/Genetic-Algorithm-Structure-Prediction

An implementation of a genetic algorithm in Python for predicting equilibrium crystal structures for a given potential. The potential implemented here is the Daoud-Cotton model, but this can be easily changed.

Primary LanguagePythonMIT LicenseMIT

Genetic-Algorithm-Structure-Prediction

An implementation of a genetic algorithm in Python for predicting equilibrium crystal structures for a given potential.

Features

  • Predicts ground-state (GS) structures in 2D for a given potential;
  • Change parameters in the program easily by acessing and editing the settings.py file;
  • Simple change of the interaction governing potential by selecting the desired one in the settings.py file (currently only has daoud-cotton model implemented);
  • Easy implementation of other potentials by adding their own files to the program folder.

Currently includes two versions of the script, one GA.py which only runs in serial and a paralellized version GA_Parallel.py which has some functions running the parellel to speed up the computations.

To do

  • Some more testing
  • Implement a potential for elastic multipole interactions
  • Do some clean up of the code
  • Attempt to paralellize more of the code to improve performance