# 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 pyspark.ml.wrapper import JavaTransformer, JavaEstimator, JavaModel
from pyspark.ml.common import inherit_doc
from mmlspark.Utils import *
from mmlspark.TypeConversionUtils import generateTypeConverter, complexTypeConverter
[docs]@inherit_doc
class FindBestModel(ComplexParamsMixin, JavaMLReadable, JavaMLWritable, JavaEstimator):
"""
Evaluates and chooses the best model from a list of models.
Args:
evaluationMetric (str): Metric to evaluate models with (default: accuracy)
models (object): List of models to be evaluated
"""
@keyword_only
def __init__(self, evaluationMetric="accuracy", models=None):
super(FindBestModel, self).__init__()
self._java_obj = self._new_java_obj("com.microsoft.ml.spark.FindBestModel")
self._cache = {}
self.evaluationMetric = Param(self, "evaluationMetric", "evaluationMetric: Metric to evaluate models with (default: accuracy)")
self._setDefault(evaluationMetric="accuracy")
self.models = Param(self, "models", "models: List of models to be evaluated", generateTypeConverter("models", self._cache, complexTypeConverter))
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
[docs] @keyword_only
def setParams(self, evaluationMetric="accuracy", models=None):
"""
Set the (keyword only) parameters
Args:
evaluationMetric (str): Metric to evaluate models with (default: accuracy)
models (object): List of models to be evaluated
"""
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
return self._set(**kwargs)
[docs] def setEvaluationMetric(self, value):
"""
Args:
evaluationMetric (str): Metric to evaluate models with (default: accuracy)
"""
self._set(evaluationMetric=value)
return self
[docs] def getEvaluationMetric(self):
"""
Returns:
str: Metric to evaluate models with (default: accuracy)
"""
return self.getOrDefault(self.evaluationMetric)
[docs] def setModels(self, value):
"""
Args:
models (object): List of models to be evaluated
"""
self._set(models=value)
return self
[docs] def getModels(self):
"""
Returns:
object: List of models to be evaluated
"""
return self._cache.get("models", None)
[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.FindBestModel"
@staticmethod
def _from_java(java_stage):
module_name=FindBestModel.__module__
module_name=module_name.rsplit(".", 1)[0] + ".FindBestModel"
return from_java(java_stage, module_name)
def _create_model(self, java_model):
return BestModel(java_model)
[docs]class BestModel(ComplexParamsMixin, JavaModel, JavaMLWritable, JavaMLReadable):
"""
Model fitted by :class:`FindBestModel`.
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.BestModel"
@staticmethod
def _from_java(java_stage):
module_name=BestModel.__module__
module_name=module_name.rsplit(".", 1)[0] + ".BestModel"
return from_java(java_stage, module_name)