tabensemb.data.AbstractProcessor

class tabensemb.data.AbstractProcessor(**kwargs)

Bases: AbstractDataStep

The base class for data-processors that change the content of the tabular dataset. The class is only directly used for those who reduce the number of data points.

Notes

If any attribute of the datamodule is set by the processor in _fit_transform(), the processor is responsible for restoring the set attribute when _transform is called. For instance, in the wrapper methods fit_transform() and transform(), we have implemented recording feature names and restoring them.

Methods

__init__(**kwargs)

fit_transform(input_data, datamodule)	Record feature names in the datamodule, fit the processor, and call _fit_transform() to transform the input data.
transform(input_data, datamodule)	Restore feature names in datamodule using recorded features and call _transform() to transform the input data.
_fit_transform(data, datamodule)
_transform(data, datamodule)