Robowiki - User contributions [en]

RobocodeJGAP

2008-12-14T17:39:09Z

Klausikm: Added JGAP logo

<center>[http://jgap.sf.net/doc/robocode/robocode.html http://testwiki.roborumble.org/w/images/7/73/RobocodeJGAP.png]</center>
= Overview =
[http://jgap.sf.net/doc/robocode/robocode.html RobocodeJGAP] is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package).
[http://jgap.sf.net JGAP] is an open source Java framework for Genetic Algorithms and Genetic Programming. RobocodeJGAP is based on JGAP.
The RobocodeJGAP project can be also [http://jgap.sourceforge.net/doc/robocode/robocode.html downloaded].
JGAP and RobocodeJGAP are maintained by [[Klaus Meffert]].

== Testbed and fitness computation ==

RobocodeJGAP evolves a robot and then let it fight against one or many robots in each evolution step (but only one opponent at a time). It can be configured whether one or more robots are registered as opponents. Currently 3 opponents are configured. The opponents are already existent, regular robots taken from the Robocode repository and the samples provides with robocode. The 3 robots are:
voidious.Dookious (the best robot in the overall ranking)
sample.TrackFire (supplied with Robocode)
bestaspirant.Aspirant_winner (an evolved robot that has quite a good score)
Two modes of operation are possible: a) fight one of the three robots each evolution step (randomly chosen) in a one-to-one battle b)fight all of the three robots and calculate the average.
Currently mode b) is set.

The fitness value of a robot is calculated after each battle. For that the ratio of the own score to the sum of both robots scores is taken. To scale the score up, it is artifically multiplied by 1000.

If fighting against three enemies one after another in a single evolution round, and the difference between the best score of our evolved robot and the worst score is too high, the fitness average is reduced a bit. The idea behind that is that an enemy that could be quite good, may have a partial weakness we just explored. But that doesn't help much, normally, in fighting against 200 other robots, like in RoboRumble.

<center>[http://jgap.sf.net http://jgap.sourceforge.net/jgap-logo.jpeg]</center>

RobocodeJGAP

2008-12-14T17:37:52Z

Klausikm: Fixed image link

RobocodeJGAP

2008-12-14T17:31:53Z

Klausikm: Added link to RobocodeJGAP

<center>[http://jgap.sf.net/doc/robocode/robocode.html [[Image:RobocodeJGAP.png]]]</center>
= Overview =
[http://jgap.sf.net/doc/robocode/robocode.html RobocodeJGAP] is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package).
[http://jgap.sf.net JGAP] is an open source Java framework for Genetic Algorithms and Genetic Programming. RobocodeJGAP is based on JGAP.
The RobocodeJGAP project can be also [http://jgap.sourceforge.net/doc/robocode/robocode.html downloaded].
JGAP and RobocodeJGAP are maintained by [[Klaus Meffert]].

== Testbed and fitness computation ==

RobocodeJGAP evolves a robot and then let it fight against one or many robots in each evolution step (but only one opponent at a time). It can be configured whether one or more robots are registered as opponents. Currently 3 opponents are configured. The opponents are already existent, regular robots taken from the Robocode repository and the samples provides with robocode. The 3 robots are:
voidious.Dookious (the best robot in the overall ranking)
sample.TrackFire (supplied with Robocode)
bestaspirant.Aspirant_winner (an evolved robot that has quite a good score)
Two modes of operation are possible: a) fight one of the three robots each evolution step (randomly chosen) in a one-to-one battle b)fight all of the three robots and calculate the average.
Currently mode b) is set.

The fitness value of a robot is calculated after each battle. For that the ratio of the own score to the sum of both robots scores is taken. To scale the score up, it is artifically multiplied by 1000.

If fighting against three enemies one after another in a single evolution round, and the difference between the best score of our evolved robot and the worst score is too high, the fitness average is reduced a bit. The idea behind that is that an enemy that could be quite good, may have a partial weakness we just explored. But that doesn't help much, normally, in fighting against 200 other robots, like in RoboRumble.

RobocodeJGAP

2008-12-14T17:31:21Z

Klausikm: Added logo

<center>[http://jgap.sf.net/doc/robocode/robocode.html [[Image:RobocodeJGAP.png]]]</center>
= Overview =
RobocodeJGAP is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package).
[http://jgap.sf.net JGAP] is an open source Java framework for Genetic Algorithms and Genetic Programming. RobocodeJGAP is based on JGAP.
The RobocodeJGAP project can be also [http://jgap.sourceforge.net/doc/robocode/robocode.html downloaded].
JGAP and RobocodeJGAP are maintained by [[Klaus Meffert]].

== Testbed and fitness computation ==

RobocodeJGAP evolves a robot and then let it fight against one or many robots in each evolution step (but only one opponent at a time). It can be configured whether one or more robots are registered as opponents. Currently 3 opponents are configured. The opponents are already existent, regular robots taken from the Robocode repository and the samples provides with robocode. The 3 robots are:
voidious.Dookious (the best robot in the overall ranking)
sample.TrackFire (supplied with Robocode)
bestaspirant.Aspirant_winner (an evolved robot that has quite a good score)
Two modes of operation are possible: a) fight one of the three robots each evolution step (randomly chosen) in a one-to-one battle b)fight all of the three robots and calculate the average.
Currently mode b) is set.

The fitness value of a robot is calculated after each battle. For that the ratio of the own score to the sum of both robots scores is taken. To scale the score up, it is artifically multiplied by 1000.

If fighting against three enemies one after another in a single evolution round, and the difference between the best score of our evolved robot and the worst score is too high, the fitness average is reduced a bit. The idea behind that is that an enemy that could be quite good, may have a partial weakness we just explored. But that doesn't help much, normally, in fighting against 200 other robots, like in RoboRumble.

Talk:RobocodeJGAP

2008-12-14T17:28:00Z

Klausikm: /* Choice of opponants */

== Choice of opponents ==

Hmm, I wonder if voidious.Dookious, sample.TrackFire, and bestaspirant.Aspirant_winner are the best choices really. I think Dookious is likely strong to such a point that the outcome of matches with it is probably predominantly random noise until you get to a certain strength level, and I think sample.TrackFire might be a bit too atypical of what's out there. In my mind it would make more sense to maybe choose something like:
* a mid strength head-on-targeter (hopefully to teach it of head-on avoidance)
* a rambot (there are a few of them out there, it's good to learn how to avoid them)
* and a non-top nano patternmatcher (there are quite a number of simple patternmatcher guns out there).
or maybe replace the rambot with some non-patternmatcher mid strength nanobot if you decide to not care about rammers. In addition, getting a learning-heavy bot like Dookious out of the mix would make it possible to run a bit faster. What do others think? --[[User:Rednaxela|Rednaxela]] 16:57, 11 December 2008 (UTC)

File:RobocodeJGAP.png

2008-12-14T17:27:27Z

Klausikm: RobocodeJGAP logo

RobocodeJGAP logo

Klaus Meffert

2008-12-10T17:42:42Z

Klausikm:

[[Image:klaus_small.jpg]]

Main author and admin of the projects [http://jgap.sf.net JGAP] and [[RobocodeJGAP]]. Currently, JGAP is driven towards grid technology. RobocodeJGAP is already being working on a prototype grid. A PHP hosted V-server is planned for distributing work requests and obtaining results. A prior experiment with Google documents as free and intermediate online store failed due to Google's undocumented restrictions.

Born 1974 in Germany, and still living there luckily.

I am working as a software consultant and developer, mainly in the SAP area, module SRM, but also in Java technologies.

[http://www.klaus-meffert.com Homepage of Klaus]

File:Klaus small.jpg

2008-12-10T17:42:26Z

Klausikm: Klaus

Klaus

User:Klausikm

2008-12-10T17:40:23Z

Klausikm: New page: See Klaus Meffert

See [[Klaus Meffert]]

Klaus Meffert

2008-12-10T17:37:33Z

Klausikm:

[[Image:http://jgap.sourceforge.net/doc/contrib/klaus_small.jpg]]
Main author and admin of the projects [http://jgap.sf.net JGAP] and [[RobocodeJGAP]]. Currently, JGAP is driven towards grid technology. RobocodeJGAP is already being working on a prototype grid. A PHP hosted V-server is planned for distributing work requests and obtaining results. A prior experiment with Google documents as free and intermediate online store failed due to Google's undocumented restrictions.

Klaus is working as a software consultant and developer, mainly in the SAP area, module SRM, but also in Java technologies.

[http://www.klaus-meffert.com Homepage of Klaus]

Klaus Meffert

2008-12-10T17:35:38Z

Klausikm:

Main author and admin of the projects [http://jgap.sf.net JGAP] and [[RobocodeJGAP]]. Currently, JGAP is driven towards grid technology. RobocodeJGAP is already being working on a prototype grid. A PHP hosted V-server is planned for distributing work requests and obtaining results. A prior experiment with Google documents as free and intermediate online store failed due to Google's undocumented restrictions.

Klaus is working as a software consultant and developer, mainly in the SAP area, module SRM, but also in Java technologies.

[http://www.klaus-meffert.com Homepage of Klaus]

Klaus Meffert

2008-12-10T17:33:22Z

Klausikm:

Main author and admin of the projects [http://jgap.sf.net JGAP] and [[RobocodeJGAP]].

Klaus is working as a software consultant and developer, mainly in the SAP area, module SRM, but also in Java technologies.

[http://www.klaus-meffert.com Homepage of Klaus]

Klaus Meffert

2008-12-10T17:33:07Z

Klausikm: New page: Main author and admin of the project [http://jgap.sf.net JGAP] and RobocodeJGAP. Klaus is working as a software consultant and developer, mainly in the SAP area, module SRM, but also ...

Main author and admin of the project [http://jgap.sf.net JGAP] and [[RobocodeJGAP]].

Klaus is working as a software consultant and developer, mainly in the SAP area, module SRM, but also in Java technologies.

[http://www.klaus-meffert.com Homepage of Klaus]

RobocodeJGAP

2008-12-10T17:30:51Z

Klausikm:

= Overview =
RobocodeJGAP is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package).
[http://jgap.sf.net JGAP] is an open source Java framework for Genetic Algorithms and Genetic Programming. RobocodeJGAP is based on JGAP.
The RobocodeJGAP project can be also [http://jgap.sourceforge.net/doc/robocode/robocode.html downloaded].
JGAP and RobocodeJGAP are maintained by [[Klaus Meffert]].

== Testbed and fitness computation ==

RobocodeJGAP evolves a robot and then let it fight against one or many robots in each evolution step (but only one opponent at a time). It can be configured whether one or more robots are registered as opponents. Currently 3 opponents are configured. The opponents are already existent, regular robots taken from the Robocode repository and the samples provides with robocode. The 3 robots are:
voidious.Dookious (the best robot in the overall ranking)
sample.TrackFire (supplied with Robocode)
bestaspirant.Aspirant_winner (an evolved robot that has quite a good score)
Two modes of operation are possible: a) fight one of the three robots each evolution step (randomly chosen) in a one-to-one battle b)fight all of the three robots and calculate the average.
Currently mode b) is set.

The fitness value of a robot is calculated after each battle. For that the ratio of the own score to the sum of both robots scores is taken. To scale the score up, it is artifically multiplied by 1000.

If fighting against three enemies one after another in a single evolution round, and the difference between the best score of our evolved robot and the worst score is too high, the fitness average is reduced a bit. The idea behind that is that an enemy that could be quite good, may have a partial weakness we just explored. But that doesn't help much, normally, in fighting against 200 other robots, like in RoboRumble.

RobocodeJGAP

2008-12-10T17:29:16Z

Klausikm: New page: = Overview = RobocodeJGAP is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package). [http://jgap.sf...

= Overview =
RobocodeJGAP is an approach to automatically evolve Robocode via the Genetic Programming engine JGAP (Java Genetic Algorithms and Genetic Programming Package).
[http://jgap.sf.net]JGAP is an open source Java framework and can be found at http://jgap.sf.net. RobocodeJGAP is based on JGAP.
The RobocodeJGAP project can be downloaded from http://jgap.sourceforge.net/doc/robocode/robocode.html
JGAP and RobocodeJGAP are maintained by [[Klaus Meffert]].

== Testbed and fitness computation ==

RobocodeJGAP evolves a robot and then let it fight against one or many robots in each evolution step (but only one opponent at a time). It can be configured whether one or more robots are registered as opponents. Currently 3 opponents are configured. The opponents are already existent, regular robots taken from the Robocode repository and the samples provides with robocode. The 3 robots are:
voidious.Dookious (the best robot in the overall ranking)
sample.TrackFire (supplied with Robocode)
bestaspirant.Aspirant_winner (an evolved robot that has quite a good score)
Two modes of operation are possible: a) fight one of the three robots each evolution step (randomly chosen) in a one-to-one battle b)fight all of the three robots and calculate the average.

Currently mode b) is set.

The fitness value of a robot is calculated after each battle. For that the ratio of the own score to the sum of both robots scores is taken. To scale the score up, it is artifically multiplied by 1000.

If fighting against three enemies one after another in a single evolution round, and the difference between the best score of our evolved robot and the worst score is too high, the fitness average is reduced a bit. The idea behind that is that an enemy that could be quite good, may have a partial weakness we just explored. But that doesn't help much, normally, in fighting against 200 other robots, like in RoboRumble.