maml.base package

Define abstract base classes.

class maml.base.BaseDataSource()

Bases: object

Abstract base class for a data source.

abstract get(*args, **kwargs)

Get data from sources.

class maml.base.BaseDescriber(**kwargs)

Bases: BaseEstimator, TransformerMixin, MSONable

Base class for a Describer. A describers converts an object to a descriptor, typically a numerical representation useful for machine learning. The output for the describers can be a single DataFrame/numpy.ndarray or a list of DataFrame/numpy.ndarray.

abc_impl( = <_abc.abc_data object )

_is_multi_output()

sklearn_auto_wrap_output_keys( = {‘transform’_ )

clear_cache()

Clear cache.

property feature_dim()

Feature dimension, useful when certain models need to specify the feature dimension, e.g., MLP models.

fit(x: Any, y: Any | None = None)

Place holder for fit API.

• Parameters

  • x – Any inputs

  • y – Any outputs

Returns: self

transform(objs: list[Any])

Transform a list of objs. If the return data is DataFrame, use df.xs(index, level=’input_index’) to get the result for the i-th object.

• Parameters objs (list) – A list of objects.

• Returns One or a list of pandas data frame/numpy ndarray

transform_one(obj: Any)

Transform an object.

class maml.base.BaseModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: object

Abstract Base class for a Model. Basically, it usually wraps around a deep learning package, e.g., the Sequential Model in Keras, but provides for transparent conversion of arbitrary input and outputs.

_predict(features: ndarray, **kwargs)

Predict the values given a set of inputs based on fitted models.

• Parameters features (np.ndarray) – array-like input features.

• Returns List of output objects.

fit(features: list | np.ndarray, targets: list | np.ndarray | None = None, val_features: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

• Parameters

  • features (list* or *np.ndarray) – Numerical input feature list or numpy array with dim (m, n) where m is the number of data and n is the feature dimension.

  • targets (list* or *np.ndarray) – Numerical output target list, or numpy array with dim (m, ).

  • val_features (list* or *np.ndarray) – validation features

  • val_targets (list* or *np.ndarray) – validation targets.

• Returns self

predict_objs(objs: list | np.ndarray)

Predict the values given a set of objects. Usually Pymatgen

Structure objects.

train(objs: list | np.ndarray, targets: list | np.ndarray | None = None, val_objs: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

Train the models from object, target pairs.

• Parameters

  • objs (list* of *objects) – List of objects

  • targets (list) – list of float or np.ndarray

  • val_objs (list* of *objects) – list of validation objects

  • val_targets (list) – list of validation targets

  • **kwargs –

Returns: self

class maml.base.DummyDescriber(**kwargs)

Bases: BaseDescriber

Dummy Describer that does nothing.

abc_impl( = <_abc.abc_data object )

sklearn_auto_wrap_output_keys( = {‘transform’_ )

transform_one(obj: Any)

Does nothing but return the original features.

• Parameters obj – Any inputs

Returns: Any outputs

class maml.base.KerasModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: BaseModel, KerasMixin

MAML models with keras models as estimators.

static _get_validation_data(val_features, val_targets, **val_kwargs)

construct validation data, the default is just returning a list of val_features and val_targets.

fit(features: list | np.ndarray, targets: list | np.ndarray | None = None, val_features: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

• Parameters

  • features (list* or *np.ndarray) – Numerical input feature list or numpy array with dim (m, n) where m is the number of data and n is the feature dimension.

  • targets (list* or *np.ndarray) – Numerical output target list, or numpy array with dim (m, ).

  • val_features (list* or *np.ndarray) – validation features

  • val_targets (list* or *np.ndarray) – validation targets.

• Returns self

class maml.base.SKLModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: BaseModel, SklearnMixin

MAML models with sklearn models as estimator.

class maml.base.SequentialDescriber(describers: list, **kwargs)

Bases: Pipeline

A thin wrapper of sklearn Pipeline.

abc_impl( = <_abc.abc_data object )

steps(: List[Any )

class maml.base.TargetScalerMixin(*args, **kwargs)

Bases: object

Mixin class for models with target scaler.

predict_objs(objs: list | np.ndarray)

Predict the values given a set of objects. Usually Pymatgen

Structure objects.

train(objs: list | np.ndarray, targets: list | np.ndarray | None = None, val_objs: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

Train the models from object, target pairs.

• Parameters

  • objs (list* of *objects) – List of objects

  • targets (list) – list of float or np.ndarray

  • val_objs (list* of *objects) – list of validation objects

  • val_targets (list) – list of validation targets

  • **kwargs –

Returns: self

maml.base.describer_type(dtype: str)

Decorate to set describers class type.

• Parameters dtype (str) – describers type, e.g., site, composition, structure etc.

• Returns wrapped class

maml.base.get_feature_batch(fb_name: str | Callable | None = None)

Providing a feature batch name, returning the function callable :param fb_name: name of the feature batch function :type fb_name: str

Returns: callable feature batch function.

maml.base.is_keras_model(model: BaseModel)

Check whether the model is keras :param model: model :type model: BaseModel

Returns: bool.

maml.base.is_sklearn_model(model: BaseModel)

Check whether the model is sklearn :param model: model :type model: BaseModel

Returns: bool.

maml.base._data module

MAML data base classes.

class maml.base._data.BaseDataSource()

Bases: object

Abstract base class for a data source.

abstract get(*args, **kwargs)

Get data from sources.

maml.base._describer module

MAML describers base classes.

class maml.base._describer.BaseDescriber(**kwargs)

Bases: BaseEstimator, TransformerMixin, MSONable

Base class for a Describer. A describers converts an object to a descriptor, typically a numerical representation useful for machine learning. The output for the describers can be a single DataFrame/numpy.ndarray or a list of DataFrame/numpy.ndarray.

abc_impl( = <_abc.abc_data object )

_is_multi_output()

sklearn_auto_wrap_output_keys( = {‘transform’_ )

clear_cache()

Clear cache.

property feature_dim()

Feature dimension, useful when certain models need to specify the feature dimension, e.g., MLP models.

fit(x: Any, y: Any | None = None)

Place holder for fit API.

• Parameters

  • x – Any inputs

  • y – Any outputs

Returns: self

transform(objs: list[Any])

Transform a list of objs. If the return data is DataFrame, use df.xs(index, level=’input_index’) to get the result for the i-th object.

• Parameters objs (list) – A list of objects.

• Returns One or a list of pandas data frame/numpy ndarray

transform_one(obj: Any)

Transform an object.

class maml.base._describer.DummyDescriber(**kwargs)

Bases: BaseDescriber

Dummy Describer that does nothing.

abc_impl( = <_abc.abc_data object )

sklearn_auto_wrap_output_keys( = {‘transform’_ )

transform_one(obj: Any)

Does nothing but return the original features.

• Parameters obj – Any inputs

Returns: Any outputs

class maml.base._describer.SequentialDescriber(describers: list, **kwargs)

Bases: Pipeline

A thin wrapper of sklearn Pipeline.

abc_impl( = <_abc.abc_data object )

steps(: List[Any )

maml.base._describer._transform_one(describer: BaseDescriber, obj: Any)

A wrapper to make a pure function.

• Parameters describer (BaseDescriber) – a describers

• Returns np.ndarray

maml.base._describer.describer_type(dtype: str)

Decorate to set describers class type.

• Parameters dtype (str) – describers type, e.g., site, composition, structure etc.

• Returns wrapped class

maml.base._feature_batch module

Batch a list of features output by describers.transform method.

maml.base._feature_batch.get_feature_batch(fb_name: str | Callable | None = None)

Providing a feature batch name, returning the function callable :param fb_name: name of the feature batch function :type fb_name: str

Returns: callable feature batch function.

maml.base._feature_batch.no_action(features: list[Any])

Return original feature lists.

maml.base._feature_batch.pandas_concat(features: list[pandas.core.frame.DataFrame])

Concatenate a list of pandas dataframe into a single one :param features: list of pandas dataframe. :type features: list

Returns: concatenated pandas dataframe

maml.base._feature_batch.stack_first_dim(features: list[numpy.ndarray])

Stack the first dimension. If the original features are a list of nxm array, the stacked features will be lxnxm, where l is the number of entries in the list :param features: list of numpy array features. :type features: list

Returns: stacked features

maml.base._feature_batch.stack_padded(features: list[numpy.ndarray])

Stack the first dimension. If the original features are a list of nxm array, the stacked features will be lxnxm, where l is the number of entries in the list :param features: list of numpy array features. :type features: list

Returns: stacked features

maml.base._mixin module

Model mixins.

class maml.base._mixin.TargetScalerMixin(*args, **kwargs)

Bases: object

Mixin class for models with target scaler.

predict_objs(objs: list | np.ndarray)

Predict the values given a set of objects. Usually Pymatgen

Structure objects.

train(objs: list | np.ndarray, targets: list | np.ndarray | None = None, val_objs: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

Train the models from object, target pairs.

• Parameters

  • objs (list* of *objects) – List of objects

  • targets (list) – list of float or np.ndarray

  • val_objs (list* of *objects) – list of validation objects

  • val_targets (list) – list of validation targets

  • **kwargs –

Returns: self

maml.base._model module

MAML models base classes.

class maml.base._model.BaseModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: object

Abstract Base class for a Model. Basically, it usually wraps around a deep learning package, e.g., the Sequential Model in Keras, but provides for transparent conversion of arbitrary input and outputs.

_predict(features: ndarray, **kwargs)

Predict the values given a set of inputs based on fitted models.

• Parameters features (np.ndarray) – array-like input features.

• Returns List of output objects.

fit(features: list | np.ndarray, targets: list | np.ndarray | None = None, val_features: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

• Parameters

  • features (list* or *np.ndarray) – Numerical input feature list or numpy array with dim (m, n) where m is the number of data and n is the feature dimension.

  • targets (list* or *np.ndarray) – Numerical output target list, or numpy array with dim (m, ).

  • val_features (list* or *np.ndarray) – validation features

  • val_targets (list* or *np.ndarray) – validation targets.

• Returns self

predict_objs(objs: list | np.ndarray)

Predict the values given a set of objects. Usually Pymatgen

Structure objects.

train(objs: list | np.ndarray, targets: list | np.ndarray | None = None, val_objs: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

Train the models from object, target pairs.

• Parameters

  • objs (list* of *objects) – List of objects

  • targets (list) – list of float or np.ndarray

  • val_objs (list* of *objects) – list of validation objects

  • val_targets (list) – list of validation targets

  • **kwargs –

Returns: self

class maml.base._model.KerasMixin()

Bases: object

keras models mixin with save and load functionality.

evaluate(eval_objs: list | np.ndarray, eval_targets: list | np.ndarray, is_feature: bool = False)

Evaluate objs, targets.

• Parameters

  • eval_objs (list) – objs for evaluation

  • eval_targets (list) – target list for the corresponding objects

  • is_feature (bool) – whether the input x is feature matrix

  • metric (callable) – metric for evaluation

classmethod from_file(filename: str, **kwargs)

Load the models from file :param filename: filename :type filename: str :param **kwargs:

Returns: object instance

static get_input_dim(describer: BaseDescriber | None = None, input_dim: int | None = None)

Get feature dimension/input_dim from describers or input_dim.

• Parameters

  • describer (Describer) – describers

  • input_dim (int) – optional input dim int

load(filename: str, custom_objects: list | None = None)

Load models parameters from filename :param filename: models file name. :type filename: str

Returns: None

save(filename: str)

Save the models and describers.

• Parameters filename (str) – filename for save

class maml.base._model.KerasModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: BaseModel, KerasMixin

MAML models with keras models as estimators.

static _get_validation_data(val_features, val_targets, **val_kwargs)

construct validation data, the default is just returning a list of val_features and val_targets.

fit(features: list | np.ndarray, targets: list | np.ndarray | None = None, val_features: list | np.ndarray | None = None, val_targets: list | np.ndarray | None = None, **kwargs)

• Parameters

  • features (list* or *np.ndarray) – Numerical input feature list or numpy array with dim (m, n) where m is the number of data and n is the feature dimension.

  • targets (list* or *np.ndarray) – Numerical output target list, or numpy array with dim (m, ).

  • val_features (list* or *np.ndarray) – validation features

  • val_targets (list* or *np.ndarray) – validation targets.

• Returns self

class maml.base._model.SKLModel(model, describer: BaseDescriber | None = None, **kwargs)

Bases: BaseModel, SklearnMixin

MAML models with sklearn models as estimator.

class maml.base._model.SklearnMixin()

Bases: object

Sklearn models save and load functionality.

evaluate(eval_objs: list | np.ndarray, eval_targets: list | np.ndarray, is_feature: bool = False, metric: str | Callable | None = None)

Evaluate objs, targets.

• Parameters

  • eval_objs (list) – objs for evaluation

  • eval_targets (list) – target list for the corresponding objects

  • is_feature (bool) – whether the input x is feature matrix

  • metric (callable) – metric for evaluation

classmethod from_file(filename: str, **kwargs)

Load the models from file :param filename: filename :type filename: str :param **kwargs:

Returns: object instance

load(filename: str)

Load models parameters from filename :param filename: models file name. :type filename: str

Returns: None

save(filename: str)

Save the models and describers.

• Parameters filename (str) – filename for save

maml.base._model.is_keras_model(model: BaseModel)

Check whether the model is keras :param model: model :type model: BaseModel

Returns: bool.

maml.base._model.is_sklearn_model(model: BaseModel)

Check whether the model is sklearn :param model: model :type model: BaseModel

Returns: bool.