[Wien] PORT, atomic position relaxation

[Stefaan Cottenier]

> > Does that mean i wont be able to relax the "y" coordinate
> > in any  of these cases.
> 
> Indeed.

Forgot to add: if you really want to relax also this y=1/4 coordinate, 
then do the following:

* Change it to 0.250001 (consistently, if you have a corresponding 3/4 
position for one of the atoms at that position, then change it to 
0.749999).

* Repeat init_lapw. sgroup and nn will reduce the symmetry of your 
cell, and you will end up with more k-points.

* Converge one scf-cycle, and see what is the force for that y-
coordinate (you will have a force now, because in the lower space 
group this position will not be fixed by symmetry any more). But very 
often, you will see that the corresponding force is zero (within the 
numerical noise).

Stefaan