Skip to content

qmof

QMOF-compatible recipes.

This set of recipes is meant to be compatible with the QMOF Database workflow. Reference: https://doi.org/10.1016/j.matt.2021.02.015

LOGGER module-attribute 

LOGGER = getLogger(__name__)

qmof_relax_job 

qmof_relax_job(
    atoms: Atoms,
    relax_cell: bool = True,
    run_prerelax: bool = True,
    copy_files: (
        SourceDirectory
        | dict[SourceDirectory, Filenames]
        | None
    ) = None,
    **calc_kwargs
) -> QMOFRelaxSchema

Relax a structure in a multi-step process for increased computational efficiency. This is all done in a single compute job. Settings are such that they are compatible with the QMOF Database.

  1. A "pre-relaxation" with BFGSLineSearch to resolve very high forces.

  2. Position relaxation with default ENCUT and coarse k-point grid.

  3. Optional: volume relaxation with coarse k-point grid.

  4. Double relaxation using production-quality settings.

  5. Static calculation.

Parameters:

  • atoms (Atoms) –

    Atoms object

  • relax_cell (bool, default: True ) –

    True if a volume relaxation should be performed. False if only the positions should be updated.

  • run_prerelax (bool, default: True ) –

    If True, a pre-relax will be carried out with BFGSLineSearch. Recommended if starting from hypothetical structures or materials with very high starting forces.

  • copy_files (SourceDirectory | dict[SourceDirectory, Filenames] | None, default: None ) –

    Files to copy (and decompress) from source to the runtime directory.

  • **calc_kwargs –

    Custom kwargs for the calculator. Set a value to None to remove a pre-existing key entirely. Applies for all jobs.

Returns:

  • QMOFRelaxSchema –

    Dictionary of results. See the type-hint for the data structure.

Source code in quacc/recipes/vasp/qmof.py 
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
@job
def qmof_relax_job(
    atoms: Atoms,
    relax_cell: bool = True,
    run_prerelax: bool = True,
    copy_files: SourceDirectory | dict[SourceDirectory, Filenames] | None = None,
    **calc_kwargs,
) -> QMOFRelaxSchema:
    """
    Relax a structure in a multi-step process for increased computational efficiency.
    This is all done in a single compute job. Settings are such that they are compatible
    with the QMOF Database.

    1. A "pre-relaxation" with BFGSLineSearch to resolve very high forces.

    2. Position relaxation with default ENCUT and coarse k-point grid.

    3. Optional: volume relaxation with coarse k-point grid.

    4. Double relaxation using production-quality settings.

    5. Static calculation.

    Parameters
    ----------
    atoms
        Atoms object
    relax_cell
        True if a volume relaxation should be performed. False if only the
        positions should be updated.
    run_prerelax
        If True, a pre-relax will be carried out with BFGSLineSearch.
        Recommended if starting from hypothetical structures or materials with
        very high starting forces.
    copy_files
        Files to copy (and decompress) from source to the runtime directory.
    **calc_kwargs
        Custom kwargs for the calculator. Set a value to `None` to remove
        a pre-existing key entirely. Applies for all jobs.

    Returns
    -------
    QMOFRelaxSchema
        Dictionary of results. See the type-hint for the data structure.
    """
    copy_files = None

    # 1. Pre-relaxation
    if run_prerelax:
        summary1 = _prerelax(atoms, **calc_kwargs)
        atoms = summary1["atoms"]
        copy_files = {summary1["dir_name"]: ["WAVECAR*"]}

    # 2. Position relaxation (loose)
    summary2 = _loose_relax_positions(atoms, copy_files=copy_files, **calc_kwargs)
    atoms = summary2["atoms"]
    copy_files = {summary2["dir_name"]: ["WAVECAR*"]}

    # 3. Optional: Volume relaxation (loose)
    if relax_cell:
        summary3 = _loose_relax_cell(atoms, copy_files=copy_files, **calc_kwargs)
        atoms = summary3["atoms"]
        copy_files = {summary3["dir_name"]: ["WAVECAR*"]}

    # 4. Double Relaxation
    # This is done for two reasons: a) because it can
    # resolve repadding issues when dV is large; b) because we can use LREAL =
    # Auto for the first relaxation and the default LREAL for the second.
    summary4 = _double_relax(
        atoms, relax_cell=relax_cell, copy_files=copy_files, **calc_kwargs
    )
    atoms = summary4[-1]["atoms"]
    copy_files = {summary4[-1]["dir_name"]: ["WAVECAR*"]}

    # 5. Static Calculation
    summary5 = _static(atoms, copy_files=copy_files, **calc_kwargs)
    summary5["prerelax_lowacc"] = summary1 if run_prerelax else None
    summary5["position_relax_lowacc"] = summary2
    summary5["volume_relax_lowacc"] = summary3 if relax_cell else None
    summary5["double_relax"] = summary4

    return summary5