# 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 import SparkContext, SQLContext
from pyspark.sql import DataFrame
from pyspark.ml.param.shared import *
from pyspark import keyword_only
from pyspark.ml.util import JavaMLReadable, JavaMLWritable
from synapse.ml.core.platform import running_on_synapse_internal
from synapse.ml.core.serialize.java_params_patch import *
from pyspark.ml.wrapper import JavaTransformer, JavaEstimator, JavaModel
from pyspark.ml.evaluation import JavaEvaluator
from pyspark.ml.common import inherit_doc
from synapse.ml.core.schema.Utils import *
from pyspark.ml.param import TypeConverters
from synapse.ml.core.schema.TypeConversionUtils import generateTypeConverter, complexTypeConverter
from synapse.ml.recommendation.RankingAdapterModel import RankingAdapterModel
[docs]@inherit_doc
class RankingAdapter(ComplexParamsMixin, JavaMLReadable, JavaMLWritable, JavaEstimator):
"""
Args:
itemCol (str): Column of items
k (int): number of items
labelCol (str): The name of the label column
minRatingsPerItem (int): min ratings for items > 0
minRatingsPerUser (int): min ratings for users > 0
mode (str): recommendation mode
ratingCol (str): Column of ratings
recommender (object): estimator for selection
userCol (str): Column of users
"""
itemCol = Param(Params._dummy(), "itemCol", "Column of items", typeConverter=TypeConverters.toString)
k = Param(Params._dummy(), "k", "number of items", typeConverter=TypeConverters.toInt)
labelCol = Param(Params._dummy(), "labelCol", "The name of the label column", typeConverter=TypeConverters.toString)
minRatingsPerItem = Param(Params._dummy(), "minRatingsPerItem", "min ratings for items > 0", typeConverter=TypeConverters.toInt)
minRatingsPerUser = Param(Params._dummy(), "minRatingsPerUser", "min ratings for users > 0", typeConverter=TypeConverters.toInt)
mode = Param(Params._dummy(), "mode", "recommendation mode", typeConverter=TypeConverters.toString)
ratingCol = Param(Params._dummy(), "ratingCol", "Column of ratings", typeConverter=TypeConverters.toString)
recommender = Param(Params._dummy(), "recommender", "estimator for selection")
userCol = Param(Params._dummy(), "userCol", "Column of users", typeConverter=TypeConverters.toString)
@keyword_only
def __init__(
self,
java_obj=None,
itemCol=None,
k=10,
labelCol="label",
minRatingsPerItem=1,
minRatingsPerUser=1,
mode="allUsers",
ratingCol=None,
recommender=None,
userCol=None
):
super(RankingAdapter, self).__init__()
if java_obj is None:
self._java_obj = self._new_java_obj("com.microsoft.azure.synapse.ml.recommendation.RankingAdapter", self.uid)
else:
self._java_obj = java_obj
self._setDefault(k=10)
self._setDefault(labelCol="label")
self._setDefault(minRatingsPerItem=1)
self._setDefault(minRatingsPerUser=1)
self._setDefault(mode="allUsers")
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
if java_obj is None:
for k,v in kwargs.items():
if v is not None:
getattr(self, "set" + k[0].upper() + k[1:])(v)
[docs] @keyword_only
def setParams(
self,
itemCol=None,
k=10,
labelCol="label",
minRatingsPerItem=1,
minRatingsPerUser=1,
mode="allUsers",
ratingCol=None,
recommender=None,
userCol=None
):
"""
Set the (keyword only) parameters
"""
if hasattr(self, "_input_kwargs"):
kwargs = self._input_kwargs
else:
kwargs = self.__init__._input_kwargs
return self._set(**kwargs)
[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.azure.synapse.ml.recommendation.RankingAdapter"
@staticmethod
def _from_java(java_stage):
module_name=RankingAdapter.__module__
module_name=module_name.rsplit(".", 1)[0] + ".RankingAdapter"
return from_java(java_stage, module_name)
[docs] def setItemCol(self, value):
"""
Args:
itemCol: Column of items
"""
self._set(itemCol=value)
return self
[docs] def setK(self, value):
"""
Args:
k: number of items
"""
self._set(k=value)
return self
[docs] def setLabelCol(self, value):
"""
Args:
labelCol: The name of the label column
"""
self._set(labelCol=value)
return self
[docs] def setMinRatingsPerItem(self, value):
"""
Args:
minRatingsPerItem: min ratings for items > 0
"""
self._set(minRatingsPerItem=value)
return self
[docs] def setMinRatingsPerUser(self, value):
"""
Args:
minRatingsPerUser: min ratings for users > 0
"""
self._set(minRatingsPerUser=value)
return self
[docs] def setMode(self, value):
"""
Args:
mode: recommendation mode
"""
self._set(mode=value)
return self
[docs] def setRatingCol(self, value):
"""
Args:
ratingCol: Column of ratings
"""
self._set(ratingCol=value)
return self
[docs] def setRecommender(self, value):
"""
Args:
recommender: estimator for selection
"""
self._set(recommender=value)
return self
[docs] def setUserCol(self, value):
"""
Args:
userCol: Column of users
"""
self._set(userCol=value)
return self
[docs] def getItemCol(self):
"""
Returns:
itemCol: Column of items
"""
return self.getOrDefault(self.itemCol)
[docs] def getK(self):
"""
Returns:
k: number of items
"""
return self.getOrDefault(self.k)
[docs] def getLabelCol(self):
"""
Returns:
labelCol: The name of the label column
"""
return self.getOrDefault(self.labelCol)
[docs] def getMinRatingsPerItem(self):
"""
Returns:
minRatingsPerItem: min ratings for items > 0
"""
return self.getOrDefault(self.minRatingsPerItem)
[docs] def getMinRatingsPerUser(self):
"""
Returns:
minRatingsPerUser: min ratings for users > 0
"""
return self.getOrDefault(self.minRatingsPerUser)
[docs] def getMode(self):
"""
Returns:
mode: recommendation mode
"""
return self.getOrDefault(self.mode)
[docs] def getRatingCol(self):
"""
Returns:
ratingCol: Column of ratings
"""
return self.getOrDefault(self.ratingCol)
[docs] def getRecommender(self):
"""
Returns:
recommender: estimator for selection
"""
return JavaParams._from_java(self._java_obj.getRecommender())
[docs] def getUserCol(self):
"""
Returns:
userCol: Column of users
"""
return self.getOrDefault(self.userCol)
def _create_model(self, java_model):
try:
model = RankingAdapterModel(java_obj=java_model)
model._transfer_params_from_java()
except TypeError:
model = RankingAdapterModel._from_java(java_model)
return model
def _fit(self, dataset):
java_model = self._fit_java(dataset)
return self._create_model(java_model)