My Community

Dear Fabien：
Thanks for your reply！I have got that with your help.
Best wishes

Dear Fabien：
I have a further question. How can I ensure that I consider the U value when calculating GW?
According to your suggestion, I add the U value to the INCAR of EXACT. According to my understanding, when calculating GW, the Green's function is iterating, and W ，the dielectric shielding matrix should be constructed by reading the WAVECAR and WAVEDER in act. I want to confirm that this part of U value is sure to be passed to GW. Is the calculation by building W(shielding matrix) ，where can I find this relevant code line in gw _ module. F and LDApU .F ， I don't seem to find the relevant code.
Please give me some help!
Best wishes!

Hi,

For steps 1 and 2 (DFT calculation) the functional is specified in INCAR. For step 3 (GW), no functional should be specified in INCAR, and the physical effect of the functional used in steps 1 and 2 is conveyed via the orbitals and eigenvalues stored in WAVECAR and WAVEDER.

Concerning iterating the eigenvalues, Eq. (3) of http://dx.doi.org/10.1103/PhysRevB.75.235102 is used and does not require the evaluation of the DFT functional.

Dear fabien_tran1：
Thanks for your help again！
Best wishes