openasce.extension package

Subpackages

• openasce.extension.debias package
  • Subpackages
    • openasce.extension.debias.common package
      • Submodules
      • openasce.extension.debias.common.utils module
  • Submodules
  • openasce.extension.debias.cfr module
  • openasce.extension.debias.debias_ips module
  • openasce.extension.debias.dice module
  • openasce.extension.debias.dmbr module
  • openasce.extension.debias.doubly_robust module
  • openasce.extension.debias.fairco module
  • openasce.extension.debias.ipw module
  • openasce.extension.debias.macr module
  • openasce.extension.debias.pda module

Submodules

openasce.extension.debias_model module

class openasce.extension.debias_model.CausalDebiasModel

Bases: Runtime

Debias Inference Class

Base class of the causal debias

Attributes:

__annotations__ = {}

__doc__ = 'Debias Inference Class\n\n    Base class of the causal debias\n\n    Attributes:\n\n    '

__init__() → None

__module__ = 'openasce.extension.debias_model'

_call(*, x: ndarray, y: ndarray, c: Dict[str, ndarray], training: bool) → Union[None, Dict[str, ndarray]]

The derived class should override this method to train the model using loss_object and optimizer or predict on the samples.

Parameters

• x – one batch of features

• y – one batch of labels

• c – one batch for each concerned columns of the samples, e.g. {‘weight’: Iterable[np.ndarray]}

• training – True means training and False for predict

Returns

None for training and Dict for predict

Raises

StopIteration – The process can be finished

_generator(**kwargs)

main loop

_predict_loop()

main loop for prediction

_train_loop(*, num_epochs, **kwargs)

main loop for train

fit(*, X: Iterable[ndarray] = None, Y: Iterable[ndarray] = None, C: Dict[str, Iterable[ndarray]] = None, Z: Iterable[Tuple[ndarray, ndarray, Dict[str, ndarray]]] = None, num_epochs: int = 1, **kwargs) → None

Feed the sample data and train the model on the samples.

Parameters

• X – Features of the samples.

• Y – Outcomes of the samples.

• C – Other concerned columns of the samples, e.g. {‘weight’: Iterable[np.ndarray]}

• Z – The iterable object returning (a batch of X, a batch of Y, a batch of C) if having

• num_epochs – number of the train epoch

Returns

None

get_result()

Get the predict result

Arguments:

Returns

predict result

predict(*, X: Iterable[ndarray] = None, C: Dict[str, Iterable[ndarray]] = None, Z: Iterable[Tuple[ndarray, ndarray, Dict[str, ndarray]]] = None, **kwargs) → None

Feed the sample data

Estimate the effect on the samples, and get_result method can be used to get the result of prediction

Parameters

• X – Features of the samples.

• C – Other concerned columns of the samples, e.g. {‘weight’: Iterable[np.ndarray]}

• Z – The iterable object returning (a batch of X, a batch of Y, a batch of C) if having

Returns

None

openasce.extension.debias_model_test module

class openasce.extension.debias_model_test.MockDebiasModel

Bases: CausalDebiasModel

__annotations__ = {}

__doc__ = None

__init__() → None

__module__ = 'openasce.extension.debias_model_test'

_call(*, x: ndarray, y: ndarray, c: Dict[str, ndarray], training: bool) → None

The derived class should override this method to train the model using loss_object and optimizer or predict on the samples.

Parameters

• x – one batch of features

• y – one batch of labels

• c – one batch for each concerned columns of the samples, e.g. {‘weight’: Iterable[np.ndarray]}

• training – True means training and False for predict

Returns

None for training and Dict for predict

Raises

StopIteration – The process can be finished

openasce.extension.debias_model_test.get_iris_dataset()