# 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 *
from mmlspark.core.schema.TypeConversionUtils import generateTypeConverter, complexTypeConverter
[docs]@inherit_doc
class MultiColumnAdapter(ComplexParamsMixin, JavaMLReadable, JavaMLWritable, JavaEstimator):
"""
Args:
baseStage (object): base pipeline stage to apply to every column
inputCols (list): list of column names encoded as a string
outputCols (list): list of column names encoded as a string
"""
@keyword_only
def __init__(self, baseStage=None, inputCols=None, outputCols=None):
super(MultiColumnAdapter, self).__init__()
self._java_obj = self._new_java_obj("com.microsoft.ml.spark.stages.MultiColumnAdapter")
self._cache = {}
self.baseStage = Param(self, "baseStage", "baseStage: base pipeline stage to apply to every column", generateTypeConverter("baseStage", self._cache, complexTypeConverter))
self.inputCols = Param(self, "inputCols", "inputCols: list of column names encoded as a string")
self.outputCols = Param(self, "outputCols", "outputCols: list of column names encoded as a string")
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
[docs] @keyword_only
def setParams(self, baseStage=None, inputCols=None, outputCols=None):
"""
Set the (keyword only) parameters
Args:
baseStage (object): base pipeline stage to apply to every column
inputCols (list): list of column names encoded as a string
outputCols (list): list of column names encoded as a string
"""
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
return self._set(**kwargs)
[docs] def setBaseStage(self, value):
"""
Args:
baseStage (object): base pipeline stage to apply to every column
"""
self._set(baseStage=value)
return self
[docs] def getBaseStage(self):
"""
Returns:
object: base pipeline stage to apply to every column
"""
return self._cache.get("baseStage", None)
[docs] def setOutputCols(self, value):
"""
Args:
outputCols (list): list of column names encoded as a string
"""
self._set(outputCols=value)
return self
[docs] def getOutputCols(self):
"""
Returns:
list: list of column names encoded as a string
"""
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.stages.MultiColumnAdapter"
@staticmethod
def _from_java(java_stage):
module_name=MultiColumnAdapter.__module__
module_name=module_name.rsplit(".", 1)[0] + ".MultiColumnAdapter"
return from_java(java_stage, module_name)
def _create_model(self, java_model):
return PipelineModel(java_model)
[docs]class PipelineModel(ComplexParamsMixin, JavaModel, JavaMLWritable, JavaMLReadable):
"""
Model fitted by :class:`MultiColumnAdapter`.
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 "org.apache.spark.ml.PipelineModel"
@staticmethod
def _from_java(java_stage):
module_name=PipelineModel.__module__
module_name=module_name.rsplit(".", 1)[0] + ".PipelineModel"
return from_java(java_stage, module_name)