search

mir-group / phoebe

A high-performance framework for solving phonon and electron Boltzmann equations
https://mir-group.github.io/phoebe/
MIT License
83 stars 19 forks source link

Make the cutoff for determining WS cell vectors slightly less strict #165

Closed jcoulter12 closed 1 year ago

jcoulter12 commented 1 year ago

The cutoff for determining which vectors are in the WS cell (in the wsWeight function of the phononH0 class) was sometimes too strict, and gave zero weight to vectors which were actually in the WS cell. This loosens that cutoff slightly to avoid error messages to users and produce the correct behavior.

There is also some draft material for changing the behavior of the LOTO correction in the case of dimensionality=2 added in phononH0, based on what is done in rigid.f90 of QE. However, for now this is turned off with a boolean variable (longRange2d) at the start of phononH0. I don't have time to test this super thoroughly right at this moment, but I want to make sure the progress is saved.

jcoulter12 commented 1 year ago

The 2D corrections here should be revisited when addressing issue #166