Running Variational Monte Carlo
From Zori Wiki
To run VMC once you have a trial wave function and walkers:
zori -i vmc.xml -p psi.xml -r walkers
If you are using HDF5 walkers rather than XML walkers use the following command:
zori -i vmc.xml -p hno-psi.xml -H -r walkers.hdf5
Should you desire to continue a previous QMC run one can running using one of the Continue options:
-c Continues a run by discarding the last few steps so that the step number in the walker files matches that of the energy file. Random number generator is reinitialized to the seed in the walker file.
-C Does the same thing as the -c option except the random number generator restarts from the saved random generator state file (rng.bin).
example:
zori -i vmc.xml -p psi.xml -H -r walkers.hdf5 -c
Implemented VMC Algorithms:
Algorithm Type="M(RT)2" Metropolis, Rosenbluth, Rosenbluth, Teller and Teller (a.k.a Metropolis Monte Carlo) 'N. METROPOLIS, A. W. ROSENBLUTH, M. N. ROSENBLUTH, N. M. TELLER, and E. TELLER, J. Chem. Phys. 21, 1087 (1953).' (http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=JCPSA6000021000006001087000001&idtype=cvips&gifs=yes)
Algorithm Type="Langevin" Langevin accelerated Metropolis C. UMRIGAR, Phys. Rev. Lett. 71, 408 (1993).' (http://prola.aps.org/abstract/PRL/v71/i3/p408_1')
Algorithm Type="LangevinSingle" Langevin accelerated Metropolis with single electron moves. C. UMRIGAR, Phys. Rev. Lett. 71, 408 (1993).' (http://prola.aps.org/abstract/PRL/v71/i3/p408_1')
Further more other Accelerated variational Monte Carlo are possible by turning off branching in diffusion Monte Carlo algorithms (VMC=”TRUE” option).
Recommended algorithm: In the current implementation of Zori the Metropolis VMC algorithm seems to be more robust with a larger time step. The other Langevin algorthims as currently implemented in Zori seem to be less black box in their application.
Return to zori manual
