Wave Function Optimization in QMCPACK

ORAL

Abstract

Wave function optimization is essential for both the accuracy and efficiency of diffusion, reptation, and variational quantum Monte Carlo (QMC). In this talk we outline the wave function optimization strategy used in the QMC software package QMCPACK developed at the University of Illinois. We use an extension of the linear optimization method originally developed by Umrigar et. al.[1] to optimize parameters in Slater-Jastrow, multi-determinant Slater-Jastrow, and Backflow-Jastrow trial wave functions. The efficiency and accuracy of this method is presented for bulk Silicon, Jellium, and the Nitrogen dimer.\\[4pt] [1] Umrigar et al. PRL 98, 110201 (2007)

*This work was supported by the U.S. Department of Energy (DOE) under Contract No. DOE-DE-FG05-08OR23336 and by the National Science Foundation under No. 0904572.

–

Authors

  • Jeremy McMinis

    • University of Illinois and NCSA

  • Miguel A. Morales

    • LLNL

  • Jeongnim Kim

    • University of Illinois at Urbana-Champaign
    • University of Illinois and NCSA
    • UIUC

  • David Ceperley

    • Dept. of Physics, U. of Illinois at Urbana-Champaign
    • Department of Physics, University of Illinois Urbana-Champaign
    • University of Illinois and NCSA
    • University of Illinois at Urbana-Champaign
    • Department of Physics, University of Illinois, Urbana-Champaign, IL 61801