sign In


Understanding through Discussion


EvC Forum active members: 64 (9164 total)
5 online now:
Newest Member: ChatGPT
Post Volume: Total: 916,765 Year: 4,022/9,624 Month: 893/974 Week: 220/286 Day: 27/109 Hour: 1/2


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   BoxCar2D - neat little genetic algorithm demonstration
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 1 of 51 (604815)
02-15-2011 7:20 AM


www.boxcar2d.com
Basically someone's hooked up a 2d physics library to a genetic algorithm to evolve little "cars" that race across a random 2d landscape. It's really quite a neat demonstration of the efficacy of evolution in a simply toy environment.

Replies to this message:
 Message 2 by slevesque, posted 02-15-2011 5:41 PM Dr Jack has not replied
 Message 6 by AZPaul3, posted 02-16-2011 8:04 PM Dr Jack has not replied
 Message 17 by DBlevins, posted 02-17-2011 7:38 PM Dr Jack has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 9 of 51 (605121)
02-17-2011 5:29 AM
Reply to: Message 8 by Dr Adequate
02-17-2011 2:16 AM


I used the standard setup except that I changed the maximum wheel number to five.
Did you drop the wheel frequency as well? If not, that's probably why yours lagged behind. I presume it actually converged to a 2 wheel design?
Have you guys played with the different landscapes yet?

This message is a reply to:
 Message 8 by Dr Adequate, posted 02-17-2011 2:16 AM Dr Adequate has replied

Replies to this message:
 Message 12 by Dr Adequate, posted 02-17-2011 6:04 AM Dr Jack has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 15 of 51 (605203)
02-17-2011 4:33 PM
Reply to: Message 14 by Taq
02-17-2011 1:28 PM


Re: ID challenge?
Check the 'Best Cars' page and its comments...

This message is a reply to:
 Message 14 by Taq, posted 02-17-2011 1:28 PM Taq has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 25 of 51 (605261)
02-18-2011 3:40 AM
Reply to: Message 19 by Dr Adequate
02-17-2011 11:03 PM


By now, I think you guys have more experience of what this will evolve than I do

This message is a reply to:
 Message 19 by Dr Adequate, posted 02-17-2011 11:03 PM Dr Adequate has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 47 of 51 (605603)
02-21-2011 4:29 AM
Reply to: Message 46 by Minnemooseus
02-21-2011 4:20 AM


Re: I got to 1135.3 at 2:02 (and ran ~190 generations)
Minnemooseus writes:
I plugged my best effort (evolved from the defaults) into a zero mutations, 2 population, default track setup. There are still some tiny variables in there, you get different results in different tries. I was only able to get these tries to a ~900 best. I couldn't repeat my 1135.3 at 2:02.
Are you running with the 'default' track option? In that case, each track is random so the peak fitness will be different.

This message is a reply to:
 Message 46 by Minnemooseus, posted 02-21-2011 4:20 AM Minnemooseus has replied

Replies to this message:
 Message 48 by Minnemooseus, posted 02-21-2011 5:17 AM Dr Jack has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.4


Message 49 of 51 (606187)
02-24-2011 4:12 AM


Version 3.0 posted
Ah! I see he has updated the software, it's now posting as version 3.0 - not sure what's new yet though.

Replies to this message:
 Message 51 by slevesque, posted 02-24-2011 4:18 AM Dr Jack has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024