Page 1 of 1

GW energies on full BZ k-grid

PostPosted: Mon Mar 04, 2019 7:21 am
by chwolf
Dear all,

I am trying to obtain the GW eigenenergies on a k-point grid generated by wannier90's kmesh.pl tool (this one just generates a k-mesh in the full BZ with weights). The goal is to read these energies with the EPW code to calculate electron phonon coupling on GW corrected bands. The EPW manual says:

Code: Select all
Can be used to read GW (or other) eigenenergies. The code expect a file called "prefix.eig" to be read. One need to provide the same number of bands as in the nscf calculations and all k-points.


To that extend I tried to use ypp -k k and copied the full grid (6x6x6=216 k points)

Code: Select all
%GWKpts                      # K points list
  0.00000000 | 0.00000000 | 0.16666667 |
  0.00000000 | 0.00000000 | 0.33333333 |
  0.00000000 | 0.00000000 | 0.50000000 |
...
  0.83333333 | 0.83333333 | 0.50000000 |
  0.83333333 | 0.83333333 | 0.66666667 |
  0.83333333 | 0.83333333 | 0.83333333 |


I then remove all the points already in the old grid and run ypp; however that results in a long file with


Code: Select all
 <---> [06.02] == GW K-grid analyzer ==
 <---> User K-points : 10
 <---> Reducing & Expanding the 2176 k-points ...
 <---> BZ -> IBZ reduction |                                        | [000%] --(E) --(X)
 <---> BZ -> IBZ reduction |########################################| [100%] --(E) --(X)
 <01s> Reduced K-grid points: 243
       ---------- Reduced K-grid ----------
   0.0000000   0.0000000   0.0000000   0.0000974  1
   0.0000000   0.0000000   0.1666667   0.0007793
   0.0000000   0.0000000   0.3333334   0.0007793
   0.0000000   0.0000000  -0.5000001   0.0003896
   0.0000000   0.1666667   0.1666667   0.0005845
   0.0000000   0.1666667   0.3333334   0.0023378
  -0.0000000   0.1666667   0.5000000   0.0023378
   0.0000000   0.1666667  -0.3333334   0.0023378
  -0.0000000   0.1666667  -0.1666667   0.0011689
   0.0000000   0.3333334   0.3333334   0.0005845
  -0.0000001   0.3333333   0.4999999   0.0023378
  -0.0000000   0.3333333  -0.3333334   0.0011689
   0.0000000  -0.5000001  -0.5000001   0.0002922
   0.1666667   0.3333334   0.5000000   0.0023378
   0.1666667   0.3333333  -0.3333334   0.0023378
   0.1666667  -0.5000001  -0.3333334   0.0011689
************  -1.9960938  -2.0078125   0.0031171
************  -1.9882812  -2.1601562   0.0044808
************  -1.8242188  -1.8320312   0.0035067
************  -2.1523438  -2.0000000   0.0035067
************  -1.9960938  -2.0078125   0.0027275
************  -1.9960938  -1.8398438   0.0044808
************  -2.1562500  -2.1679688   0.0035067
************  -1.8320312  -2.0039062   0.0044808
************  -1.9882812  -2.0039062   0.0040912
************  -1.9921875  -2.3398438   0.0035067
************  -1.6523438  -1.6601562   0.0038964
************  -2.3242188  -2.0000000   0.0042860
************  -1.9843750  -2.0039062   0.0046756


Can someone tell me where my mistake lies?

TIA!
Chris

Re: GW energies on full BZ k-grid

PostPosted: Mon Mar 04, 2019 9:48 am
by Daniele Varsano
Dear Chris,
Hard to say what went wrong, you can post your input files as it is needed to reproduce the problem in order to spot it.
Anyway the k mesh generator it is meant to generate a new k point grid, having the same q grid to the original one containing the desired k points?
Is this what you need?
If you need just to expand your k point grid from the IBZ to the full BZ may be the easiest way is to remove the symmetries from your calculation.
Am I missing something?

Best,
Daniele