|
|
| (5 intermediate revisions by the same user not shown) |
| Line 1: |
Line 1: |
| | I'm DavidR from Portugal and I'm a new user at Robocode and I'm doing research in Swarm intelligence, Synchronization and Networks, Robocode is a very interesting way to develop some of the ideas I have. I'm coding my first robot at the moment and hope to have it ready by the end of the year (If I manage to get the time to code as my PhD is taking 99% of my time now.) | | I'm DavidR from Portugal and I'm a new user at Robocode and I'm doing research in Swarm intelligence, Synchronization and Networks, Robocode is a very interesting way to develop some of the ideas I have. I'm coding my first robot at the moment and hope to have it ready by the end of the year (If I manage to get the time to code as my PhD is taking 99% of my time now.) |
| | | | |
| − | == MiniR101 == | + | == My Robots == |
| | | | |
| − | 8Dez09 - v0.6 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.6 dmsr.MiniR101 0.6] [http://m.sixhat.net/robocode/dmsr.MiniR101_0.6_src.zip src]
| + | [[MiniR101]] - MiniR101 is my first Bot at Robocode. It started by being a Bot for experiments and learning. When I first packed it the size was near the [[MiniBot]] category, so I decided to keep it in that category. At the moment this is only a MiniMelee bot. It does very poor in 1o1 and this is probably the next thing I'll try to improve. |
| − | | |
| − | In this version I just did basically two things that were missing.
| |
| − | | |
| − | * One was to update the fitness function to take into account the danger of each enemy. I got this idea when looking to [[HawkOnFire]] code.
| |
| − | * The second one was to fix the open source code problem. When I package the Bot it isn't packaging the source code (for what ever reason that I don't want to go into now) so I decided to put the source code in my own server and you can download it as a zip file.
| |
| − | | |
| − | v0.5 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.5 dmsr.MiniR101 0.5]
| |
| − | | |
| − | Small changes make nice improvements...
| |
| − | | |
| − | v0.4 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.4 dmsr.MiniR101 0.4]
| |
| − | | |
| − | This one was an attempt to bring 0.3 to a decent position. Although 0.3 did perform better than 0.2 in terms of Average Score, it was failing in Survival... I didn't had the time to check what was happening but I started making some adjustments that ended in 0.4. I happier with it now. I wont touch it until 2010 so...
| |
| − | | |
| − | * Movement: Uses a fitness function to determine which placement is best to move to next (So it my take on [[Minimum Risk Movement]])... Considers enemies position and geographical position.
| |
| − | | |
| − | * Targeting: tracks enemies positions and tries to determine Deltas to estimate future position and then fires a bit randomly into that direction (I'll have to check this in the future)
| |
| − | | |
| − | * Radar: Spins, spins, and when it finds a target it tries to invert direction, so it momentarily locks
| |
| − | | |
| − | v0.3 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.3 dmsr.MiniR101 0.3]
| |
| − | | |
| − | This one is a mess... It was coded during a weekend from v0.2 and I changed so many things that I ended up with almost 3000 Bytes of code... I then had a huge battle to make him fit in the Mini category! The big difference is that I changed from a Random Firing Gun to a Random Firing Linear Targeting Gun (don't ask as I'll have to read the code to understand what is going on there).
| |
| − | | |
| − | Next: Take a deep look at this code before trying anything fancy, and see what small improvements can be made to recover the % survival that dropped 20% with this version...
| |
| − | | |
| − | v0.2 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.2 dmsr.MiniR101 0.2]
| |
| − | | |
| − | Basically it's v0.1 with parameters optimized. This greatly improved the survival of the damn thing... (or at least made it dye a little bit latter)
| |
| − | | |
| − | v0.1 [http://darkcanuck.net/rumble/RatingsDetails?game=minimeleerumble&name=dmsr.MiniR101%200.1 dmsr.MiniR101 0.1]
| |
| − | | |
| − | First alfa release of a Robot with some ideas I was playing around. As I was coding I was first playing with the movement algorithm and was happy with it. Then I just added a simple Radar and Targeting ideas and packaged it. I saw it was 1599 Bytes so it was only 99 Bytes above the Mini category. I tweaked it a bit to make it fit in the Mini category and put it in the Melee Rumble just to see how it behaves there. I didn't do any parameter optimization or nothing... so what ever he does is what he is capable of just after birth...
| |
| | | | |
| | == Other Info About Me == | | == Other Info About Me == |
| Line 42: |
Line 9: |
| | | | |
| | My [http://complexsystemsstudies.eu/ PhD] | | My [http://complexsystemsstudies.eu/ PhD] |
| | + | |
| | + | My [http://twitter.com/sixhat Twitter] |
| | | | |
| | [[Category:Bot Authors|DavidR]] | | [[Category:Bot Authors|DavidR]] |
I'm DavidR from Portugal and I'm a new user at Robocode and I'm doing research in Swarm intelligence, Synchronization and Networks, Robocode is a very interesting way to develop some of the ideas I have. I'm coding my first robot at the moment and hope to have it ready by the end of the year (If I manage to get the time to code as my PhD is taking 99% of my time now.)
My Robots
MiniR101 - MiniR101 is my first Bot at Robocode. It started by being a Bot for experiments and learning. When I first packed it the size was near the MiniBot category, so I decided to keep it in that category. At the moment this is only a MiniMelee bot. It does very poor in 1o1 and this is probably the next thing I'll try to improve.
Other Info About Me
My Blog: Sixhat Pirate Parts
My PhD
My Twitter
