FormatPreservingTransformer#
synthesized.privacy.FormatPreservingTransformer
- class FormatPreservingTransformer(name, pattern, seed=None)#
Consistently transforms by generating a format preserved hash.
Examples
“AAPL” -> “LAN” “AMZN” -> “ORM” “GOOG” -> “ELD” “AAPL” -> “LAN”
- name#
the data frame column to transform.
- Type
str
- pattern#
regex pattern of the string to be generated.
- Type
str
- seed#
a random seed for the transformer. Optional.
- Type
Optional[int]
Methods
fit
(df)Fits the given dataframe to the transformer
fit_transform
(df)- rtype
DataFrame
from_meta
(meta)Construct a Transformer from a transformer class name and a meta.
from_name_and_meta
(class_name, meta)Construct a Transformer from a transformer class name and a meta.
get_registry
()- rtype
Dict
[str
,Type
[TypeVar
(TransformerType
, bound=Transformer
)]]
inverse_transform
(df, **kwargs)Inverse transforms the transformed dataframe to give the original dataset
is_fitted
()- rtype
bool
transform
(df, **kwargs)Transforms the given dataframe using fitted transformer
Attributes
in_columns
- rtype
List
[str
]
out_columns
- rtype
List
[str
]
- fit(df)#
Fits the given dataframe to the transformer
- Parameters
df (
DataFrame
) – Dataset to fit- Return type
- Returns
self
- classmethod from_meta(meta)#
Construct a Transformer from a transformer class name and a meta. This is an abstract class, must be implemented in each transformer subclass.
- Return type
- classmethod from_name_and_meta(class_name, meta)#
Construct a Transformer from a transformer class name and a meta.
See also
Transformer.from_meta: construct a Transformer from a meta
- Return type
Transformer
- inverse_transform(df, **kwargs)#
Inverse transforms the transformed dataframe to give the original dataset
- Parameters
df (
DataFrame
) – Transformed dataset- Return type
DataFrame
- Returns
Original dataset
- transform(df, **kwargs)#
Transforms the given dataframe using fitted transformer
- Parameters
df (
DataFrame
) – Dataset to transform- Return type
DataFrame
- Returns
Transformed dataset