format of PARCHG file for noncollinear calculations

[sophie_weber]

Hello,

I am currently doing calculations were I am calculating a band-decomposed charge density for bands in a window around the Fermi energy. I know that for collinear spin-polarized calculations (ISPIN=2) the first section of the PARCHG file outputs the charge density, and then the second section gives the magnetization density n(up)-n(down). The calculations I am performing include spin-orbit coupling, so I would expect the PARCHG to contain four sections, one for charge, and then mx, my, and mz components of magnetization density. However, the PARCHG with spin orbit coupling has only one entry. I initially assumed this was just charge density, but if I upload the file to VESTA for visualisation, it has both negative and postive components.

Could someone clarify what the format of the PARCHG with spin-orbit-coupling is, and if indeed it should just give charge density, why I might be getting positive and negative components? I attach the input files include the INCARs for both self-consistent and the decomposed charge density calculation, as well as the OUTCAR. Unfortunately, the PARCHG is too large to attach, but I can email it if necessary.

Thank you very much in advance.

[manuel_engel1]

Hi,

Unfortunately, LPARD=true is currently not supported for LNONCOLLINEAR=true. This is a known issue. I will make efforts that this is documented more clearly.

[sophie_weber]

Ahh, ok. That explains a lot. Thank you for clarifying, I appreciate it.