ballet.transformer module

class ballet.transformer.ConversionApproach(name, convert, caught, isokay)

Bases: tuple

caught: Collection[Type[Exception]]

Alias for field number 2

convert: Callable

Alias for field number 1

isokay: Optional[Callable[[Exception], bool]]

Opportunity to catch other exceptions that match a condition

name: str

Alias for field number 0

class ballet.transformer.DelegatingRobustTransformer(transformer)

Bases: ballet.util.DeepcopyMixin, ballet.eng.base.BaseTransformer

Robust transformer that delegates to underlying transformer

This transformer is robust against different typed and shaped input data. It tries a variety of input data conversion approaches and passes the result to the underlying transformer, using the first approach that works.

Parameters

transformer (BaseTransformer) – a transformer object with fit and transform methods

Raises

UnsuccessfulInputConversionError – If none of the conversion approaches work.

CONVERSION_APPROACHES = [ConversionApproach(name='identity', convert=<function identity>, caught=(<class 'ValueError'>, <class 'TypeError'>), isokay=<function ConversionApproach.<lambda>>), ConversionApproach(name='series', convert=<class 'pandas.core.series.Series'>, caught=(<class 'ValueError'>, <class 'TypeError'>), isokay=<function catch_bare_exception_sanitize_array>), ConversionApproach(name='dataframe', convert=<class 'pandas.core.frame.DataFrame'>, caught=(<class 'ValueError'>, <class 'TypeError'>), isokay=<function ConversionApproach.<lambda>>), ConversionApproach(name='array', convert=<function asarray>, caught=(<class 'ValueError'>, <class 'TypeError'>), isokay=<function ConversionApproach.<lambda>>), ConversionApproach(name='asarray2d', convert=<function asarray2d>, caught=(), isokay=<function ConversionApproach.<lambda>>)]

DEFAULT_CAUGHT = (<class 'ValueError'>, <class 'TypeError'>)

fit(X, y=None, **kwargs)

get_params(deep=False)

Get parameters for this estimator.

Parameters

deep (bool, default=True) – If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns

params – Parameter names mapped to their values.

Return type

dict

transform(X, y=None, **kwargs)

ballet.transformer.catch_bare_exception_sanitize_array(e)

See pandas-dev/pandas#35744

ballet.transformer.desugar_transformer(transformer)

Replace transformer syntactic sugar with actual transformer

The following syntactic sugar is supported: - None is replaced with an IdentityTransformer - a callable (function or lambda) is replaced with a FunctionTransformer

that wraps that callable

• a tuple (input, transformer) is replaced with a SubsetTransformer

Return type

BaseTransformer

ballet.transformer.get_transformer_primitives(transformer)

Get the primitives used in this transformer or pipeline

The primitives are just the class names underlying the transformer or pipeline.

Return type

List[str]

ballet.transformer.make_robust_transformer(transformer)

Convert to robust transformer or pipeline

Convert to either a single DelegatingRobustTransformer or a TransformerPipeline where each transformer in the pipeline is a DelegatingRobustTransformer.

Return type

Union[TransformerPipeline, DelegatingRobustTransformer]

ballet.transformer.make_transformer_pipeline(steps)

Construct a TransformerPipeline from the given estimators.

Source: sklearn_pandas.cont_method

Return type

TransformerPipeline