Implicit basis for SAXIS

[rolando_saniz1]

When studying non-collinear magnetic order one can specify the spin quantization axis, i.e., set SAXIS = s_x s_y s_z. We are assuming that s_x, s_y, and s_z here are coordinates in the basis of lattice vectors, and not cartesian coordinates. Could you please comment on this? (i.e., whether it is correct or not). In cubic systems, it doesn't really make a difference, but for trigonal or monoclinic systems, for example, the difference is important. Actually, the same question applies to MAGMOM.

[henrique_miranda]

I am not sure I understand your question correctly but I will try to answer nonetheless.
SAXIS defines a basis of vectors with respect to the cartesian coordinates.
The transformation and inverse transformation are documented on this page:
wiki/index.php/SAXIS

Let me know if this answers your question.

[henrique_miranda]

I updated the wiki regarding this issue to make it clearer that SAXIS and MAGMOM are defined with respect to cartesian coordinates and so independent of the lattice vectors.