Clarification on Born Effective Charges and Dielectric Constatns Output from G1W1 for a 2D Material

[Zimmi]

Dear Developers and Community,

I am working with G1W1 calculations for a 2D material and would like to extract dielectric-related quantities and Born effective charges from the outputs. In my results, I do not see any files with the extension `.eps`, and I am therefore unsure whether the dielectric response has been computed or stored elsewhere.

In particular, I noticed that in DFPT-related outputs (e.g. `tensors.xml`), there are two forms of effective charge tensors:
`EFFECTIVE_CHARGES_EU`, corresponding to the derivative of polarization with respect to atomic displacement (∂P/∂u), and
`EFFECTIVE_CHARGES_UE`, corresponding to the derivative of atomic forces with respect to an electric field (∂F/∂E), which are related by index transposition.

Could you please clarify:
1. Whether G1W1 produces any dielectric or effective-charge-related output (such as ε∞ or Born effective charges), and if so, in which files they are written.
2. Whether specific flags or options are required in the `gw.in` input file to enable the calculation or printing of dielectric quantities.
I have attached the `gw.in` input file used for my calculations for reference.

Thank you very much for your time and help.

[Daniele Varsano]

Dear Zimmi,

a GW calculation does not print the dielectric matrix even if it is calculated to evaluate the screened potential.
As you are emplying the plasmon-pole approximation, the dielectric matrix can be reconstructed by reading the ndb.pp* files e.g. using yambo-py.
You have a file for each q points eps_GG'(q,omega). For pp calculations, you have it calculated at zero frequency and at the imaginary PP frequency.
Note that what is stored is the quantity vX from wich you can build the inverse dielectric matrix adding a delta_gg'.

Also note that as you are considering a 2D material, the macroscopic component eps_00(q=0) is exactly 1.

Yambo does not calculate Born effective charges.

Best,

Daniele