Feature encoding

class: scorecardbundle.feature_encoding.WOE.WOE_Encoder

Perform WOE transformation for features and calculate the information value (IV) of features with reference to the target variable y.

Parameters

epslon: float, optional(default=1e-10)
        Replace 0 with a very small number during division 
        or logrithm to avoid infinite value.       

output_dataframe: boolean, optional(default=False)
        if output_dataframe is set to True. The transform() function will
        return pandas.DataFrame. If it is set to False, the output will
        be numpy ndarray.
  

Attributes

iv_: a dictionary that contains feature names and their IV

result_dict_: a dictionary that contains feature names and 
    their WOE result tuple. Each WOE result tuple contains the
    woe value dictionary and the iv for the feature.
  

Methods

fit(X, y): 
        fit the WOE transformation to the feature.

transform(X): 
        transform the feature using the WOE fitted.

fit_transform(X, y): 
        fit the WOE transformation to the feature and transform it.         
  