aboratino / stick-creatures-climbers

Simulates evolution of a population of virtual stick creatures. They learn to climb using a genetic algorithm. Uses pygame for graphics

Home Page:https://www.youtube.com/watch?v=XqUoxJUXihs

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

stick-creatures-climbers

Simulates evolution of a population of virtual stick creatures.
They learn to climb using a genetic algorithm / evolutionary algorithm.

Example video on Youtube: https://www.youtube.com/watch?v=XqUoxJUXihs

Uses pygame for graphics.

Each stick creature consists of the following:

  • A given number of segments that independently swing back and forth.
  • A head, represented by a larger green dot.
  • A tail end, represented by a smaller green dot.
  • A white halo that appears around the head of the 'winner'.

In this scheme, 'breeding' occurs at static intervals. Whichever creature has come closest to the top of the screen is the winner, or stud. The winner then breeds with the rest of the population.

Breeding consists of averaging out the rotation and rotation rates of each segment of each creature with the stud. In addition there is a chance to mutate the rotation rate by a small amount.

Anthony Boratino 2013-1019

About

Simulates evolution of a population of virtual stick creatures. They learn to climb using a genetic algorithm. Uses pygame for graphics

https://www.youtube.com/watch?v=XqUoxJUXihs

License:MIT License


Languages

Language:Python 100.0%