Structure Descriptors for Machine Learning

[jan-janssen]

I started to add interfaces for the SNAP #71 and SOAP #74 descriptors. But as we already have a series of machine learning projects for grain boundaries and other defect structures @ligerzero-ai and @pmrv can you add your descriptors to the sructuretoolkit as well?

[ligerzero-ai]

Are we only looking for structure descriptors?

I've only the voronoi computations in that case. And I thought you said you would like to use pyscal for the voronoi computations? I use the pymatgen package for the computations.

For the chemical descriptors, yes I can add them.

this will require time as i've not configured them separately.

[jan-janssen]

@ligerzero-ai The idea is simply to get an overview of the descriptors which were already successfully applied in the department, so that new students can start with a reasonable selection. That is why I try to integrate the descriptors inside the structuretoolkit package. In this way I would be very interested if you could share your experience which descriptors worked for you and maybe a short example how to calculate them for individual ASE structures.