Hydrogen Combustion supplemetary data from an active learning study

This data is generated in an active learning study of hydrogen combustion, with energies and forces generated using the ωB97X-V DFT functional with the cc-pVTZ basis set. It is build on top of the original HCombustion dataset: https://figshare.com/articles/dataset/Hydrogen_Combustion_using_IRC_AIMD_and_normal_modes/19601689

There are two parts to this data. al_added_npz which are data added during the active learning cycle, and irc_dialation which are the irc dialation data described in the paper Beyond potential energy surface benchmarking: a complete application of machine learning to chemical reactivity (link to be added).

All data are stored in npz format. They can be loaded through numpy by using:
np.load("name.npz", allow_pickle=True)
This will load a dictionary with the following keys:
'R': positions, array of (n_snapshots, n_atoms, 3)
'Z': atomic numbers, array of (n_snapshots, n_atoms)
'E': energy, array of (n_snapshots, 1)
'F': forces, array of (n_snapshots, n_atoms, 3)
coordinates are in the unit of Angstrom, energy in kcal/mol, forces in kcal/mol/A.