Atomic Global Minimum Locator: Genetic Algorithm

Genetic algorithms (GA) are an efficient means of searching for global minima. While my focus has been on developing 'new' algorithms such as particle swarm optimization, GA algorithms are very important and widely used. The genetic algorithm in this application is well designed, but it needs a new mating algorithm (see: the makeChild function in moleculeSet.cc). If you are interested in doing this, please contact me. If you want the most efficient algorithm, you might think about implementing a differential evolution algorithm. In the mean time, use simulated annealing or particle swarm optimization. Simulated annealing with a small population size, single point energy calculations, and the -i option is quite efficient. Afterwards it's a good idea to perform local geometry optimization on the list of best structures using an optimization file. This was my approach in reference 1.