different VASP5.3.3 version, different RPA results
Posted: Tue Feb 04, 2014 6:21 pm
I have to VASP5.3.3 version compiled on two different machines, at different times of the year. They both work fine and give me the exact same results for "normal" calculations of DFT energies.
However, when I run an RPA calculation, the 3rd and the 4th steps ( the "determination of all virtual states by an exact diagonalisation of the DFT Hamiltoninan" and the calculation of the ACFDT-RPA correlation, Ec) report different kinds of values. I've noticed that in the fourth step, the version i think is the latest one, gives "total correlation" instead of "RPA correlation" in the OUTCAR. Does this mean that I have to subtract the RPA correlation from the total? If so, how?
I am using the exact same files (INCAR, POTCAR, POSCAR and KPOINTS).
Any help would be much appreciated.
Best,
N.
However, when I run an RPA calculation, the 3rd and the 4th steps ( the "determination of all virtual states by an exact diagonalisation of the DFT Hamiltoninan" and the calculation of the ACFDT-RPA correlation, Ec) report different kinds of values. I've noticed that in the fourth step, the version i think is the latest one, gives "total correlation" instead of "RPA correlation" in the OUTCAR. Does this mean that I have to subtract the RPA correlation from the total? If so, how?
I am using the exact same files (INCAR, POTCAR, POSCAR and KPOINTS).
Any help would be much appreciated.
Best,
N.