Hands-On Neuroevolution with Python

MultiNEAT usage example

The following is an example of using the MultiNEAT library to implement the XOR solver using neuroevolution. This is just an overview, without the implementation of the XOR fitness score evaluator (evaluate_xor), which will be discussed in the next chapter. Let's get started:

1. Create the NEAT configuration settings:

params = NEAT.Parameters()params.PopulationSize = 100# The rest of the settings omitted for brevity

2. Create a minimal genome configuration and spawn a population of organisms from this genome:

g = NEAT.Genome(0, 3, 0, 1, False,       NEAT.ActivationFunction.UNSIGNED_SIGMOID,      NEAT.ActivationFunction.UNSIGNED_SIGMOID, 0, params, 0)pop = NEAT.Population(g, params, True, 1.0, i)

3. Run the evolution process over ...