Wang-Landau sampling with self-adaptive range

Area of Science:

• Computational physics
• Statistical mechanics
• Numerical methods

Background:

• The Wang-Landau algorithm is a powerful Monte Carlo method for calculating the density of states.
• Complex systems often exhibit intricate density of states, posing challenges for traditional algorithms.
• High-precision numerical integration of sharply peaked functions in multidimensional spaces is computationally demanding.

Purpose of the Study:

• To introduce a self-adapting variant of the Wang-Landau algorithm.
• To demonstrate its suitability for systems with complex density of states structures.
• To showcase its application in high-precision numerical integration tasks.

Main Methods:

• Development of a self-adapting mechanism within the Wang-Landau algorithm.
• Application to determine two-dimensional densities of states.
• Utilizing the algorithm for numerical integration of sharply peaked functions over multidimensional domains.

Main Results:

• The self-adapting Wang-Landau algorithm efficiently handles systems with complicated density of states.
• Accurate determination of two-dimensional densities of states was achieved.
• High-precision numerical integration of sharply peaked functions was successfully performed.

Conclusions:

• The self-adapting Wang-Landau algorithm offers a robust and efficient solution for complex computational problems.
• This enhanced algorithm expands the applicability of Monte Carlo methods in physics and numerical analysis.
• It provides a valuable tool for researchers dealing with intricate density of states and multidimensional integration.