[Octopus-devel] [Octopus-notify] svn commit: r4981 - in trunk/src: system by dstrubbe

[Xavier Andrade]

Hi David,

On Fri, 20 Feb 2009, octopus-notify at tddft.org wrote:

> *) The application of perturbations involved first applying phases
>    e^ikr.  In fact, this is incorrect because when pseudopotential
>    projectors are used, they apply these phases themselves.  The phases
>    are no longer applied in the perturbations.

The projectors only apply the difference of the phase, not the whole 
phase, if you have a look at projector.F90, lines 201-203 you see it:

   kr = sum(kpoints(1:ndim, ik) * &
     (this%sphere%x(is, 1:ndim) - this%sphere%mesh%x(this%sphere%jxyz(is), 1:ndim)))

this is done in this way because when applying the hamiltonian you already 
have the phase in the wave-function.

X