Is it possible to write the Hamiltonian into wannier90 format?

[Dongsheng-Wen]

Dear Developer,
Thanks for the instructive codes and I find it very helpful. I wonder if it is possible to convert the model Hamiltonian into the wannier90.dat format, which is quite widely used in calculations by wannierberri (https://github.com/wannier-berri/wannier-berri) and wanniertools (https://github.com/quanshengwu/wannier_tools)?

Regards,
Dongsheng

[BertJorissen]

Hi
I'm working on that, but some files for the wannier90 input/output related to the symmetries seem to be badly defined in the manual.
In the development version, 'pip install pybinding-dev', I'm experimenting with the Wannier class (though, I don't think it was working already).
I'll give it a try in a few weeks again.
Best
Bert

[BertJorissen]

By the way, the berry phase can be calculated in the development version with the right units etc. Give it a try with the 'pybinding-dev' release, a small example here: https://bertjorissen.github.io/pybinding/tutorial/wavefunction.html . Please let me know if you used this code, as this is already an extension to pybinding. The formalism is the same as the one from pythtb, using the idea from the king-smith paper.

[Dongsheng-Wen]

Hi I'm working on that, but some files for the wannier90 input/output related to the symmetries seem to be badly defined in the manual. In the development version, 'pip install pybinding-dev', I'm experimenting with the Wannier class (though, I don't think it was working already). I'll give it a try in a few weeks again. Best Bert

Thank you. Really looking forward to this feature. For the simplest hamiltonian format, wannier90 wrote it this way: https://github.com/wannier-developers/wannier90/blob/2f4aed6a35ab7e8b38dbe196aa4925ab3e9deb1b/src/hamiltonian.F90#L698-L799. For other matrices like spin and position matrices, I'm not quite sure.