Bulk Silicium with space groups

This dataset aims at training models to predict crystallographic spacegroup from finite-temperature atomistic positions. It contains pure Silicium configurations (namely all atomic positions) and their associated spacegroup international number for five different bulk phases : 3128 monoclinic configurations (C2/m - spacegroup international number 12 - Materials Project identifier mp-1079297) 3000 β-Sn configurations (I4₁/amd - spacegroup international number 141 - Materials Project identifier mp-92) 3000 hexagonalconfigurations (P6₃/mmc - spacegroup international number 194 - Materials Project identifier mp-165) 3000 cubic configurations (Pm3̅n - spacegroup international number 223 - Materials Project identifier mp-971662) 112 diamond configurations (Fd3̅m - spacegroup international number 227 - Materials Project identifier mp-149) Those configurations have been obtained using VASP ab-initio molecular dynamics as explained in Sec. III.C of : BISSUEL, Dylan, ALBARET, Tristan, et NIEHAUS, Thomas A. Critical assessment of machine-learned repulsive potentials for the density functional based tight-binding method: A case study for pure silicon. The Journal of Chemical Physics, 2022, vol. 156, no 6.