A1 Refereed original research article in a scientific journal
Updates to the DScribe library: New descriptors and derivatives
Authors: Laakso Jarno, Himanen Lauri, Homm Henrietta, Morooka Eiaki V., Jäger Marc O. J., Todorović Milika, Rinke Patrick
Publisher: AIP Publishing
Publication year: 2023
Journal: Journal of Chemical Physics
Journal acronym: J CHEM PHYS
Article number: 234802
Volume: 158
Issue: 23
Number of pages: 8
ISSN: 0021-9606
eISSN: 1089-7690
DOI: https://doi.org/10.1063/5.0151031(external)
Web address : https://doi.org/10.1063/5.0151031(external)
Preprint address: https://arxiv.org/abs/2303.14046(external)
Abstract
We present an update of the DScribe package, a Python library for atomistic descriptors. The update extends DScribe's descriptor selection with the Valle-Oganov materials fingerprint and provides descriptor derivatives to enable more advanced machine learning tasks, such as force prediction and structure optimization. For all descriptors, numeric derivatives are now available in DScribe. For the many-body tensor representation (MBTR) and the Smooth Overlap of Atomic Positions (SOAP), we have also implemented analytic derivatives. We demonstrate the effectiveness of the descriptor derivatives for machine learning models of Cu clusters and perovskite alloys.
We present an update of the DScribe package, a Python library for atomistic descriptors. The update extends DScribe's descriptor selection with the Valle-Oganov materials fingerprint and provides descriptor derivatives to enable more advanced machine learning tasks, such as force prediction and structure optimization. For all descriptors, numeric derivatives are now available in DScribe. For the many-body tensor representation (MBTR) and the Smooth Overlap of Atomic Positions (SOAP), we have also implemented analytic derivatives. We demonstrate the effectiveness of the descriptor derivatives for machine learning models of Cu clusters and perovskite alloys.
UTU Research Portal