My Community

I am trying to calculate the projections of the wave function onto the spherical harmonics with the spin orbit interactions on.
What are the formats of the PROOUT.1-4 files for spin orbit calculations?
Why are some of these files showing lots of NaN?

Hello!

Please provide more information about your problem. Which files are showing NaNs and where?

Did you have a look at the Wiki entry for the files: wiki/index.php/PROOUT ?

If you cannot find the answer there can you please upload the necessary files according to the forum guidelines?

Thank you!

Best,

Andreas Singraber

Thank you for your reply.

I understood the format of PROOUT file when there is no SOC as only one PROOUT file ( PROOUT.1) is generated. With SOC, four PROOUT files are being generated ( PROOUT.1, PROOUT.2, PROOUT.3 and PROOUT.4). My question is: what do these four files correspond to?

In PROOUT.3 and PROOUT.4 files, there are lots of NaN.

Also, is there any documentation on how exactly the augmentation part in the PROOUT file is calculated?

Sorry for my delay in answering this question.
Indeed there seems to be a serious problem with the PROOUT file for noncollinear calculations.
The quantities that are written refer to the WF states so in the case of collinear (ISPIN=2) or non-collinear calculations (LNONCOLLINEAR) two files should be produced: one for each spin channel.
The handling of non-collinear calculations is not implemented correctly (I added a warning about it on the wiki page)
Thanks for reporting the bug, I will work on fixing it for the next release.

In general, the support for the PROOUT file is not great so it is preferable if you get the quantity that you need from another file.
One option would be to look at the LOCPROJ file:
wiki/index.php/LOCPROJ

Another option is to use the wannier90 interface and read the projections from the wannier90.amn file.
wiki/index.php/LWANNIER90

It depends on what exact quantity you need.