Quickstart

This guide will explain how to install and configure a genetic algorithm using the library.

1. Install the package

The package can be installed with pip and git:

pip install git+https://github.com/AdrianvEik/KTH_dfm_control#egg=dfmcontrol

2. Run a simple optimization algorithm.

To show the basic functionality of the library, we will optimize a simple function defined here as tfx(). The function is defined as: \(f(x) = (3 x^2 + 2x + 1) \cdot \sin{(x)}\). With its minimum at \(x = 0\).

2.1 Import the necessary modules

import numpy as np
from dfmcontrol import GeneticAlgorithm
from dfmcontrol.test_functions import tfx

2.2 Initialize the genetic algorithm

The following code initializes the genetic algorithm with a bitsize of 16. The bitsize is the number of bits used to represent the individual in the population. This binary value is then converted to a numerical value using the defined b2n with a function from Helper, ndbit2int(). The attribute b2nkwargs defines the keyword arguments for the conversion function, setting the lower and upper bound of the numerical value.

# We use a bit size of 16 to represent the values in the range [-10, 10]
ga = genetic_algorithm(bitsize=16)

# Initialize the conversion function for binary to numerical values
ga.b2n = dfmh.ndbit2int
ga.b2nkwargs = {"factor": 10} # This defines the search space [-10, 10]

Note

The conversion function can be any function that takes a binary array and returns a numerical value. The function must be defined in the module dfmcontrol.helpers. This function will also define the population initailisation that is available in the module.

2.3 Initialise the population.

The population is initialised using the init_pop() method of the genetic algorithm. Which as the following arguments:

• Method: The method used to initialise the population. This can be either uniform_bit_pop or cauchy_bit_pop or bitpop from .