bwfn from pw-scf

General discussion of the Cambridge quantum Monte Carlo code CASINO; how to install and setup; how to use it; what it does; applications.
Post Reply
Katharina Doblhoff
Posts: 84
Joined: Tue Jun 17, 2014 6:50 am

bwfn from pw-scf

Post by Katharina Doblhoff »

I am generating bwnf.data.b1 files directly from pw-scf (using pw.x -pw2casino and having a pw2casino.dat file). And have several questions concerning this:
1.) I need to make sure my blip conversion is converged and would thus like to perform kinetic energy tests. Is this possible or can I only perform overlap tests? Since I am no experienced user, the overlaps are hard for me to interpret - errors in the energy would give me physical numbers and I know which energy threshold I am aiming at.

2.) In the out file I find
Energies determined by pw2casino tool
-------------------------------------
Kinetic energy 50.4668872148199 au = 100.933774429640 Ry

Is this the kinetic energy after conversion to blip? If so, does pwscf somewhere give the kinetic energy before the conversion?
If, on the other hand, this is the kinetic energy before the conversion, am I then correct in assuming that a good check for the blip convergence would be to calculate the vmc kinetic energy (using no jastrow) and comparing the two, or is there a more intelligent method (e.g. using the blip utility and doing a kinetic energy check)?

3.) with the blip utility one can shift the atoms for reduced dimensional calculations. Can this be done directly from pwscf, or do I need to input the geometry accordingly?
Katharina Doblhoff
Posts: 84
Joined: Tue Jun 17, 2014 6:50 am

Re: bwfn from pw-scf

Post by Katharina Doblhoff »

I have figured some of the answers out myself:
ad 1.) performing the numerical integral using the blip facility is numerically prohibitive for larger systems at the moment.
ad 2.) these energies thus have to be pwfn kinetic energies and can therefore be compared with the vmc kinetic enegies for periodic systems with no Jastrow included.
Post Reply