Runs with fixed magnetization

[nthiliniek]

Hello,

Does Yambo support the QE runs with fixed magnetization? Or should I stick with the option " starting magnetization"?

Thanks.

[andrea.ferretti]

Dear nthiliniek,

QE implements fixed magnetization in spin-collinear calculations by using two fermi levels.
As far as I remember this option was not directly compatible with Yambo (which works only with one fermi level)...

One possibility, though, is to first converge with fixed mag and two fermi levels, and then check whether
one could go back to a description with 1 fermi level.

This could be done eg at the QE level by having a follow up scf run, starting with the existing spin densities, without the magnetic constraint.
If it stays with the desired magnetization you are done.
Within Yambo, if the system allows (eg you are working with a magnetic semiconductor), one could identify a unique fermi level even if QE has used two...
Then one needs to hack the datafile_schema.xml in QE $prefix.save to trick Yambo (this is a bit for advanced users, though).

short answer: not obvious to make fix mag working with Yambo (though not always impossible)
take care

Andrea