How to get PROJ_x.OUT file under specified high symmetry points in exciting?

exciting
1

Dear Developers,
Thank you for taking time to answer my questions.
I employ the element inside , and get output files with the name ‘PROJ_x.OUT’. Also, I could get the projection of each evolved KS wavefunction onto the groundstate in uniform k-point with the attribute ‘ngridk’ inside . But how can I get the projection under specified high symmetry points?

<xs nempty="5" ngridk="10 10 1" nosym="true" reducek="false" xstype="RT-TDDFT">
        <realTimeTDDFT 
                endTime="405.d0"
                printAfterIterations="10" 
                printTimingGeneral="true" 
                propagator="AETRS" 
                timeStep="0.5d0">
                <laser>
                        <sinSq 
                              amplitude="9.78d0" 
                              direction="x" 
                              omega="0.15d0" 
                              phase="0.d0" 
                              pulseLength="400.d0" 
                              t0="0.1d0"/>
                </laser>
                        <screenshots niter="100" printAbsProjCoeffs="true" />
        </realTimeTDDFT>
</xs>
2

The projection is carried out for the same k-points as given in the attribute “ngridk” inside the “xs” element.
Thus, the simplest way to obtain the projection under some desired high-symmetry point is to make sure that they are included in the k-grid generated from “ngridk” inside “xs”.

3

Thanks for your suggestion. It’s fine for computer to execute if the k-points, eg M:(0.0, 1/2, 0.0) and K:(-1/3, 2/3, 0.0), are simple. And I only need to set “ngridk=6 6 1 or 12 12 1” for that situation. But when i want to deal with the points along the high-symmetry, the “ngridk” must be very huge causing a high computational cost. Therefore, is there any other methods to solve it? Is interpolation feasible？

4

In my opinion, an interpolation can provide a reasonable approximation.
You can carry out your calculation for the available k-points and then interpolate your projections for the desired k-points.
I would recommend to check if the interpolated k-points are close to the ones employed for the interpolation.

1 Like