Histogram Reweighting Method for Dynamic Properties

TL;DR

This paper extends the histogram reweighting technique to dynamic properties from Molecular Dynamics simulations, enabling efficient calculation of correlation functions and transport coefficients across different thermodynamic conditions.

Contribution

It demonstrates that histogram reweighting can be applied to dynamic properties, allowing reconstruction of initial state distributions without additional simulations.

Findings

Correlation functions can be obtained from few simulation datasets.

Transport coefficients can be accurately estimated across conditions.

Method reduces computational effort for dynamic property analysis.

Abstract

The histogram reweighting technique, widely used to analyze Monte Carlo data, is shown to be applicable to dynamic properties obtained from Molecular Dynamics simulations. The theory presented here is based on the fact that the correlation functions in systems in thermodynamic equilibrium are averages over initial conditions of functions of the trajectory of the system in phase-space, the latter depending on the volume, the total number of particles and the classical Hamiltonian. Thus, the well-known histogram reweighting method can almost straightforwardly be applied to reconstruct the probability distribution of initial states at different thermodynamic conditions, without extra computational effort. Correlation functions and transport coefficients are obtained with this method from few simulation data sets.