ERROR: k-point mismatch between WFN_fi and WFNq_fi
Eleanor Davison
Dear BGW developers,
I am trying to run a finite Q calculation for MoS_2, and I am running into an error in the absorption.x step. The error occurs when it tries to read the WFNq_fi file. Here is the output just before it stops running:
……
Shift vector : 0.00000 0.00000 0.00000 Length = 0.02185
Exciton momentum : 0.01500 0.00500 0.00000 Length = 0.02185
No symmetries used in the shifted-grid sampling
Could not find point equivalent to 1.388888900000000E-002
1.388888900000000E-002 0.000000000000000E+000
ERROR: k-point mismatch between WFN_fi and WFNq_fi
forrtl: error (76): Abort trap signal
Image PC Routine Line Source
absorption.cplx.x 0000000003BA352B Unknown Unknown Unknown
……
I have shifted the WFNq_fi grid by (-0.001 0 0) compared to WFN_fi (full lists of k-points attached). However, this output says that my “shift vector” is (0 0 0). I’m guessing this is why it can’t find the first point in WFNq_fi, as it is expecting it to be the same as the first point in WFN_fi, due to the shift vector between them being zero?
My questions are:
- Why has it found a zero shift between my fine grids? (or where does BGW extract the shift vector from?)
- Have I made a mistake using this q-shift? Should the shift between WFN_fi and WFNq_fi be equal to zero? Or maybe equal to the finite exciton momentum Q?
Thank you in advance for your help with this!
Best wishes,
Eleanor
Bill Ash
Have you solved this problem yet? I'm having the same problem.
Thank you in advance!
Bill