robert-forrest / evomatic

Evolutionary algorithm for development of glassy alloy materials

Evomatic

Evomatic is an genetic algorithm tool for the development of new alloys.

Installation

The evomatic package can be installed from pypi using pip:

pip install evomatic

Evomatic makes heavy use of the metallurgy package to manipulate and approximate properties of alloys. The cerebral package can be used by evomatic to obtain predictions of alloy properties on-the-fly during evolution.

Usage

Basic usage of evomatic involves setting some targets and running an evolution, the following example shows evolution of a small population towards alloys with high mass:

>>> import evomatic as evo

>>> evolver = evo.Evolver(targets={"maximise": ["mass"]}, population_size=50)

>>> history = evolver.evolve()

>>> history["alloys"]
                       alloy        mass  generation  rank   fitness
                       Og100  295.000000           7     0  1.000000
                 Og92.4Fl7.6  294.544000          11     1  0.998433
            Og91.4Lv7.4Cm1.2  294.276000           9     2  0.997512
          Og78Mc14Lv6.2Hs1.8  293.726000          11     3  0.995622
                       Lv100  293.000000           2     4  0.993127
                         ...         ...         ...   ...       ...
  Ar38.8Li28.5B24.9Sc5.7I2.1   25.396895          10   587  0.073521
                     Si79H21   22.398830          10   588  0.063218
                Li83.7Zr16.3   20.678292           6   589  0.057305
                 Li97.9Db2.1   12.464260           5   590  0.029078
                       He100    4.002602           4   591  0.000000

In this simple example, there is no better material for the objective of maximum mass than pure Oganesson.

Documentation

Documentation is available here.