logml.data.transformers.decomposition

Classes

BaseDecompositionTransformer(**kwargs)	The idea is to produce features on top of a given dataframe's columns.
NMFDecompositionTransformer(**kwargs)	Provides NMF decomposition functionality.
PCADecompositionTransformer(**kwargs)	Provides PCA decomposition functionality.

class logml.data.transformers.decomposition.BaseDecompositionTransformer(**kwargs)

Bases: logml.data.base.BaseTransformer

The idea is to produce features on top of a given dataframe’s columns. Two main options for decomposition are PCA and NMF.

CONFIG_CLASS

alias of logml.data.config.DecompositionTransformerParams

LABEL = None

fit(dataframe: pandas.core.frame.DataFrame, dataset_metadata: Optional[logml.data.metadata.DatasetMetadata] = None, **kwargs)

Fits the decomposer for all affected column.

transform(dataframe: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame

Applies transformations and returns the result dataframe.

update_metadata(dataset_metadata: Optional[logml.data.metadata.DatasetMetadata] = None, dataframe: Optional[pandas.core.frame.DataFrame] = None) → None

Update metadata according to the change made.

params: BaseTransformerParams

global_params: Dict

metadata_cfg: ModelingTaskSpec

affected_columns_: List[str]

class logml.data.transformers.decomposition.PCADecompositionTransformer(**kwargs)

Bases: logml.data.transformers.decomposition.BaseDecompositionTransformer

Provides PCA decomposition functionality.

LABEL = 'pca'

params: BaseTransformerParams

global_params: Dict

metadata_cfg: ModelingTaskSpec

affected_columns_: List[str]

new_columns_: List[Tuple[str, np.dtype]]

class logml.data.transformers.decomposition.NMFDecompositionTransformer(**kwargs)

Bases: logml.data.transformers.decomposition.BaseDecompositionTransformer

Provides NMF decomposition functionality.

LABEL = 'nmf'

params: BaseTransformerParams

global_params: Dict

metadata_cfg: ModelingTaskSpec

affected_columns_: List[str]

new_columns_: List[Tuple[str, np.dtype]]