Energy Estimators and Calculation of Energy Expectation Values in the Path Integral Formalism

TL;DR

This paper extends a path integral method to accurately compute energy expectation values, demonstrating improved convergence and significant computational speedups through analytical energy estimators and Monte Carlo simulations.

Contribution

It introduces analytical energy estimators within the hierarchy of effective actions, enhancing the convergence of energy calculations in path integral formalism.

Findings

Energy expectation values converge as 1/N^p

Significant speedup in numerical calculations

Validated by Monte Carlo simulations

Abstract

A recently developed method, introduced in Phys. Rev. Lett. 94 (2005) 180403, Phys. Rev. B 72 (2005) 064302, Phys. Lett. A 344 (2005) 84, systematically improved the convergence of generic path integrals for transition amplitudes. This was achieved by analytically constructing a hierarchy of $N$-fold discretized effective actions $S^{(p)}_N$ labeled by a whole number $p$ and starting at $p=1$ from the naively discretized action in the mid-point prescription. The derivation guaranteed that the level $p$ effective actions lead to discretized transition amplitudes differing from the continuum limit by a term of order $1/N^p$. Here we extend the applicability of the above method to the calculation of energy expectation values. This is done by constructing analytical expressions for energy estimators of a general theory for each level $p$. As a result of this energy expectation values converge to the continuum as $1/N^p$. Finally, we perform a series of Monte Carlo simulations of several models, show explicitly the derived increase in convergence, and the ensuing speedup in numerical calculation of energy expectation values of many orders of magnitude.