dabl.EasyPreprocessor

class dabl.EasyPreprocessor(scale=True, force_imputation=True, verbose=0, types=None)[source]

A simple preprocessor.

Detects variable types, encodes everything as floats for use with sklearn.

Applies one-hot encoding, missing value imputation and scaling.

Parameters:
scaleboolean, default=True

Whether to scale continuous data.

force_imputationbool, default=True

Whether to create imputers even if no training data is missing.

verboseint, default=0

Control output verbosity.

Attributes:
ct_ColumnTransformer

Main container for all transformations.

columns_pandas columns

Columns of training data.

dtypes_Series of dtypes

Dtypes of training data columns.

types_something

Inferred input types.

__init__(scale=True, force_imputation=True, verbose=0, types=None)[source]