Annotated GW output
This page explains the output from 1-shot GW and QSGW calculations. The lmgw1-shot script is used for the 1-shot GW calculation and the lmgwsc script is used for the QSGW calculation. The output from both lmgw1-shot and lmgwsc are similar. The main difference is the construction of a full self-energy (not just the diagonal terms). This is converted into an effective exchange-correlation potential in the lqpe step.
Consider the following output from lmgw1-shot broken down into individual lines:
The above line sets up a call to the GW package to extract the k-point list.
Make k-point list
Generate input data for GW at given k-points (eigenvectors, projection of evecs…)
Translates output of lmfgwd into intermediate form.
Translates intermdiate form of lmfgwd output into form used by GW. Also makes information about matrix elements.
Finds Fermi level.
Finds energy range.
Product basis for core levels.
Hartree matrix elements for core
Exchange matrix elements for core
Product basis for valence
Hartree matrix elements for valence
Exchange matrix elements for valence
Dielectric function and screened Coulomb interaction
Self-energy
Assemble components of potential and make the QPU file.