Some classes to make life bit easy
Contains some easy but required hacks
IndexMapper
IndexMapper (categorical_columns:Union[str,list], verbose:bool=False)
A multi column categorical labeller :param categorical_columns: :param verbose: Verbosity
| Type | Default | Details | |
|---|---|---|---|
| categorical_columns | Union | Columns to create Label for | |
| verbose | bool | False | Verbosity |
VariableTransformation
VariableTransformation (variables:Union[str,list], strategy='zscore', custom_func=None)
Variable Transformation for all features independently
| Type | Default | Details | |
|---|---|---|---|
| variables | Union | Variables for which we need to perform the transformation | |
| strategy | str | zscore | strategy on how to transform variables ‘zscore’,‘min-max’,‘median-iqr’, mean-iqr’,‘median-qd’, ‘median-std’,‘median-range’,‘mean-range’,‘min-qd’,‘min-iqr’, ‘min’,‘max’,‘mean’,‘median’,‘std’,‘iqr’,‘qd’,‘range’,‘custom’ |
| custom_func | NoneType | None | The function should take 4 arguments custom_func(value,group,groups_estimates,inverse) |
GroupedVariableTransformation
GroupedVariableTransformation (key, target, monotone_constraints=None, strategy='zscore', custom_func=None)
Variable Transformation at group level
| Type | Default | Details | |
|---|---|---|---|
| key | Grouping / ID column against which aggregation would happen | ||
| target | Target column name which needs to be used for creating aggregation | ||
| monotone_constraints | NoneType | None | Monotonic Constraints |
| strategy | str | zscore | strategy on how to transform variables ‘zscore’,‘min-max’,‘median-iqr’, mean-iqr’,‘median-qd’, ‘median-std’,‘median-range’,‘mean-range’,‘min-qd’,‘min-iqr’, ‘min’,‘max’,‘mean’,‘median’,‘std’,‘iqr’,‘qd’,‘range’,‘custom’ |
| custom_func | NoneType | None | A custom function of form that take 4 arguments : custom_func(value,group,groups_estimates,inverse) |