logml.data.transformers.randomization

Classes

`AddRandomColumnsTransformer`(params[, ...])	For a given set of target columns, creates additional 'randomized' columns for later significance assessment.
`ShuffleTransformer`(params[, metadata_cfg, ...])	Randomly permutes a given dataset's rows.

class logml.data.transformers.randomization.ShuffleTransformer(params: logml.data.config.BaseTransformerParams, metadata_cfg: logml.configuration.modeling.ModelingTaskSpec = None, cfg: GlobalConfig = None, global_params: Dict = None, logger=None)

Randomly permutes a given dataset’s rows.

fit(dataframe: pandas.core.frame.DataFrame, dataset_metadata: Optional[logml.data.metadata.DatasetMetadata] = None, **kwargs): Operates on rows, so no need to filter columns.

transform(dataframe: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame: Applies transformations and returns the result dataframe.

class logml.data.transformers.randomization.AddRandomColumnsTransformer(params: logml.data.config.BaseTransformerParams, metadata_cfg: logml.configuration.modeling.ModelingTaskSpec = None, cfg: GlobalConfig = None, global_params: Dict = None, logger=None)

For a given set of target columns, creates additional ‘randomized’ columns for later significance assessment.

CONFIG_CLASS: alias of logml.data.config.AddRandomColumnsTransformerParams

transform(dataframe: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame: Adds randomized columns.

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