Determining the density of states for classical statistical models: a random walk algorithm to produce a flat...

Area of Science:

• Statistical Mechanics
• Computational Physics
• Materials Science

Background:

• Accurate estimation of the density of states is crucial for understanding classical statistical models.
• Conventional Monte Carlo simulations face challenges in complex systems and at phase transitions.
• Calculating thermodynamic quantities like free energy and entropy can be computationally intensive.

Purpose of the Study:

• To develop an efficient Monte Carlo algorithm for accurate density of states estimation.
• To enable precise calculation of thermodynamic properties, including free energy and entropy.
• To overcome limitations of conventional methods in complex systems and at phase transitions.

Main Methods:

• A novel random walk algorithm in energy space is employed.
• The density of states is adaptively modified to achieve a flat histogram.
• The method is applied to Potts, Ising, and spin-glass models, including multi-dimensional walks.

Main Results:

• Highly accurate density of states estimates were obtained for large 2D Potts and Ising models.
• The algorithm successfully estimated thermodynamic quantities (internal energy, specific heat, entropy) at various temperatures.
• The method demonstrated efficiency in overcoming tunneling barriers at first-order phase transitions and characterizing spin-glass models.

Conclusions:

• The proposed Monte Carlo algorithm provides an efficient and accurate approach for studying classical statistical models.
• This method significantly enhances the calculation of thermodynamic properties and is applicable to complex systems.
• The random walk technique is versatile and can be extended to calculate the density of states for various parameters beyond energy.