Structure & Identifiers

A Structure is the core qcdata object for representing a molecule or molecular super structure in 3D space. Structure objects can be created directly from symbol and geometry information (geometry must be in Bohr), from SMILES strings, from xyz files, or opened from Structure objects previously saved to disk.

qcdata.Structure

Structure(**data: Any)

A Structure object with atoms and their corresponding cartesian coordinates, charge, multiplicity, and identifiers such as name, smiles, etc.

Attributes:

Name	Type	Description
symbols	list[str]	The atomic symbols of the structure.
geometry	SerializableNDArray	The geometry of the structure in Cartesian coordinates. Units are Bohr (AU).
identifiers	Identifiers	Identifiers for the structure such as name, smiles, etc.
charge	int	The molecular charge.
multiplicity	int	The molecular multiplicity.
connectivity	list[tuple[int, int, float]]	Explicit description of the bonds between atoms. Each tuple contains the indices of the atoms in the bond and the order of the bond. E.g., [(0, 1, 1.0), (1, 2, 2.0)] indicates a single bond between atoms 0 and 1 and a double bond between atoms 1 and 2.
extras	Dict[str, Any]	Additional information to bundle with the object. Use for schema development and scratch space.
ids	Identifiers	@property Shortcut to access identifiers.
geometry_angstrom	ndarray	@property The geometry of the structure in Angstrom.
atomic_numbers	list[int]	@property The atomic numbers of the atoms in the structure.
formula	str	@property The molecular formula of the structure using the Hill System.

Example

from qcdata import Structure

structure = Structure(
    symbols=["H", "O", "H"],
    geometry=[[0.0, 0.0, 0.0], [0.0, 0.0, 1.0], [0.0, 0.0, 2.0]],
    charge=0,  # optional; defaults to 0
    multiplicity=1,  # optional; defaults to 1
    identifiers={"smiles": "O"},  # optional
)

Source code in src/qcdata/models/structure.py

def __init__(self, **data: Any):
    """Create a new Structure object.

    Example:
        ```python
        from qcdata import Structure

        structure = Structure(
            symbols=["H", "O", "H"],
            geometry=[[0.0, 0.0, 0.0], [0.0, 0.0, 1.0], [0.0, 0.0, 2.0]],
            charge=0,  # optional; defaults to 0
            multiplicity=1,  # optional; defaults to 1
            identifiers={"smiles": "O"},  # optional
        )

        ```
    """
    # Backwards compatibility for 'ids' attribute.
    if identifiers := data.pop("ids", None):
        warnings.warn(
            "Passing 'ids' is deprecated and will be removed in a future "
            "release. Please use 'identifiers' instead. Once instantiated, "
            "you can use structure.ids to access the identifiers as a shortcut.",
            category=FutureWarning,
            stacklevel=2,
        )
        data["identifiers"] = identifiers
    super().__init__(**data)

from_xyz classmethod

from_xyz(
    xyz_str: str,
    *,
    charge: int | None = None,
    multiplicity: int | None = None,
) -> Self

Create a Structure from an XYZ file or string.

Parameters:

Name	Type	Description	Default
xyz_str	str	The XYZ string.	required
charge	int | None	The molecular charge of the structure. If not provided, will read from the XYZ string if set or default to 0.	None
multiplicity	int | None	The molecular multiplicity of the structure. If not provided, will read from the XYZ string if set or default to 1.	None

Note

Will read qcdata data such as charge and multiplicity from the comments line with a qcdata_key=value format (if it is present). Also will read in qcdata__identifiers_* keys and additional non-qcdata comments.

Example

struct = Structure.from_xyz(xyz_str)

Source code in src/qcdata/models/structure.py

@classmethod
def from_xyz(
    cls,
    xyz_str: str,
    *,
    charge: int | None = None,
    multiplicity: int | None = None,
) -> Self:
    """Create a Structure from an XYZ file or string.

    Args:
        xyz_str: The XYZ string.
        charge: The molecular charge of the structure. If not provided, will read
            from the XYZ string if set or default to 0.
        multiplicity: The molecular multiplicity of the structure. If not provided,
            will read from the XYZ string if set or default to 1.

    Note:
        Will read qcdata data such as `charge` and `multiplicity` from the comments
        line with a `qcdata_key=value` format (if it is present). Also will read in
        qcdata__identifiers_* keys and additional non-qcdata comments.

    Example:
        ```python
        struct = Structure.from_xyz(xyz_str)
        ```
    """

    lines = xyz_str.split("\n")

    num_atoms = int(lines[0])

    # Collect comments
    structure_kwargs: dict[str, Any] = {}
    identifier_kwargs: dict[str, Any] = {}
    other_comments: list[str] = []

    for item in lines[1].strip().split():
        if item.startswith("qcdata__identifiers_"):
            key = item.split("=")[0].replace("qcdata__identifiers_", "")
            value = item.split("=")[1]
            identifier_kwargs[key] = value
        elif item.startswith("qcdata_"):
            key = item.split("=")[0].replace("qcdata_", "")
            value = item.split("=")[1]
            structure_kwargs[key] = value
        else:
            other_comments.append(item)

    if charge is not None and "charge" in structure_kwargs:
        raise ValueError("Charge cannot be set in the file and as an argument.")
    if multiplicity is not None and "multiplicity" in structure_kwargs:
        raise ValueError(
            "Multiplicity cannot be set in the file and as an argument."
        )

    # Set charge and multiplicity if provided
    if charge is not None:
        structure_kwargs["charge"] = charge
    if multiplicity is not None:
        structure_kwargs["multiplicity"] = multiplicity

    symbols = []
    geometry = []
    for line in lines[2 : 2 + num_atoms]:
        split_line = line.split()
        symbols.append(split_line[0])
        geometry.append([float(val) / BOHR_TO_ANGSTROM for val in split_line[1:]])

    return cls(
        symbols=symbols,
        geometry=geometry,
        **structure_kwargs,
        identifiers=Identifiers(**identifier_kwargs),
        extras={cls._xyz_comment_key: other_comments},
    )

to_xyz

to_xyz(precision: int = 17) -> str

Return an xyz string representation of the structure.

Parameters:

Name	Type	Description	Default
precision	int	The number of decimal places to include in the xyz file. Default 17 which captures all precision of float64.	17

Notes: Will add qcdata data such as charge and multiplicity to the comments line with a qcdata_key=value format.

Source code in src/qcdata/models/structure.py

def to_xyz(self, precision: int = 17) -> str:
    """Return an xyz string representation of the structure.

    Args:
        precision: The number of decimal places to include in the xyz file. Default
            17 which captures all precision of float64.
    Notes:
        Will add qcdata data such as charge and multiplicity to the comments line with
        a `qcdata_key=value` format.
    """

    qcdata_data = {  # These get added to comments line (line 2) in xyz file
        "qcdata_charge": self.charge,
        "qcdata_multiplicity": self.multiplicity,
    }

    # Add identifiers to qcdata_data
    for key, value in self.identifiers.__dict__.items():
        if key != "extras" and value:
            qcdata_data[f"qcdata__identifiers_{key}"] = value

    assert isinstance(self.geometry, np.ndarray)  # For mypy
    geometry_angstrom = self.geometry * BOHR_TO_ANGSTROM

    xyz_lines = []
    xyz_lines.append(f"{len(self.symbols)}")
    # Add qcdata data to comments line
    comments = f"{' '.join([f'{k}={v}' for k, v in qcdata_data.items()])}"
    # Add any other comments
    if xyz_comments := self.extras.get(self._xyz_comment_key, []):
        comments += " " + " ".join(xyz_comments)
    xyz_lines.append(comments)

    # Create a format string using the precision parameter
    format_str = f"{{:2s}} {{: >18.{precision}f}} {{: >18.{precision}f}} {{: >18.{precision}f}}"  # noqa: E501

    for symbol, (x, y, z) in zip(self.symbols, geometry_angstrom):
        xyz_lines.append(format_str.format(symbol, x, y, z))
    xyz_lines.append("")  # Append newline to end of file
    return "\n".join(xyz_lines)

distance

distance(i: int, j: int, units: LengthUnit = BOHR) -> float

Calculate the distance between two atoms.

Parameters:

Name	Type	Description	Default
i	int	The index of the first atom.	required
j	int	The index of the second atom.	required
units	LengthUnit	The units to return the distance in. Defaults to "bohr". May be "bohr" or "angstrom".	BOHR

Returns:

Type	Description
float	The distance between the atoms in units (Bohr or Angstrom).

Example

struct.distance(0, 1)
1.34

Source code in src/qcdata/models/structure.py

def distance(self, i: int, j: int, units: LengthUnit = LengthUnit.BOHR) -> float:
    """Calculate the distance between two atoms.

    Args:
        i: The index of the first atom.
        j: The index of the second atom.
        units: The units to return the distance in. Defaults to "bohr".
            May be "bohr" or "angstrom".

    Returns:
        The distance between the atoms in units (Bohr or Angstrom).

    Example:
        ```python
        struct.distance(0, 1)
        1.34
        ```
    """
    distance = np.linalg.norm(self.geometry[i] - self.geometry[j])
    if units == LengthUnit.ANGSTROM:
        return float(distance * BOHR_TO_ANGSTROM)
    return float(distance)

add_smiles

add_smiles(
    *, program: str = "rdkit", hydrogens: bool = False
) -> None

!! DEPRECATED !!

This helper has been removed to qcinf (see qcinf.structure_to_smiles). It will be removed from qcdata in a future release.

Source code in src/qcdata/models/structure.py

def add_smiles(
    self: "Structure",
    *,
    program: str = "rdkit",
    hydrogens: bool = False,
) -> None:
    """
    !! DEPRECATED !!

    This helper has been removed to **qcinf** (see `qcinf.structure_to_smiles`).
    It will be removed from qcdata in a future release.
    """
    warnings.warn(
        "`Structure.add_smiles()` has moved to `qcinf` and is no longer "
        "implemented here.\n\n"
        "Install qcinf and replace your call with:\n\n"
        "    from qcinf import structure_to_smiles\n"
        "    smiles = structure_to_smiles(struct, backend='rdkit|openbabel')\n"
        "    struct.add_identifiers(smiles=smiles)\n\n",
        DeprecationWarning,  # use FutureWarning if you want it visible by default
        stacklevel=2,
    )
    raise NotImplementedError(
        "Structure.add_smiles() is removed. "
        "Use qcinf.structure_to_smiles and struct.add_identifiers instead."
    )

qcdata.Identifiers

Structure identifiers.

Attributes:

Name	Type	Description
name	str | None	A human-readable, common name for the structure.
name_IUPAC	str | None	The IUPAC name of the structure.
smiles	str | None	The SMILES representation of the structure.
canonical_smiles	str | None	The canonical SMILES representation of the structure.
canonical_smiles_program	str | None	The program used to generate the canonical SMILES.
canonical_explicit_hydrogen_smiles	str | None	The canonical explicit hydrogen SMILES representation of the structure.
canonical_isomeric_smiles	str | None	The canonical isomeric SMILES representation of the structure.
canonical_isomeric_explicit_hydrogen_smiles	str | None	The canonical isomeric explicit hydrogen SMILES representation of the structure.
canonical_isomeric_explicit_hydrogen_mapped_smiles	str | None	The canonical isomeric explicit hydrogen mapped SMILES representation of the structure.
inchi	str | None	The InChI representation of the structure.
inchikey	str | None	The InChIKey representation of the structure.
pubchem_cid	str | None	The PubChem Compound ID of the structure.
pubchem_sid	str | None	The PubChem Substance ID of the structure.
pubchem_conformerid	str | None	The PubChem Conformer ID of the structure.
extras	Dict[str, Any]	Additional information to bundle with the object. Use for schema development and scratch space.