# Copyright (C) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License. See LICENSE in project root for information.
import sys
if sys.version >= '3':
basestring = str
from pyspark.ml.param.shared import *
from pyspark import keyword_only
from pyspark.ml.util import JavaMLReadable, JavaMLWritable
from mmlspark.core.serialize.java_params_patch import *
from pyspark.ml.wrapper import JavaTransformer, JavaEstimator, JavaModel
from pyspark.ml.common import inherit_doc
from mmlspark.core.schema.Utils import *
[docs]@inherit_doc
class CleanMissingData(ComplexParamsMixin, JavaMLReadable, JavaMLWritable, JavaEstimator):
"""
Args:
cleaningMode (str): Cleaning mode (default: Mean)
customValue (str): Custom value for replacement
inputCols (list): The names of the input columns
outputCols (list): The names of the output columns
"""
@keyword_only
def __init__(self, cleaningMode="Mean", customValue=None, inputCols=None, outputCols=None):
super(CleanMissingData, self).__init__()
self._java_obj = self._new_java_obj("com.microsoft.ml.spark.featurize.CleanMissingData")
self.cleaningMode = Param(self, "cleaningMode", "cleaningMode: Cleaning mode (default: Mean)")
self._setDefault(cleaningMode="Mean")
self.customValue = Param(self, "customValue", "customValue: Custom value for replacement")
self.inputCols = Param(self, "inputCols", "inputCols: The names of the input columns")
self.outputCols = Param(self, "outputCols", "outputCols: The names of the output columns")
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
[docs] @keyword_only
def setParams(self, cleaningMode="Mean", customValue=None, inputCols=None, outputCols=None):
"""
Set the (keyword only) parameters
Args:
cleaningMode (str): Cleaning mode (default: Mean)
customValue (str): Custom value for replacement
inputCols (list): The names of the input columns
outputCols (list): The names of the output columns
"""
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
return self._set(**kwargs)
[docs] def setCleaningMode(self, value):
"""
Args:
cleaningMode (str): Cleaning mode (default: Mean)
"""
self._set(cleaningMode=value)
return self
[docs] def getCleaningMode(self):
"""
Returns:
str: Cleaning mode (default: Mean)
"""
return self.getOrDefault(self.cleaningMode)
[docs] def setCustomValue(self, value):
"""
Args:
customValue (str): Custom value for replacement
"""
self._set(customValue=value)
return self
[docs] def getCustomValue(self):
"""
Returns:
str: Custom value for replacement
"""
return self.getOrDefault(self.customValue)
[docs] def setOutputCols(self, value):
"""
Args:
outputCols (list): The names of the output columns
"""
self._set(outputCols=value)
return self
[docs] def getOutputCols(self):
"""
Returns:
list: The names of the output columns
"""
return self.getOrDefault(self.outputCols)
[docs] @classmethod
def read(cls):
""" Returns an MLReader instance for this class. """
return JavaMMLReader(cls)
[docs] @staticmethod
def getJavaPackage():
""" Returns package name String. """
return "com.microsoft.ml.spark.featurize.CleanMissingData"
@staticmethod
def _from_java(java_stage):
module_name=CleanMissingData.__module__
module_name=module_name.rsplit(".", 1)[0] + ".CleanMissingData"
return from_java(java_stage, module_name)
def _create_model(self, java_model):
return CleanMissingDataModel(java_model)
[docs]class CleanMissingDataModel(ComplexParamsMixin, JavaModel, JavaMLWritable, JavaMLReadable):
"""
Model fitted by :class:`CleanMissingData`.
This class is left empty on purpose.
All necessary methods are exposed through inheritance.
"""
[docs] @classmethod
def read(cls):
""" Returns an MLReader instance for this class. """
return JavaMMLReader(cls)
[docs] @staticmethod
def getJavaPackage():
""" Returns package name String. """
return "com.microsoft.ml.spark.featurize.CleanMissingDataModel"
@staticmethod
def _from_java(java_stage):
module_name=CleanMissingDataModel.__module__
module_name=module_name.rsplit(".", 1)[0] + ".CleanMissingDataModel"
return from_java(java_stage, module_name)