qc2.ase.pyscf

This module defines an ASE interface to PySCF.

https://pyscf.org/ => Official documentation https://github.com/pyscf/pyscf => GitHub page

Note: Adapted from https://github.com/pyscf/pyscf/blob/master/pyscf/pbc/tools/pyscf_ase.py & https://github.com/pyscf/pyscf/issues/624

Module Contents

Classes

PySCF

An ASE calculator for the PySCF quantum chemistry package.

Functions

ase_atoms_to_pyscf(→ List[List[Union[str, numpy.ndarray]]])

Converts ASE atoms to PySCF atom.
qc2.ase.pyscf.ase_atoms_to_pyscf(ase_atoms: ase.Atoms) → List[List[str | numpy.ndarray]][source]

Converts ASE atoms to PySCF atom.

Parameters:

ase_atoms (Atoms) – ASE Atoms object.

Returns:

PySCF atom.

Return type:

List[List[Union[str, np.ndarray]]]

class qc2.ase.pyscf.PySCF(restart: bool | None = None, ignore_bad_restart: bool | None = False, label: str | None = 'PySCF', atoms: ase.Atoms | None = None, command: str | None = None, directory: str = '.', **kwargs)[source]

Bases: ase.calculators.calculator.Calculator, qc2.ase.qc2_ase_base_class.BaseQc2ASECalculator

An ASE calculator for the PySCF quantum chemistry package.

Parameters:

  • Calculator (Calculator) – Base-class for all ASE calculators.

  • BaseQc2ASECalculator (BaseQc2ASECalculator) – Base class for ase calculartors in qc2.

Raises:

  • InputError – If attributes other than ‘method’, ‘xc’, ‘basis’, ‘multiplicity’, ‘charge’, ‘relativistic’, ‘cart’, ‘scf_addons’, ‘verbose’, and ‘output’ are input as Calculator.

  • CalculatorSetupError – If abinitio methods other than ‘scf.RHF’, ‘scf.UHF’, ‘scf.ROHF’, ‘dft.RKS’, ‘dft.UKS’, and ‘dft.ROKS’ are selected.

implemented_properties: List[str] = ['energy', 'forces'][source]
default_parameters: Dict[str, Any][source]
check_pyscf_attributes() → None[source]

Checks for any missing and/or mispelling PySCF input attribute.

Notes

it can also be used to eventually set specific environment variables, ios, etc.

calculate(atoms: ase.Atoms | None = None, properties: List[str] = ['energy'], system_changes: List[str] = all_changes) → None[source]

This method is the core responsible for the actual calculation.

Notes

Implementation based on Calculator.calculate() method. see also ase/ase/calculators/calculator.py.

Parameters:

  • atoms (Atoms, optional) – Atoms object to which the calculator is attached. Defaults to None.

  • properties (list[str], optional) – List of what properties need to be calculated. Defaults to [‘energy’].

  • system_changes (list[str], optional) – List of what has changed since last calculation. Can be any combination of these six: ‘positions’, ‘numbers’, ‘cell’, ‘pbc’, ‘initial_charges’ and ‘initial_magmoms’. Defaults to all_changes.

Raises:

  • CalculatorSetupError – If a proper geometry is not provided.

  • CalculatorSetupError – If abinitio methods other than

  • 'scf.RHF', 'scf.UHF', 'scf.ROHF', 'dft.RKS', 'dft.UKS', – and ‘dft.ROKS’ are selected.

save(datafile: h5py.File | str) → None[source]

Dumps qchem data to a datafile using QCSchema or FCIDump formats.

Parameters:

datafile (Union[h5py.File, str]) – file to save the data to.

Returns:

None

Notes

files are written following the QCSchema or FCIDump formats.

Example

>>> from ase.build import molecule
>>> from qc2.ase import PySCF
>>>
>>> molecule = molecule('H2')
>>> molecule.calc = PySCF()  # => RHF/STO-3G
>>> molecule.calc.schema_format = "qcschema"
>>> molecule.calc.get_potential_energy()
>>> molecule.calc.save('h2.hdf5')
>>>
>>> molecule = molecule('H2')
>>> molecule.calc = PySCF()  # => RHF/STO-3G
>>> molecule.calc.schema_format = "fcidump"
>>> molecule.calc.get_potential_energy()
>>> molecule.calc.save('h2.fcidump')
load(datafile: h5py.File | str) → qiskit_nature.second_q.formats.qcschema.QCSchema | qiskit_nature.second_q.formats.fcidump.FCIDump[source]

Loads electronic structure data from a datafile.

Parameters:

datafile (Union[h5py.File, str]) – file to read the data from.

Returns:

Instances of QCSchema or FCIDump dataclasses containing qchem data.

Notes

files are read following the qcschema or fcidump formats.

Example

>>> from ase.build import molecule
>>> from qc2.ase import PySCF
>>>
>>> molecule = molecule('H2')
>>> molecule.calc = PySCF()     # => RHF/STO-3G
>>> molecule.calc.schema_format = "qcschema"
>>> qcschema = molecule.calc.load('h2.h5')
>>>
>>> molecule = molecule('H2')
>>> molecule.calc = PySCF()     # => RHF/STO-3G
>>> molecule.calc.schema_format = "fcidump"
>>> fcidump = molecule.calc.load('h2.fcidump')
get_integrals_mo_basis() → Tuple[numpy.ndarray, numpy.ndarray, numpy.ndarray, numpy.ndarray, numpy.ndarray, numpy.ndarray][source]

Retrieves 1- & 2-body integrals in MO basis from PySCF routines.

Returns:

  • one_body_int_a & one_body_int_b: Numpy arrays containing alpha and beta components of the one-body integrals.

  • two_body_int_aa, two_body_int_bb, two_body_int_ab & two_body_int_ba: Numpy arrays containing alpha-alpha, beta-beta, alpha-beta & beta-alpha components of the two-body integrals.

Return type:

A tuple containing np.ndarray types

get_integrals_ao_basis() → Tuple[numpy.ndarray, numpy.ndarray][source]

Retrieves 1- & 2-e integrals in AO basis from PySCF routines.

Returns:

Tuple containing the 1-electron integrals and the 2-electron integrals in the AO basis.

Return type:

Tuple[np.ndarray, np.ndarray]

get_molecular_orbitals_coefficients() → Tuple[numpy.ndarray, numpy.ndarray][source]

Retrieves alpha and beta MO coeffs from PySCF routines.

Returns:

Tuple containing the alpha and beta MO coefficients.

Return type:

Tuple[np.ndarray, np.ndarray]

get_molecular_orbitals_energies() → Tuple[numpy.ndarray, numpy.ndarray][source]

Retrieves alpha and beta MO energies from PySCF routines.

Returns:

Tuple containing the alpha and beta MO energies.

Return type:

Tuple[np.ndarray, np.ndarray]

get_overlap_matrix() → numpy.ndarray[source]

Retrieves overlap matrix from PySCF routines.

Returns:

The overlap matrix.

Return type:

np.ndarray

_expand_mo_object(mo_object: Tuple[numpy.ndarray | None, numpy.ndarray | None] | numpy.ndarray, array_dimension: int = 2) → Tuple[numpy.ndarray, numpy.ndarray][source]

Expands the mo object into alpha- and beta-spin components.

Parameters:

  • mo_object – the molecular orbital object to expand.

  • array_dimension – This argument specifies the dimension of the numpy array (if a tuple is not encountered). Making this configurable permits this function to be used to expand both, MO coefficients (3D array) and MO energies (2D array).

Returns:

The (alpha, beta) tuple of MO data.

Notes

Adapted from Qiskit-Nature pyscfdriver.py.