Macroscopic dielectric matrix from GW0

[ssharifz]

Hello,

I would like to know information about the calculated macroscopic dielectric matrix from GW0. A lot of information is printed out in the OUTCAR, but it is not clear what the information means. There are sets of numbers printed after the heading:

"INVERSE MACROSCOPIC DIELECTRIC TENSOR (including local field effects in RPA (Hartree))".

Can anybody tell me how I can construct the dielectric matrix based on this output? Thanks!

[ssharifz]

I should also add that I am using the tag NWRITE = 3, and so there are many numbers printed after the dielectric constant. They are under the headings "real part", "imag part", "real part wings", etc. I would like to know what these numbers mean.

[giacomo giorgi]

Hello,
Same request....
Thanks in advance to anybody able to give info about the matter,

G

[panda]

http://cms.mpi.univie.ac.at/vasp/vasp/node174.html

Details are explained here with complete references

The Inverse Macroscopic Dielectric Tensor is a Tensor. A tensor is a vector representation independent of coordinate system. To construct the dielectric matrix you would first need to parse the OUTCAR file and then store the Inverse Macroscopic Dielectric Tensor as a tuple. Remember that a tensor in matrix form is rank 2 (matrixes are tensors of rank 2, vectors are tensors of rank 1, scalars are tensors of rank 0). So to construct the dielectric matrix ALL you need to do is parse the OUTCAR and store the values as a tuple. If you still need help I will do this for you, because I have to do it anyway, it just will not be immediately.

[panda]

OK, so now I have written a parser to obtain the Macroscopic Dielectric Tensor, which I have stored as a tuple. But what to do with it? Ideas? Plot it? Post-data processing? Anyone? Help!!!!!

[monazam]

Dear Panda

Where do you have this parser? is there any way to find dielectric constant from inverse macroscopic dielectric tensor?
Thanks very much in advance.