what do "n_spin" and "n_spinor" mean in module electrons

[hongtang2023]

Dear Yambo team,

I have a simple question about Yambo code.
On one calculation, I used Quantum ESPRESSO to do a SOC and noncollinear calculation, in which I set "nspin=4" in this QE calculation.
However, when I use Yambo to initialize from the QE results, in the r_setup file, I got
" Spinor components : 2
Spin polarizations : 1 "

Spinor = 2 is understandable. but why " Spin polarizations : 1 " ?

Moreover, when I use TDDFT ALDA to do calculations, I got "n_spin=2". The "n_spin" is defined in module electrons and was used in subroutine TDDFT_ALDA_G_space(E,k).

So why I got "n_spin=2" in subroutine TDDFT_ALDA_G_space(E,k)? and why I got "Spin polarizations : 1" in the r_setup file?

Another question, there is "n_spinor" also defined in module electrons. Is "n_spinor" the same as " Spinor components : 2" in the r_setup file?


Appreciate your help very much!


Best,
Hong

[Davide Sangalli]

Dear Hong,
in yambo the WFs and the energies are stored with the following indexes.

Code: Select all

psi_{ib,ik,i_sp_pol}(ir/ig,i_spinor)
epsilon_{ib,ik,i_sp_pol}

Here
- ir/ig are the index in the real/reciprocal space grid, and i_spinor=1,n_spinor runs over the spinorial component of the WFs
- ib,ik, and i_sp_pol=1,n_sp_pol run over the possible quantum numbers which define the eigenvalues.

n_spin=max(n_sp_pol,n_spinor)

Possible values are
- n_spin=n_sp_pol=n_spinor=1 (case without spin)
- n_spin=n_sp_pol=2 ; n_spinor=1 (magnetic system without SOC, Sz is a good quantum number)
- n_spin=n_spinor=2 ; n_sp_pol=1 (systems with SOC, Sz is not a good quantum number)


Best,
D.

[hongtang2023]

Thanks! Davide,

in
epsilon_{ik,ik,i_sp_pol}

the first ik should be ib, right?


Best,
Hong

[Davide Sangalli]

Yeah. I edited the reply.