Hands-On Neuroevolution with Python

Exercise 2

In the source code for this chapter, we also provide the experiment runner implementation based on the MultiNEAT Python library that we introduced in Chapter 2, Python Libraries and Environment Setup.

You can try to use it to solve the hard maze task as follows:

1. Update the current Anaconda environment by installing the MultiNEAT Python library with the following command:

$ conda install -c conda-forge multineat

2. Run the experiment runner implementation based on the MultiNEAT library:

$ python maze_experiment_multineat.py -m hard -g 500 -t 10 --width 200 --height 200

These commands install the MultiNEAT library in the current Anaconda environment and start 10 trials (with 500 generations each) of the hard maze experiment using ...