search

Probe-Particle / PPSTM

Code simulating variousSTM techniques, especially for tilting tip (dependent on https://github.com/Probe-Particle/ppafm)l)
MIT License
7 stars 1 forks source link

problem with FHI-aims and cube files starting not exactly at (0,0,0) #4

Open ondrejkrejci opened 1 year ago

ondrejkrejci commented 1 year ago

Problem:

if in FHI-aims control.in one is having:

output cube hartree_potential

the cube starting vector is not (0,0,0): e.g.:

CUBE FILE written by FHI-AIMS
 *****************************
  175    0.094308    0.094531    0.094486
  190    0.188616    0.000000    0.000000
  126    0.000000    0.189061    0.000000
  200    0.000000    0.000000    0.188973
...

even though the difference is only 0.05 Å - this is not ideal in terms of simulations via PPSTM_simple.py, or the GUI.py, where we are using geometry.in directly within the FHI-aims reading procedure. The shift in the cube file causes shift of the internal coordinates in the PP-AFM precalculations.

Current solution:

The only way around is to use the latest version of PP-AFM and mv geometry.in geometry-aims.in and use ASE to get new geometry.in from any of the xsf files, that PP-AFM produces.

We should find a better (and more automatized) way around it!

ondrejkrejci commented 1 year ago

We could add a way into PPSTM_simple.py and GUI to upload the geometry from npy or xsf files directly in the FHI-aims reading procedure. There also should be some "error proof" check on the way somewhere as on can make a mistake easily.