Comment on "Calculation of microcanonical entropy differences from configurational averages"

TL;DR

This paper presents an improved method for calculating equilibrium entropy differences in classical systems, enhancing accuracy especially with closely spaced energy levels or importance sampling Monte Carlo techniques.

Contribution

It introduces a simple modification to Davis's existing method, making it more effective for certain energy level spacings and sampling methods.

Findings

Improved accuracy over original method for closely spaced energy levels

Enhanced performance with importance sampling Monte Carlo

Provides guidelines for using the method with molecular dynamics simulations

Abstract

We introduce a simple improvement on the method to calculate equilibrium entropy differences between classical energy levels proposed by Davis (S. Davis, Phys. Rev. E, 050101, 2011). We demonstrate that the modification is superior to the original whenever the energy levels are sufficiently closely spaced or whenever the microcanonical averaging needed in the method is carried out by importance sampling Monte Carlo. We also point out the necessary adjustments if Davis's method (improved or not) is to be used with molecular dynamics simulations.