Diamond

From Robowiki
Revision as of 17:22, 31 May 2012 by Voidious (talk | contribs)
Jump to navigation Jump to search
Diamond
Diamond collage thumb200.png
Full Size
Author(s) Voidious
Extends AdvancedRobot
Targeting Dynamic Clustering, Displacement Vectors, GuessFactors
Movement Minimum Risk Movement, DC/GF-Wave Surfing
Current Version 1.5.30
Code License zlib
Download
Sub-pages:
Version History - DiamondHawk
Youtube
Youtube has a video of Diamond in action: click here to watch

Background Information

What's special about it?
Oh, nothing, really. Well, since you asked...
How competitive is it?
Very. =) #2 in Melee right behind DemonicRage, #2 in 1v1 behind DrussGT.

Strategy

How does it move?
In melee, it uses a Minimum Risk Movement, evaluating a bunch of points around itself and choosing the best (least bad / most good) one. I'm always tweaking the parameters, but some of them are:
  • Considers points in a (randomized radius) circle around itself.
  • Stay far away from other bots, weighted by their energy.
  • Stay perpendicular to other bots.
  • Don't be the closest bot to any other bots (avoid being targeted).
  • Some randomizing based on past locations.
In 1v1, it uses Wave Surfing (derived from Dookious) with Dynamic Clustering.
How does it fire?
The gun uses Dynamic Clustering for data analysis - i.e., k-nearest neighbors to find similar states, and kernel density among those states to settle on a firing angle. In Melee, it uses displacement vectors for recording and reconstructing firing angles - I thought that made more sense than GuessFactors, and it's a lot simpler and more efficient than real Play It Forward. In 1v1, it uses GuessFactors with precise MEA. It uses precise intersection and interpolates missed scans when necessary.
I'm also experimenting with various clustering and optimization techniques for the 1v1 guns, with the help of WaveSim.
How does it select a target to attack/avoid in melee?
As of 1.30, it no longer does. It aims at everyone on the battle field at once and tries to hit as many enemies as possible. (Big thanks to Shadow/Melee Gun for the excellent idea.)
How does the melee strategy differ from one-on-one strategy?
In 1v1, it switches to a Wave Surfing movement and switches to its 1v1 Virtual Guns.
What does it save between rounds and matches?
Nothing between matches. Between rounds, it saves all its targeting and movement data, which includes general info about each enemy, a bunch of kd-trees, and maps of the trees' data points to displacement vectors and/or GuessFactors.

Additional Information

Where did you get the name?
Diamond is the ex-superhero name of the main character in Powers, my favorite comic book series. I like to name all my bots after warriors of some sort, but I'd be lying if I said I didn't just think it sounds cool.
Can I use your code?
Yes, it's released under the zlib license.
What's next for your robot?
I'd love to take the 1v1 throne. A lot. =) And it will take some work to reclaim the Melee throne, especially with Demon coming.
Does it have any White Whales?
DrussGT
What other robot(s) is it based on?
I tried to start with a clean slate on this one. But I've taken plenty of ideas from the wiki and utility code from a few places.
  • BrokenSword - General guidance on various things, especially the minimum risk movement and bullet power management.
  • Dookious - Diamond's surfing algorithm was originally ported from Dookious, along with a variety of utility functions, including precise prediction code based on Albert's FuturePosition.
  • Lukious - The tiniest bit of kernel density code and general guidance on scan distancing.
  • HawkOnFire - I never did peek at the code, but it deserves a nod just for the wisdom rozu shared about minimum risk movement via this bot.
  • Corbos, Simonton - Corbos for the idea to use a kd-tree to speed up our DC guns, Simonton for the bucket variant.
  • Shadow - Pioneered the Shadow/Melee Gun, which is an amazing advancement in Melee targeting.
  • Rednaxela - Credit for precise intersection, Gun Heat Waves, and a lightning fast kd-tree.