pfd.exploration.inference package

Submodules

pfd.exploration.inference.eval_dp module

class pfd.exploration.inference.eval_dp.DPTest(model: Path | str | None = None, data: Path | str | None = None, **kwargs)[source]

Bases: EvalModel

classmethod args()[source]

classmethod doc()[source]

evaluate(name: str = 'default', prefix: Path | str = './', **kwargs)[source]

inference(name: str = 'default', prefix: Path | str = './', **kwargs)[source]

load_model(model: Path | str, **kwargs)[source]

read_data(data, fmt='deepmd/npy', **kwargs)[source]

read_data_unlabeled(data, fmt='deepmd/npy', **kwargs)[source]

pfd.exploration.inference.eval_model module

class pfd.exploration.inference.eval_model.EvalModel(model: Path | str | None = None, data: Path | str | None = None, **kwargs)[source]

Bases: ABC

The base class for inference and evaluation.

Parameters:: ABC (_type_) – _description_
Returns:: _description_
Return type:: _type_

clear_data()[source]

clear_model()[source]

property data

abstractmethod evaluate(**kwargs)[source]

static get_driver(key: str)[source]

Get a driver for ModelEval

Parameters:: key (str) – _description_
Raises:: RuntimeError – _description_
Returns:: _description_
Return type:: _type_

static get_drivers() → dict[source]

Get all drivers

Returns:: all drivers
Return type:: dict

abstractmethod inference(**kwargs)[source]

abstractmethod load_model(model: Path | str, **kwargs)[source]

property model

abstractmethod read_data(data: Path | str, **kwargs)[source]

abstractmethod read_data_unlabeled(data: Path | str, **kwargs)[source]

static register(key: str)[source]

Register a model interface. Used as decorators

Parameters:: key (str) – key of the model

class pfd.exploration.inference.eval_model.TestReport(name: str = 'default_system', system: dpdata.system.System | None = None, atom_numb: int = 0, numb_frame: int = 0, mae_f: float = 0, rmse_f: float = 0, mae_e: float = 0, rmse_e: float = 0, mae_e_atom: float = 0, rmse_e_atom: float = 0, mae_v: float = 0, rmse_v: float = 0, lab_e: numpy.ndarray | None = None, pred_e: numpy.ndarray | None = None, lab_f: numpy.ndarray | None = None, pred_f: numpy.ndarray | None = None, lab_v: numpy.ndarray | None = None, pred_v: numpy.ndarray | None = None)[source]

Bases: object

atom_numb: int = 0

lab_e: ndarray | None = None

lab_f: ndarray | None = None

lab_v: ndarray | None = None

mae_e: float = 0

mae_e_atom: float = 0

mae_f: float = 0

mae_v: float = 0

name: str = 'default_system'

numb_frame: int = 0

pred_e: ndarray | None = None

pred_f: ndarray | None = None

pred_v: ndarray | None = None

report()[source]

rmse_e: float = 0

rmse_e_atom: float = 0

rmse_f: float = 0

rmse_v: float = 0

system: System | None = None

class pfd.exploration.inference.eval_model.TestReports(name: str = 'default_reports')[source]