Spin polarized calculation using initial spin non collinear

[hemant arora]

Hello,

I am using spin-polarized GGA calculation and setting the initial spin noncollinear. will this noncollinear initial spin affect the output?
Let's assume In my calculation I have set spin-polarized GGA FHI pseudopotential. The noncollinear initial spin is first put in the x-direction and then all in the z-direction. Will the output characteristic (Mulliken population or transmission spectra) be the same or different?

[Anders Blom]

Not necessarily, it's after all just a starting guess for the relative populations. Of course there is always a risk you precondition it towards a local minimum, but that chance is always there no matter how you select the initial guess.