`autopycoin.layers`.BaseLayer#

class autopycoin.layers.BaseLayer(*args, **kwargs)[source]#

Base layer which defines pre/post-processing methods to override.

This layer aims to be inherited and brings four functionality. - preprocessing : Preprocess the inputs data - post_processing : Preprocess the outputs data - init_params : initialize parameters before build method This three wrappers have to be overriden - Typing check.

checks(inputs_shape: Union[tensorflow.python.framework.tensor_shape.TensorShape, List[tensorflow.python.framework.tensor_shape.TensorShape]], **kwargs: dict) → None[source]#: Public API to initialize parameters before build method.

init_params(inputs_shape: Union[tensorflow.python.framework.tensor_shape.TensorShape, List[tensorflow.python.framework.tensor_shape.TensorShape]], **kwargs: dict) → None[source]#: Public API to initialize parameters before build method.

post_processing(output: Union[List[Union[tensorflow.python.framework.ops.Tensor, autopycoin.extension_type.QuantileTensor, autopycoin.extension_type.UnivariateTensor]], tensorflow.python.framework.ops.Tensor, autopycoin.extension_type.QuantileTensor, autopycoin.extension_type.UnivariateTensor]) → None[source]#: Public API to apply post-processing logics to your outputs data.

preprocessing(inputs: Union[List[Union[tensorflow.python.framework.ops.Tensor, autopycoin.extension_type.QuantileTensor, autopycoin.extension_type.UnivariateTensor]], tensorflow.python.framework.ops.Tensor, autopycoin.extension_type.QuantileTensor, autopycoin.extension_type.UnivariateTensor]) → None[source]#: Public API to apply preprocessing logics to your inputs data.