A Python library for Cartesian genetic programming.
Run python -m unittest discover -v in the root folder.
Install the package using pip
pip install cgpy
The following code creates a random CGP and evaluates its function
op_table = [Operation("+"), Operation("*"), Operation("sin")]
dims = 2
nr_of_parameters = 0
nr_of_nodes = 5
cgp = create_random_cgp(dims, nr_of_parameters, op_table, nr_of_nodes)
pnt = [0.5, 1.5]
print(cgp.eval(pnt))
See more examples in the example folder and the documentation at www.cgpy.org
Cgpy is developed by Viktor Wase with contributions and assistance from John Brynte Turesson and Johannes Wennberg.