dxtb.config.Config

class dxtb.config.Config(*, file=None, strict=False, exclude=[], method=1, grad=False, batch_mode=0, int_cutoff=50.0, int_driver=0, int_level=2, int_uplo='l', anomaly=False, device=device(type='cpu'), dtype=torch.float32, maxiter=100, mixer=2, mix_guess=True, damp=0.5, damp_init=0.1, damp_dynamic=False, damp_dynamic_factor=0.99, damp_soft_start=True, damp_generations=5, damp_diagonal_offset=0.01, guess=0, scf_mode=0, scp_mode=2, x_atol=0.0001, x_atol_max=1e-05, f_atol=0.0001, force_convergence=False, fermi_etemp=300.0, fermi_maxiter=200, fermi_thresh=None, fermi_partition=0, cache_enabled=False, cache_hcore=False, cache_overlap=False, cache_dipole=False, cache_quadrupole=False, cache_charges=True, cache_coefficients=False, cache_density=False, cache_fock=False, cache_iterations=True, cache_mo_energies=False, cache_occupation=False, cache_potential=False, max_element=86, skip_compat_checks=False)

Bases: object

Configuration of the calculation.

Methods

from_args	Create a configuration from command-line arguments.
from_dict	Create a configuration from a dictionary.
from_json	Create a configuration from a JSON file.
info	Return a dictionary with the configuration information.
to_json	Serialize the configuration to a JSON-formatted string.

Attributes

batch_mode	Whether multiple systems or a single one are handled.
strict	Strict mode for SCF configuration.
file	The input file or directory.
exclude	The tight-binding components to exclude from the calculation.
method	The xTB method to use.
grad	Whether to compute the gradient.
max_element	The maximum element number in the system.
anomaly	Whether to run PyTorch in anomaly detection mode.
device	The device to use for the calculation.
dtype	The data type to use for the calculation.
cache	The cache configuration.
ints	The integral configuration.
scf	The SCF configuration.

Parameters:

• file (str | Path | None)

• strict (bool)

• exclude (str | list[str])

• method (int)

• grad (bool)

• batch_mode (int)

• int_cutoff (float)

• int_driver (str | int)

• int_level (int)

• int_uplo (str)

• anomaly (bool)

• device (device)

• dtype (dtype)

• maxiter (int)

• mixer (str | int)

• mix_guess (bool)

• damp (float)

• damp_init (float)

• damp_dynamic (bool)

• damp_dynamic_factor (float)

• damp_soft_start (bool)

• damp_generations (int)

• damp_diagonal_offset (float)

• guess (str | int)

• scf_mode (str | int)

• scp_mode (str | int)

• x_atol (float)

• x_atol_max (float)

• f_atol (float)

• force_convergence (bool)

• fermi_etemp (float)

• fermi_maxiter (int)

• fermi_thresh (float | int | None)

• fermi_partition (str | int)

• cache_enabled (bool)

• cache_hcore (bool)

• cache_overlap (bool)

• cache_dipole (bool)

• cache_quadrupole (bool)

• cache_charges (bool)

• cache_coefficients (bool)

• cache_density (bool)

• cache_fock (bool)

• cache_iterations (bool)

• cache_mo_energies (bool)

• cache_occupation (bool)

• cache_potential (bool)

• max_element (int)

• skip_compat_checks (bool)

classmethod from_args(args)

Create a configuration from command-line arguments.

Return type:

Self

Parameters:

args (Namespace) – The parsed command-line arguments.

Returns:

The configuration object.

Return type:

Self

Parameters:

args (Namespace)

classmethod from_dict(cfg)

Create a configuration from a dictionary.

Return type:

Self

Parameters:

cfg (dict[str, Any]) – The configuration dictionary.

Returns:

The configuration object.

Return type:

Self

Parameters:

cfg (dict[str, Any])

classmethod from_json(path)

Create a configuration from a JSON file.

Return type:

Self

Parameters:

path (PathLike) – The path to the JSON file.

Returns:

The configuration object.

Return type:

Self

Raises:

FileNotFoundError – If the file does not exist.

Parameters:

path (str | Path)

info()

Return a dictionary with the configuration information.

Return type:

dict[str, dict[str, Any]]

Returns:

The configuration information.

Return type:

dict[str, dict[str, Any]]

to_json(path=None)

Serialize the configuration to a JSON-formatted string.

Return type:

str

Parameters:

path (str | Path | None)

Returns:

str: A JSON-formatted string representing the configuration.

anomaly: bool

Whether to run PyTorch in anomaly detection mode.

property batch_mode: int

Whether multiple systems or a single one are handled.

The following batch modes are available:

• 0: Single system

• 1: Multiple systems with padding

• 2: Multiple systems with no padding (conformer ensemble)

Returns:

The batch mode.

Return type:

int

cache: ConfigCache

The cache configuration.

device: device

The device to use for the calculation.

dtype: dtype

The data type to use for the calculation.

exclude: str | list[str]

The tight-binding components to exclude from the calculation.

file: str | Path | None

The input file or directory.

grad: bool

Whether to compute the gradient.

ints: ConfigIntegrals

The integral configuration.

max_element: int

The maximum element number in the system.

method: int

The xTB method to use.

scf: ConfigSCF

The SCF configuration.

strict: bool = False

Strict mode for SCF configuration. Always throws errors if True.