DataTransform

Data Transform is the most basic component of Fate Algorithm module. It transforms the input Table to a Table whose values are Instance Object defined here, and what's more, the transformed table is the input data format of all other algorithm module, such as intersect、 homo LR and hetero LR、SecureBoost and so on.

Data IO module accepts the following input data format and transforms them to desired output Table.

• dense input format
  input Table's value item is a list of single element, e.g. :

  1.0,2.0,3.0,4.5
  1.1,2.1,3.4,1.3
  2.4,6.3,1.5,9.0
  

• svm-light input format
  first item of input Table's value is label, following by a list of complex "feature_id:value" items, e.g. :

  1 1:0.5 2:0.6
  0 1:0.7 3:0.8 5:0.2
  

• tag input format
  the input Table's value is a list of tag, data transform module first aggregates all tags occurred in input table, then changes all input line to one-hot representation in sorting the occurred tags by lexicographic order, e.g. assume values is :

  a c
  a b d
  

  after processing, the new values became: :

  1 0 1 0
  1 1 0 1
  

• :tag:value input format: the input Table's value is a list of , like a mixed svm-light and tag input-format. data transform module first aggregates all tags occurred in input table, then changes all input line to one-hot representation in sorting the occurred tags by lexicographic order, then fill the occur item with value. e.g. assume values is

  a:0.2 c:1.5
  a:0.3 b:0.6 d:0.7
  

  after processing, the new values became: :

  0.2 0 0.5 0
  0.3 0.6 0 0.7
  

Param

data_transform_param

Classes

DataTransformParam(input_format='dense', delimitor=',', data_type='float64', exclusive_data_type=None, tag_with_value=False, tag_value_delimitor=':', missing_fill=False, default_value=0, missing_fill_method=None, missing_impute=None, outlier_replace=False, outlier_replace_method=None, outlier_impute=None, outlier_replace_value=0, with_label=False, label_name='y', label_type='int', output_format='dense', need_run=True, with_match_id=False, match_id_name='', match_id_index=0)

Bases: BaseParam

Define data transform parameters that used in federated ml.

Parameters:

Name	Type	Description	Default
input_format		please have a look at this tutorial at "DataTransform" section of federatedml/util/README.md. Formally, dense input format data should be set to "dense", svm-light input format data should be set to "sparse", tag or tag:value input format data should be set to "tag". Note: in fate's version >= 1.9.0, this params can be used in uploading/binding data's meta	'dense'
delimitor	str	the delimitor of data input, default: ','	','
data_type	int	{'float64','float','int','int64','str','long'} the data type of data input	'float64'
exclusive_data_type	dict	the key of dict is col_name, the value is data_type, use to specified special data type of some features.	None
tag_with_value		use if input_format is 'tag', if tag_with_value is True, input column data format should be tag[delimitor]value, otherwise is tag only	False
tag_value_delimitor		use if input_format is 'tag' and 'tag_with_value' is True, delimitor of tag[delimitor]value column value.	':'
missing_fill	bool	need to fill missing value or not, accepted only True/False, default: False	False
default_value	None or object or list	the value to replace missing value. if None, it will use default value define in federatedml/feature/imputer.py, if single object, will fill missing value with this object, if list, it's length should be the sample of input data' feature dimension, means that if some column happens to have missing values, it will replace it the value by element in the identical position of this list.	0
missing_fill_method		the method to replace missing value, should be one of [None, 'min', 'max', 'mean', 'designated']	None
missing_impute		element of list can be any type, or auto generated if value is None, define which values to be consider as missing	None
outlier_replace		need to replace outlier value or not, accepted only True/False, default: True	False
outlier_replace_method		the method to replace missing value, should be one of [None, 'min', 'max', 'mean', 'designated']	None
outlier_impute		element of list can be any type, which values should be regard as missing value	None
outlier_replace_value		the value to replace outlier. if None, it will use default value define in federatedml/feature/imputer.py, if single object, will replace outlier with this object, if list, it's length should be the sample of input data' feature dimension, means that if some column happens to have outliers, it will replace it the value by element in the identical position of this list.	0
with_label	bool	True if input data consist of label, False otherwise. default: 'false' Note: in fate's version >= 1.9.0, this params can be used in uploading/binding data's meta	False
label_name	str	column_name of the column where label locates, only use in dense-inputformat. default: 'y'	'y'
label_type		use when with_label is True	'int','int64','float','float64','long','str'
output_format		output format	'dense'
with_match_id		True if dataset has match_id, default: False Note: in fate's version >= 1.9.0, this params can be used in uploading/binding data's meta	False
match_id_name		Valid if input_format is "dense", and multiple columns are considered as match_ids, the name of match_id to be used in current job Note: in fate's version >= 1.9.0, this params can be used in uploading/binding data's meta	''
match_id_index		Valid if input_format is "tag" or "sparse", and multiple columns are considered as match_ids, the index of match_id, default: 0 This param works only when data meta has been set with uploading/binding.	0

Source code in python/federatedml/param/data_transform_param.py

def __init__(self, input_format="dense", delimitor=',', data_type='float64',
             exclusive_data_type=None,
             tag_with_value=False, tag_value_delimitor=":",
             missing_fill=False, default_value=0, missing_fill_method=None,
             missing_impute=None, outlier_replace=False, outlier_replace_method=None,
             outlier_impute=None, outlier_replace_value=0,
             with_label=False, label_name='y',
             label_type='int', output_format='dense', need_run=True,
             with_match_id=False, match_id_name='', match_id_index=0):
    self.input_format = input_format
    self.delimitor = delimitor
    self.data_type = data_type
    self.exclusive_data_type = exclusive_data_type
    self.tag_with_value = tag_with_value
    self.tag_value_delimitor = tag_value_delimitor
    self.missing_fill = missing_fill
    self.default_value = default_value
    self.missing_fill_method = missing_fill_method
    self.missing_impute = missing_impute
    self.outlier_replace = outlier_replace
    self.outlier_replace_method = outlier_replace_method
    self.outlier_impute = outlier_impute
    self.outlier_replace_value = outlier_replace_value
    self.with_label = with_label
    self.label_name = label_name
    self.label_type = label_type
    self.output_format = output_format
    self.need_run = need_run
    self.with_match_id = with_match_id
    self.match_id_name = match_id_name
    self.match_id_index = match_id_index

Attributes

input_format = input_format instance-attribute

delimitor = delimitor instance-attribute

data_type = data_type instance-attribute

exclusive_data_type = exclusive_data_type instance-attribute

tag_with_value = tag_with_value instance-attribute

tag_value_delimitor = tag_value_delimitor instance-attribute

missing_fill = missing_fill instance-attribute

default_value = default_value instance-attribute

missing_fill_method = missing_fill_method instance-attribute

missing_impute = missing_impute instance-attribute

outlier_replace = outlier_replace instance-attribute

outlier_replace_method = outlier_replace_method instance-attribute

outlier_impute = outlier_impute instance-attribute

outlier_replace_value = outlier_replace_value instance-attribute

with_label = with_label instance-attribute

label_name = label_name instance-attribute

label_type = label_type instance-attribute

output_format = output_format instance-attribute

need_run = need_run instance-attribute

with_match_id = with_match_id instance-attribute

match_id_name = match_id_name instance-attribute

match_id_index = match_id_index instance-attribute

Functions

check()

Source code in python/federatedml/param/data_transform_param.py

def check(self):

    descr = "data_transform param's"

    self.input_format = self.check_and_change_lower(self.input_format,
                                                    ["dense", "sparse", "tag"],
                                                    descr)

    self.output_format = self.check_and_change_lower(self.output_format,
                                                     ["dense", "sparse"],
                                                     descr)

    self.data_type = self.check_and_change_lower(self.data_type,
                                                 ["int", "int64", "float", "float64", "str", "long"],
                                                 descr)

    if type(self.missing_fill).__name__ != 'bool':
        raise ValueError("data_transform param's missing_fill {} not supported".format(self.missing_fill))

    if self.missing_fill_method is not None:
        self.missing_fill_method = self.check_and_change_lower(self.missing_fill_method,
                                                               ['min', 'max', 'mean', 'designated'],
                                                               descr)

    if self.outlier_replace_method is not None:
        self.outlier_replace_method = self.check_and_change_lower(self.outlier_replace_method,
                                                                  ['min', 'max', 'mean', 'designated'],
                                                                  descr)

    if type(self.with_label).__name__ != 'bool':
        raise ValueError("data_transform param's with_label {} not supported".format(self.with_label))

    if self.with_label:
        if not isinstance(self.label_name, str):
            raise ValueError("data transform param's label_name {} should be str".format(self.label_name))

        self.label_type = self.check_and_change_lower(self.label_type,
                                                      ["int", "int64", "float", "float64", "str", "long"],
                                                      descr)

    if self.exclusive_data_type is not None and not isinstance(self.exclusive_data_type, dict):
        raise ValueError("exclusive_data_type is should be None or a dict")

    if not isinstance(self.with_match_id, bool):
        raise ValueError("with_match_id should be boolean variable, but {} find".format(self.with_match_id))

    if not isinstance(self.match_id_index, int) or self.match_id_index < 0:
        raise ValueError("match_id_index should be non negative integer")

    if self.match_id_name is not None and not isinstance(self.match_id_name, str):
        raise ValueError("match_id_name should be str")

    return True

Other Features of DataTransform

• Missing value impute, provides $$"mean", "designated", "min", "max"$$ methods to impute missing value
• Outlier value replace, also provides several outlier replace method like missing value impute.
• Parameters of data meta should be set when uploading or binding data since FATE-v1.9.0, refer to upload guides please.

---

最后更新: 2022-08-25