[QE-users] Issue using "b3lypv1r" keyword

[Zack Windom]

Hi all, 

I am going through the process of testing and comparing Espresso’s version of B3LYP with other GTO-based software. I think I may have stumbled across a bug in trying to use “b3lypv1r”: notably, the output file reads:

Exchange-correlation      = B3LYPV1R ( 0  3  0  0 0 0)


Additionally, when I list the full name as "b3lp+b3lpv1r+b3lp+b3lp”, the output file reads:

Exchange-correlation      = B3LP+B3LPV1R+B3LP+B3LP ( 7  0  9  7 0 0)

Things change for the better when I use the short name “b3lyp-v1r”, as the output reads correctly:

Exchange-correlation      = B3LYP-V1R ( 7 13  9  7 0 0)


As I am trying to modify the source code to include a new functional based off B3LYP, this technicality is rather important. It seems that even using "b3lp-v1r” as the correlation in the full name, Espresso still interprets this to mean no correlation, or icorr=0. This seems like a parsing error somewhere; I was hoping someone could provide a patch, or give general guidance on files that I might need to look at to fix this. 


Thanks,
Zachary Windom



Graduate Student, Chemical Physics
Bartlett Group, Quantum Theory Project
University of Florida, Gainesville, FL