Dataset Augmentation WorkChain

Dataset Augmentation WorkChain#

Dataset Augmentation WorkChain
workchainaiida_trains_pot.datasetaugmentation.datasetaugmentation_wc.datasetaugmentation_wc.DatasetAugmentationWorkChain

WorkChain to generate a training dataset.

Inputs:

  • clusters, Namespace
    • interatomic_distance, (Int, Float, NoneType), optional – Interatomic distance. Default: uuid: 9472632f-efb9-4804-8c71-9d3ace2a876b (unstored) value: 1.5
    • max_atoms, (Int, NoneType), optional – Maximum number of atoms in each cluster. Default: uuid: 741c76fb-bd3d-4d64-922b-8bb441e3c52b (unstored) value: 10
    • n_clusters, (Int, NoneType), optional – Number of clusters to generate. Default: uuid: 86b55342-9829-42fc-9fe2-efc34876f7e5 (unstored) value: 20
  • do_check_vacuum, (Bool, NoneType), optional – Check if vacuum along non periodic directions is enough and add it if necessary. Default: uuid: 81fff088-6874-4068-b6c4-8694442d78f5 (unstored) value: True
  • do_clusters, (Bool, NoneType), optional – Add clusters to the dataset. Default: uuid: 40639a29-63d8-4a8e-a997-0cb3c02e5e57 (unstored) value: True
  • do_input, (Bool, NoneType), optional – Add input structures to the dataset. Default: uuid: 86852a06-9f28-41fa-b1cc-fa62fad2459c (unstored) value: True
  • do_isolated, (Bool, NoneType), optional – Add isolated atoms configurations to the dataset. Default: uuid: ce511a43-3026-4c02-91f1-7ea4a656020f (unstored) value: True
  • do_rattle_strain_defects, (Bool, NoneType), optional – Perform rattle calculations (random atomic displacements, cell stretch/compression, vacancies. Permutations and replacements are not yet implemented). Default: uuid: 086f19e7-a149-4c73-a044-ca51f6d1f042 (unstored) value: True
  • do_replication, (Bool, NoneType), optional – Replicate structures to have a minimum distance between atoms greater than min_dist. Default: uuid: 90a54d9d-de65-4dcd-bede-1ad1652fa1a4 (unstored) value: True
  • do_slabs, (Bool, NoneType), optional – Add slabs to the dataset. Default: uuid: a7ca8792-7f71-4ed6-8246-e7030a2ae989 (unstored) value: True
  • do_substitution, (Bool, NoneType), optional – Add substituted structures to the dataset. Default: uuid: db8bf706-363c-437f-92f0-3007899bce63 (unstored) value: True
  • metadata, Namespace
    • call_link_label, str, optional, is_metadata – The label to use for the CALL link if the process is called by another process.
    • description, (str, NoneType), optional, is_metadata – Description to set on the process node.
    • disable_cache, (bool, NoneType), optional, is_metadata – Do not consider the cache for this process, ignoring all other caching configuration rules.
    • label, (str, NoneType), optional, is_metadata – Label to set on the process node.
    • store_provenance, bool, optional, is_metadata – If set to False provenance will not be stored in the database.
  • replicate, Namespace
    • max_atoms, (Int, NoneType), optional – Maximum number of atoms in the supercell. Stronger criteria respect to min_dist. Default: uuid: 42dfeeb1-2da2-42ad-97d3-9fbcaa07823c (unstored) value: 450
    • min_dist, (Int, Float, NoneType), optional – Minimum distance between atoms in PBC replicas, unless max_atoms is reached. Default: uuid: 6b1258f0-7356-4ad2-aa73-53cbe2369876 (unstored) value: 18.0
  • rsd, Namespace
    • params, Namespace
      • frac_vacancies, (Int, Float, NoneType), optional – Fraction of configurations with vacancies. Default: uuid: 0ed5b07a-3986-4904-90fa-1dd411bef0a8 (unstored) value: 0.3
      • max_compressive_strain, (Int, Float, NoneType), optional – Maximum compressive strain factor. Cell can be compressed up to this fraction of cell parameters. Default: uuid: aa6cee35-2f05-4295-9783-34575c272395 (unstored) value: 0.2
      • max_tensile_strain, (Int, Float, NoneType), optional – Maximum tensile strain factor. Cell can be stretched up to this fraction of cell parameters. Default: uuid: 4aa05e8b-a472-4524-b2c0-26c76b5e57fe (unstored) value: 0.6
      • n_configs, (Int, NoneType), optional – Number of configurations to generate per each input structure. Default: uuid: d1e15c9f-3c36-4984-9d07-9d80aba21b04 (unstored) value: 50
      • rattle_fraction, (Int, Float, NoneType), optional – Atoms are displaced by a rattle_fraction of the minimum interatomic distance. Default: uuid: 019d1508-4abd-4ff9-9c58-fa9321cf15ac (unstored) value: 0.3
      • vacancies_per_config, (Int, NoneType), optional – Number of vacancies per configuration. Default: uuid: e4f0c1ee-cd77-427f-b8fe-eaeade3ba2ce (unstored) value: 2
  • slabs, Namespace
    • max_atoms, (Int, NoneType), optional – Maximum number of atoms. Default: uuid: 1df78ee4-a4ad-4385-8209-8911a7d5a505 (unstored) value: 450
    • miller_indices, (List, NoneType), optional – List of lists with the Miller indices. Default: uuid: aec04ed9-fbf6-4701-87fb-3a221caab0ee (unstored) value: [[1, 1, 1], [1, 1, 0], [1, 0, 0]]
    • min_thickness, (Int, Float, NoneType), optional – Minimum thickness of the slab. Default: uuid: b2197763-3229-4ec3-9269-ecb5e07b2989 (unstored) value: 10.0
  • structures, PESData, required – PESData, dataset containing input structures.
  • substitution, Namespace
    • structures_fraction, (Int, Float, NoneType), optional – Fraction of structures to be substituted. Default: uuid: c9bf439b-b7ab-4961-b0bf-8b46c2bafa69 (unstored) value: 0.2
    • switches_fraction, (Int, Float, NoneType), optional – Fraction of atoms to be substituted. Default: uuid: 0c97d4e1-8b29-4a89-8d6b-e4daa0661a05 (unstored) value: 0.2
  • vacuum, (Int, Float, NoneType), optional – Minimum vacuum along non periodic directions. Default: uuid: b4ccd51f-113f-4874-be02-144c0495dd84 (unstored) value: 15.0

Outputs:

  • structures, Namespace – Augmented datasets.

Outline:

check_inputs(Check inputs.)
if(do_replication)
    replicate(Replicate structures.)
run_dataset_generation(Generate datasets.)