Wrong occupations in magnetic system

[sangallidavide]

There is an issue in the yambo code which wrongly determines the occupations in CrI3 with n_sp_pol=2.
E.g. we obtained occupations like
[X] Average metallic occ. [spin UP]: 0.999987

A workaround is to edit the following file
https://github.com/yambo-code/yambo/blob/master/src/common/OCCUPATIONS_Fermi.F#L121
increasing the tolerance

- nel_diff_zero=1.E-5
+ nel_diff_zero=1.E-4