amaciarey/pathintegralgroundstate

This is a Path Integral Ground State Monte Carlo code written in Fortran. It is intended to the description of 4He in three dimensions but it can be easily extended to any other homogeneous or trapped system in any dimensions

pathintegralgroundstate's Issues

Revise the coefficients of Staging algorithm

Revise the staging transformation coefficients implemented in the different staging routines (for example line 612 of vpi_mod.f90 file). There is maybe a little bug.

Open MP parallelization

A first version of the Open MP parallelization has been implemented but, for now, there is an small problem concerning detailed balance. It must be carefully studied.

Harmonic confinement

The possibility of work with a trapped system must be implemented. It would be ideal to be able to work with harmonic confining potentials that can have different trapping frequencies in the different spatial dimensions.

Considerations about Bisection vs Staging

Concerning to the efficiency of the two different sampling methods one can see that bisection is clearly more efficient (about 50% faster than staging), however, the most computationally expensive step of the simulation are the center of mass movements.

When we are simulating an off-diagonal configuration to evaluate the OBDM it seems that staging performs much better than bisection. This last point must be studied carefully (work in progress).

Evaluation of the acceptance of off-diagonal movements

The acceptance of off-diagonal updates must be corrected. At the moment, the normalization factors suppose that there are Np off-diagonal particles which is not true, so this is a bug in the code that can be easily corrected.