A portable and flexible implementation of the Wang–Landau algorithm in order to determine the density of states

Published: 25 January 2022| Version 1 | DOI: 10.17632/gcx9k8wnhh.1


In this work we develop an implementation of the Wang–Landau algorithm Wang and Landau (2001) [4]arXiv e-prints. This algorithm allows us to find the density of states (DOS), a function that, for a given system, describes the proportion of states that have a certain energy. The implementation uses the Python and C++ languages for the algorithm itself, and it can take advantage of any library, such as the powerful LAMMPS library, for the computation of energy. Therefore, the resulting implementation is simple and flexible without sacrificing efficiency. This implementation also considers recent developments in the parallelization of the code for faster computation. We establish the soundness and effectiveness of our implementation by studying well-known systems such as the Ising model, the Lennard–Jones and EAM solids. We have found that our implementation can find the DOS with very good precision in a reasonable amount of time. Therefore, we are equipped with a very powerful and flexible implementation that can be easily used in order to study more realistic models of matter.



Computational Physics, Parallel Computing, Density of States