pyscal / pyscal

Python library written in C++ for calculation of local atomic structural environment
https://pyscal.org
BSD 3-Clause "New" or "Revised" License
57 stars 16 forks source link

Voronoi Analysis works fine for 128 atoms but fails for 250 atoms and larger #93

Closed jan-janssen closed 3 years ago

jan-janssen commented 3 years ago

Describe the bug I calculate:

np.mean(lst), np.max(lst), np.min(lst), np.std(lst)
(17.652945, 17.65295, 17.652939, 2.5246266e-06)  # small cell 
(26.47941835024731, 70.12143498554614, 17.3288579518855, 14.405461275280086)   # large cell 

To Reproduce https://github.com/jan-janssen/ovito-vs-pyscal/blob/master/ovito-vs-pyscal-voronoi.ipynb

Expected behavior The mean, maximum, minimum and standard deviation should not change when changing the simulation cell

Additional context The melting point calculations seem to be all fine for fcc, bcc and diamond structures just hcp fails: https://github.com/jan-janssen/pyiron_meltingpoint

srmnitc commented 3 years ago

Thanks for pointing this out. Now fixed. I will make a release soon.

jan-janssen commented 3 years ago

I can confirm it is fixed with the new release. I now opened a pull request for migrating the melting point protocol to pyscal and the results look very promising https://github.com/pyiron/pyiron_meltingpoint/pull/9#issuecomment-738643029 - maybe you can even use it to compare your melting points.