#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
from pyspark import since, keyword_only
from pyspark.ml.util import *
from pyspark.ml.wrapper import JavaEstimator, JavaModel
from pyspark.ml.param.shared import *
from pyspark.mllib.common import inherit_doc
__all__ = ['BisectingKMeans', 'BisectingKMeansModel',
'KMeans', 'KMeansModel',
'GaussianMixture', 'GaussianMixtureModel',
'LDA', 'LDAModel', 'LocalLDAModel', 'DistributedLDAModel']
[docs]class GaussianMixtureModel(JavaModel, JavaMLWritable, JavaMLReadable):
"""
.. note:: Experimental
Model fitted by GaussianMixture.
.. versionadded:: 2.0.0
"""
@property
@since("2.0.0")
def weights(self):
"""
Weight for each Gaussian distribution in the mixture.
This is a multinomial probability distribution over the k Gaussians,
where weights[i] is the weight for Gaussian i, and weights sum to 1.
"""
return self._call_java("weights")
@property
@since("2.0.0")
def gaussiansDF(self):
"""
Retrieve Gaussian distributions as a DataFrame.
Each row represents a Gaussian Distribution.
The DataFrame has two columns: mean (Vector) and cov (Matrix).
"""
return self._call_java("gaussiansDF")
@inherit_doc
[docs]class GaussianMixture(JavaEstimator, HasFeaturesCol, HasPredictionCol, HasMaxIter, HasTol, HasSeed,
HasProbabilityCol, JavaMLWritable, JavaMLReadable):
"""
.. note:: Experimental
GaussianMixture clustering.
>>> from pyspark.ml.linalg import Vectors
>>> data = [(Vectors.dense([-0.1, -0.05 ]),),
... (Vectors.dense([-0.01, -0.1]),),
... (Vectors.dense([0.9, 0.8]),),
... (Vectors.dense([0.75, 0.935]),),
... (Vectors.dense([-0.83, -0.68]),),
... (Vectors.dense([-0.91, -0.76]),)]
>>> df = sqlContext.createDataFrame(data, ["features"])
>>> gm = GaussianMixture(k=3, tol=0.0001,
... maxIter=10, seed=10)
>>> model = gm.fit(df)
>>> weights = model.weights
>>> len(weights)
3
>>> model.gaussiansDF.show()
+--------------------+--------------------+
| mean| cov|
+--------------------+--------------------+
|[-0.0550000000000...|0.002025000000000...|
|[0.82499999999999...|0.005625000000000...|
|[-0.87,-0.7200000...|0.001600000000000...|
+--------------------+--------------------+
...
>>> transformed = model.transform(df).select("features", "prediction")
>>> rows = transformed.collect()
>>> rows[4].prediction == rows[5].prediction
True
>>> rows[2].prediction == rows[3].prediction
True
>>> gmm_path = temp_path + "/gmm"
>>> gm.save(gmm_path)
>>> gm2 = GaussianMixture.load(gmm_path)
>>> gm2.getK()
3
>>> model_path = temp_path + "/gmm_model"
>>> model.save(model_path)
>>> model2 = GaussianMixtureModel.load(model_path)
>>> model2.weights == model.weights
True
>>> model2.gaussiansDF.show()
+--------------------+--------------------+
| mean| cov|
+--------------------+--------------------+
|[-0.0550000000000...|0.002025000000000...|
|[0.82499999999999...|0.005625000000000...|
|[-0.87,-0.7200000...|0.001600000000000...|
+--------------------+--------------------+
...
.. versionadded:: 2.0.0
"""
k = Param(Params._dummy(), "k", "number of clusters to create",
typeConverter=TypeConverters.toInt)
@keyword_only
def __init__(self, featuresCol="features", predictionCol="prediction", k=2,
probabilityCol="probability", tol=0.01, maxIter=100, seed=None):
"""
__init__(self, featuresCol="features", predictionCol="prediction", k=2, \
probabilityCol="probability", tol=0.01, maxIter=100, seed=None)
"""
super(GaussianMixture, self).__init__()
self._java_obj = self._new_java_obj("org.apache.spark.ml.clustering.GaussianMixture",
self.uid)
self._setDefault(k=2, tol=0.01, maxIter=100)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
def _create_model(self, java_model):
return GaussianMixtureModel(java_model)
@keyword_only
@since("2.0.0")
[docs] def setParams(self, featuresCol="features", predictionCol="prediction", k=2,
probabilityCol="probability", tol=0.01, maxIter=100, seed=None):
"""
setParams(self, featuresCol="features", predictionCol="prediction", k=2, \
probabilityCol="probability", tol=0.01, maxIter=100, seed=None)
Sets params for GaussianMixture.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
@since("2.0.0")
[docs] def setK(self, value):
"""
Sets the value of :py:attr:`k`.
"""
return self._set(k=value)
@since("2.0.0")
[docs] def getK(self):
"""
Gets the value of `k`
"""
return self.getOrDefault(self.k)
[docs]class KMeansModel(JavaModel, JavaMLWritable, JavaMLReadable):
"""
Model fitted by KMeans.
.. versionadded:: 1.5.0
"""
@since("1.5.0")
[docs] def clusterCenters(self):
"""Get the cluster centers, represented as a list of NumPy arrays."""
return [c.toArray() for c in self._call_java("clusterCenters")]
@since("2.0.0")
[docs] def computeCost(self, dataset):
"""
Return the K-means cost (sum of squared distances of points to their nearest center)
for this model on the given data.
"""
return self._call_java("computeCost", dataset)
@inherit_doc
[docs]class KMeans(JavaEstimator, HasFeaturesCol, HasPredictionCol, HasMaxIter, HasTol, HasSeed,
JavaMLWritable, JavaMLReadable):
"""
K-means clustering with a k-means++ like initialization mode
(the k-means|| algorithm by Bahmani et al).
>>> from pyspark.ml.linalg import Vectors
>>> data = [(Vectors.dense([0.0, 0.0]),), (Vectors.dense([1.0, 1.0]),),
... (Vectors.dense([9.0, 8.0]),), (Vectors.dense([8.0, 9.0]),)]
>>> df = sqlContext.createDataFrame(data, ["features"])
>>> kmeans = KMeans(k=2, seed=1)
>>> model = kmeans.fit(df)
>>> centers = model.clusterCenters()
>>> len(centers)
2
>>> model.computeCost(df)
2.000...
>>> transformed = model.transform(df).select("features", "prediction")
>>> rows = transformed.collect()
>>> rows[0].prediction == rows[1].prediction
True
>>> rows[2].prediction == rows[3].prediction
True
>>> kmeans_path = temp_path + "/kmeans"
>>> kmeans.save(kmeans_path)
>>> kmeans2 = KMeans.load(kmeans_path)
>>> kmeans2.getK()
2
>>> model_path = temp_path + "/kmeans_model"
>>> model.save(model_path)
>>> model2 = KMeansModel.load(model_path)
>>> model.clusterCenters()[0] == model2.clusterCenters()[0]
array([ True, True], dtype=bool)
>>> model.clusterCenters()[1] == model2.clusterCenters()[1]
array([ True, True], dtype=bool)
.. versionadded:: 1.5.0
"""
k = Param(Params._dummy(), "k", "number of clusters to create",
typeConverter=TypeConverters.toInt)
initMode = Param(Params._dummy(), "initMode",
"the initialization algorithm. This can be either \"random\" to " +
"choose random points as initial cluster centers, or \"k-means||\" " +
"to use a parallel variant of k-means++",
typeConverter=TypeConverters.toString)
initSteps = Param(Params._dummy(), "initSteps", "steps for k-means initialization mode",
typeConverter=TypeConverters.toInt)
@keyword_only
def __init__(self, featuresCol="features", predictionCol="prediction", k=2,
initMode="k-means||", initSteps=5, tol=1e-4, maxIter=20, seed=None):
"""
__init__(self, featuresCol="features", predictionCol="prediction", k=2, \
initMode="k-means||", initSteps=5, tol=1e-4, maxIter=20, seed=None)
"""
super(KMeans, self).__init__()
self._java_obj = self._new_java_obj("org.apache.spark.ml.clustering.KMeans", self.uid)
self._setDefault(k=2, initMode="k-means||", initSteps=5, tol=1e-4, maxIter=20)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
def _create_model(self, java_model):
return KMeansModel(java_model)
@keyword_only
@since("1.5.0")
[docs] def setParams(self, featuresCol="features", predictionCol="prediction", k=2,
initMode="k-means||", initSteps=5, tol=1e-4, maxIter=20, seed=None):
"""
setParams(self, featuresCol="features", predictionCol="prediction", k=2, \
initMode="k-means||", initSteps=5, tol=1e-4, maxIter=20, seed=None)
Sets params for KMeans.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
@since("1.5.0")
[docs] def setK(self, value):
"""
Sets the value of :py:attr:`k`.
"""
return self._set(k=value)
@since("1.5.0")
[docs] def getK(self):
"""
Gets the value of `k`
"""
return self.getOrDefault(self.k)
@since("1.5.0")
[docs] def setInitMode(self, value):
"""
Sets the value of :py:attr:`initMode`.
"""
return self._set(initMode=value)
@since("1.5.0")
[docs] def getInitMode(self):
"""
Gets the value of `initMode`
"""
return self.getOrDefault(self.initMode)
@since("1.5.0")
[docs] def setInitSteps(self, value):
"""
Sets the value of :py:attr:`initSteps`.
"""
return self._set(initSteps=value)
@since("1.5.0")
[docs] def getInitSteps(self):
"""
Gets the value of `initSteps`
"""
return self.getOrDefault(self.initSteps)
[docs]class BisectingKMeansModel(JavaModel, JavaMLWritable, JavaMLReadable):
"""
.. note:: Experimental
Model fitted by BisectingKMeans.
.. versionadded:: 2.0.0
"""
@since("2.0.0")
[docs] def clusterCenters(self):
"""Get the cluster centers, represented as a list of NumPy arrays."""
return [c.toArray() for c in self._call_java("clusterCenters")]
@since("2.0.0")
[docs] def computeCost(self, dataset):
"""
Computes the sum of squared distances between the input points
and their corresponding cluster centers.
"""
return self._call_java("computeCost", dataset)
@inherit_doc
[docs]class BisectingKMeans(JavaEstimator, HasFeaturesCol, HasPredictionCol, HasMaxIter, HasSeed,
JavaMLWritable, JavaMLReadable):
"""
.. note:: Experimental
A bisecting k-means algorithm based on the paper "A comparison of document clustering
techniques" by Steinbach, Karypis, and Kumar, with modification to fit Spark.
The algorithm starts from a single cluster that contains all points.
Iteratively it finds divisible clusters on the bottom level and bisects each of them using
k-means, until there are `k` leaf clusters in total or no leaf clusters are divisible.
The bisecting steps of clusters on the same level are grouped together to increase parallelism.
If bisecting all divisible clusters on the bottom level would result more than `k` leaf
clusters, larger clusters get higher priority.
>>> from pyspark.ml.linalg import Vectors
>>> data = [(Vectors.dense([0.0, 0.0]),), (Vectors.dense([1.0, 1.0]),),
... (Vectors.dense([9.0, 8.0]),), (Vectors.dense([8.0, 9.0]),)]
>>> df = sqlContext.createDataFrame(data, ["features"])
>>> bkm = BisectingKMeans(k=2, minDivisibleClusterSize=1.0)
>>> model = bkm.fit(df)
>>> centers = model.clusterCenters()
>>> len(centers)
2
>>> model.computeCost(df)
2.000...
>>> transformed = model.transform(df).select("features", "prediction")
>>> rows = transformed.collect()
>>> rows[0].prediction == rows[1].prediction
True
>>> rows[2].prediction == rows[3].prediction
True
>>> bkm_path = temp_path + "/bkm"
>>> bkm.save(bkm_path)
>>> bkm2 = BisectingKMeans.load(bkm_path)
>>> bkm2.getK()
2
>>> model_path = temp_path + "/bkm_model"
>>> model.save(model_path)
>>> model2 = BisectingKMeansModel.load(model_path)
>>> model.clusterCenters()[0] == model2.clusterCenters()[0]
array([ True, True], dtype=bool)
>>> model.clusterCenters()[1] == model2.clusterCenters()[1]
array([ True, True], dtype=bool)
.. versionadded:: 2.0.0
"""
k = Param(Params._dummy(), "k", "number of clusters to create",
typeConverter=TypeConverters.toInt)
minDivisibleClusterSize = Param(Params._dummy(), "minDivisibleClusterSize",
"the minimum number of points (if >= 1.0) " +
"or the minimum proportion",
typeConverter=TypeConverters.toFloat)
@keyword_only
def __init__(self, featuresCol="features", predictionCol="prediction", maxIter=20,
seed=None, k=4, minDivisibleClusterSize=1.0):
"""
__init__(self, featuresCol="features", predictionCol="prediction", maxIter=20, \
seed=None, k=4, minDivisibleClusterSize=1.0)
"""
super(BisectingKMeans, self).__init__()
self._java_obj = self._new_java_obj("org.apache.spark.ml.clustering.BisectingKMeans",
self.uid)
self._setDefault(maxIter=20, k=4, minDivisibleClusterSize=1.0)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
@keyword_only
@since("2.0.0")
[docs] def setParams(self, featuresCol="features", predictionCol="prediction", maxIter=20,
seed=None, k=4, minDivisibleClusterSize=1.0):
"""
setParams(self, featuresCol="features", predictionCol="prediction", maxIter=20, \
seed=None, k=4, minDivisibleClusterSize=1.0)
Sets params for BisectingKMeans.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
@since("2.0.0")
[docs] def setK(self, value):
"""
Sets the value of :py:attr:`k`.
"""
return self._set(k=value)
@since("2.0.0")
[docs] def getK(self):
"""
Gets the value of `k` or its default value.
"""
return self.getOrDefault(self.k)
@since("2.0.0")
[docs] def setMinDivisibleClusterSize(self, value):
"""
Sets the value of :py:attr:`minDivisibleClusterSize`.
"""
return self._set(minDivisibleClusterSize=value)
@since("2.0.0")
[docs] def getMinDivisibleClusterSize(self):
"""
Gets the value of `minDivisibleClusterSize` or its default value.
"""
return self.getOrDefault(self.minDivisibleClusterSize)
def _create_model(self, java_model):
return BisectingKMeansModel(java_model)
@inherit_doc
[docs]class LDAModel(JavaModel):
"""
.. note:: Experimental
Latent Dirichlet Allocation (LDA) model.
This abstraction permits for different underlying representations,
including local and distributed data structures.
.. versionadded:: 2.0.0
"""
@since("2.0.0")
[docs] def isDistributed(self):
"""
Indicates whether this instance is of type DistributedLDAModel
"""
return self._call_java("isDistributed")
@since("2.0.0")
[docs] def vocabSize(self):
"""Vocabulary size (number of terms or words in the vocabulary)"""
return self._call_java("vocabSize")
@since("2.0.0")
[docs] def topicsMatrix(self):
"""
Inferred topics, where each topic is represented by a distribution over terms.
This is a matrix of size vocabSize x k, where each column is a topic.
No guarantees are given about the ordering of the topics.
WARNING: If this model is actually a :py:class:`DistributedLDAModel` instance produced by
the Expectation-Maximization ("em") `optimizer`, then this method could involve
collecting a large amount of data to the driver (on the order of vocabSize x k).
"""
return self._call_java("topicsMatrix")
@since("2.0.0")
[docs] def logLikelihood(self, dataset):
"""
Calculates a lower bound on the log likelihood of the entire corpus.
See Equation (16) in the Online LDA paper (Hoffman et al., 2010).
WARNING: If this model is an instance of :py:class:`DistributedLDAModel` (produced when
:py:attr:`optimizer` is set to "em"), this involves collecting a large
:py:func:`topicsMatrix` to the driver. This implementation may be changed in the future.
"""
return self._call_java("logLikelihood", dataset)
@since("2.0.0")
[docs] def logPerplexity(self, dataset):
"""
Calculate an upper bound bound on perplexity. (Lower is better.)
See Equation (16) in the Online LDA paper (Hoffman et al., 2010).
WARNING: If this model is an instance of :py:class:`DistributedLDAModel` (produced when
:py:attr:`optimizer` is set to "em"), this involves collecting a large
:py:func:`topicsMatrix` to the driver. This implementation may be changed in the future.
"""
return self._call_java("logPerplexity", dataset)
@since("2.0.0")
[docs] def describeTopics(self, maxTermsPerTopic=10):
"""
Return the topics described by their top-weighted terms.
"""
return self._call_java("describeTopics", maxTermsPerTopic)
@since("2.0.0")
[docs] def estimatedDocConcentration(self):
"""
Value for :py:attr:`LDA.docConcentration` estimated from data.
If Online LDA was used and :py:attr:`LDA.optimizeDocConcentration` was set to false,
then this returns the fixed (given) value for the :py:attr:`LDA.docConcentration` parameter.
"""
return self._call_java("estimatedDocConcentration")
@inherit_doc
[docs]class DistributedLDAModel(LDAModel, JavaMLReadable, JavaMLWritable):
"""
.. note:: Experimental
Distributed model fitted by :py:class:`LDA`.
This type of model is currently only produced by Expectation-Maximization (EM).
This model stores the inferred topics, the full training dataset, and the topic distribution
for each training document.
.. versionadded:: 2.0.0
"""
@since("2.0.0")
[docs] def toLocal(self):
"""
Convert this distributed model to a local representation. This discards info about the
training dataset.
WARNING: This involves collecting a large :py:func:`topicsMatrix` to the driver.
"""
return LocalLDAModel(self._call_java("toLocal"))
@since("2.0.0")
[docs] def trainingLogLikelihood(self):
"""
Log likelihood of the observed tokens in the training set,
given the current parameter estimates:
log P(docs | topics, topic distributions for docs, Dirichlet hyperparameters)
Notes:
- This excludes the prior; for that, use :py:func:`logPrior`.
- Even with :py:func:`logPrior`, this is NOT the same as the data log likelihood given
the hyperparameters.
- This is computed from the topic distributions computed during training. If you call
:py:func:`logLikelihood` on the same training dataset, the topic distributions
will be computed again, possibly giving different results.
"""
return self._call_java("trainingLogLikelihood")
@since("2.0.0")
[docs] def logPrior(self):
"""
Log probability of the current parameter estimate:
log P(topics, topic distributions for docs | alpha, eta)
"""
return self._call_java("logPrior")
@since("2.0.0")
[docs] def getCheckpointFiles(self):
"""
If using checkpointing and :py:attr:`LDA.keepLastCheckpoint` is set to true, then there may
be saved checkpoint files. This method is provided so that users can manage those files.
Note that removing the checkpoints can cause failures if a partition is lost and is needed
by certain :py:class:`DistributedLDAModel` methods. Reference counting will clean up the
checkpoints when this model and derivative data go out of scope.
:return List of checkpoint files from training
"""
return self._call_java("getCheckpointFiles")
@inherit_doc
[docs]class LocalLDAModel(LDAModel, JavaMLReadable, JavaMLWritable):
"""
.. note:: Experimental
Local (non-distributed) model fitted by :py:class:`LDA`.
This model stores the inferred topics only; it does not store info about the training dataset.
.. versionadded:: 2.0.0
"""
pass
@inherit_doc
[docs]class LDA(JavaEstimator, HasFeaturesCol, HasMaxIter, HasSeed, HasCheckpointInterval,
JavaMLReadable, JavaMLWritable):
"""
.. note:: Experimental
Latent Dirichlet Allocation (LDA), a topic model designed for text documents.
Terminology:
- "term" = "word": an el
- "token": instance of a term appearing in a document
- "topic": multinomial distribution over terms representing some concept
- "document": one piece of text, corresponding to one row in the input data
Original LDA paper (journal version):
Blei, Ng, and Jordan. "Latent Dirichlet Allocation." JMLR, 2003.
Input data (featuresCol):
LDA is given a collection of documents as input data, via the featuresCol parameter.
Each document is specified as a :py:class:`Vector` of length vocabSize, where each entry is the
count for the corresponding term (word) in the document. Feature transformers such as
:py:class:`pyspark.ml.feature.Tokenizer` and :py:class:`pyspark.ml.feature.CountVectorizer`
can be useful for converting text to word count vectors.
>>> from pyspark.ml.linalg import Vectors, SparseVector
>>> from pyspark.ml.clustering import LDA
>>> df = sqlContext.createDataFrame([[1, Vectors.dense([0.0, 1.0])],
... [2, SparseVector(2, {0: 1.0})],], ["id", "features"])
>>> lda = LDA(k=2, seed=1, optimizer="em")
>>> model = lda.fit(df)
>>> model.isDistributed()
True
>>> localModel = model.toLocal()
>>> localModel.isDistributed()
False
>>> model.vocabSize()
2
>>> model.describeTopics().show()
+-----+-----------+--------------------+
|topic|termIndices| termWeights|
+-----+-----------+--------------------+
| 0| [1, 0]|[0.50401530077160...|
| 1| [0, 1]|[0.50401530077160...|
+-----+-----------+--------------------+
...
>>> model.topicsMatrix()
DenseMatrix(2, 2, [0.496, 0.504, 0.504, 0.496], 0)
>>> lda_path = temp_path + "/lda"
>>> lda.save(lda_path)
>>> sameLDA = LDA.load(lda_path)
>>> distributed_model_path = temp_path + "/lda_distributed_model"
>>> model.save(distributed_model_path)
>>> sameModel = DistributedLDAModel.load(distributed_model_path)
>>> local_model_path = temp_path + "/lda_local_model"
>>> localModel.save(local_model_path)
>>> sameLocalModel = LocalLDAModel.load(local_model_path)
.. versionadded:: 2.0.0
"""
k = Param(Params._dummy(), "k", "number of topics (clusters) to infer",
typeConverter=TypeConverters.toInt)
optimizer = Param(Params._dummy(), "optimizer",
"Optimizer or inference algorithm used to estimate the LDA model. "
"Supported: online, em", typeConverter=TypeConverters.toString)
learningOffset = Param(Params._dummy(), "learningOffset",
"A (positive) learning parameter that downweights early iterations."
" Larger values make early iterations count less",
typeConverter=TypeConverters.toFloat)
learningDecay = Param(Params._dummy(), "learningDecay", "Learning rate, set as an"
"exponential decay rate. This should be between (0.5, 1.0] to "
"guarantee asymptotic convergence.", typeConverter=TypeConverters.toFloat)
subsamplingRate = Param(Params._dummy(), "subsamplingRate",
"Fraction of the corpus to be sampled and used in each iteration "
"of mini-batch gradient descent, in range (0, 1].",
typeConverter=TypeConverters.toFloat)
optimizeDocConcentration = Param(Params._dummy(), "optimizeDocConcentration",
"Indicates whether the docConcentration (Dirichlet parameter "
"for document-topic distribution) will be optimized during "
"training.", typeConverter=TypeConverters.toBoolean)
docConcentration = Param(Params._dummy(), "docConcentration",
"Concentration parameter (commonly named \"alpha\") for the "
"prior placed on documents' distributions over topics (\"theta\").",
typeConverter=TypeConverters.toListFloat)
topicConcentration = Param(Params._dummy(), "topicConcentration",
"Concentration parameter (commonly named \"beta\" or \"eta\") for "
"the prior placed on topic' distributions over terms.",
typeConverter=TypeConverters.toFloat)
topicDistributionCol = Param(Params._dummy(), "topicDistributionCol",
"Output column with estimates of the topic mixture distribution "
"for each document (often called \"theta\" in the literature). "
"Returns a vector of zeros for an empty document.",
typeConverter=TypeConverters.toString)
keepLastCheckpoint = Param(Params._dummy(), "keepLastCheckpoint",
"(For EM optimizer) If using checkpointing, this indicates whether"
" to keep the last checkpoint. If false, then the checkpoint will be"
" deleted. Deleting the checkpoint can cause failures if a data"
" partition is lost, so set this bit with care.",
TypeConverters.toBoolean)
@keyword_only
def __init__(self, featuresCol="features", maxIter=20, seed=None, checkpointInterval=10,
k=10, optimizer="online", learningOffset=1024.0, learningDecay=0.51,
subsamplingRate=0.05, optimizeDocConcentration=True,
docConcentration=None, topicConcentration=None,
topicDistributionCol="topicDistribution", keepLastCheckpoint=True):
"""
__init__(self, featuresCol="features", maxIter=20, seed=None, checkpointInterval=10,\
k=10, optimizer="online", learningOffset=1024.0, learningDecay=0.51,\
subsamplingRate=0.05, optimizeDocConcentration=True,\
docConcentration=None, topicConcentration=None,\
topicDistributionCol="topicDistribution", keepLastCheckpoint=True):
"""
super(LDA, self).__init__()
self._java_obj = self._new_java_obj("org.apache.spark.ml.clustering.LDA", self.uid)
self._setDefault(maxIter=20, checkpointInterval=10,
k=10, optimizer="online", learningOffset=1024.0, learningDecay=0.51,
subsamplingRate=0.05, optimizeDocConcentration=True,
topicDistributionCol="topicDistribution", keepLastCheckpoint=True)
kwargs = self.__init__._input_kwargs
self.setParams(**kwargs)
def _create_model(self, java_model):
if self.getOptimizer() == "em":
return DistributedLDAModel(java_model)
else:
return LocalLDAModel(java_model)
@keyword_only
@since("2.0.0")
[docs] def setParams(self, featuresCol="features", maxIter=20, seed=None, checkpointInterval=10,
k=10, optimizer="online", learningOffset=1024.0, learningDecay=0.51,
subsamplingRate=0.05, optimizeDocConcentration=True,
docConcentration=None, topicConcentration=None,
topicDistributionCol="topicDistribution", keepLastCheckpoint=True):
"""
setParams(self, featuresCol="features", maxIter=20, seed=None, checkpointInterval=10,\
k=10, optimizer="online", learningOffset=1024.0, learningDecay=0.51,\
subsamplingRate=0.05, optimizeDocConcentration=True,\
docConcentration=None, topicConcentration=None,\
topicDistributionCol="topicDistribution", keepLastCheckpoint=True):
Sets params for LDA.
"""
kwargs = self.setParams._input_kwargs
return self._set(**kwargs)
@since("2.0.0")
[docs] def setK(self, value):
"""
Sets the value of :py:attr:`k`.
>>> algo = LDA().setK(10)
>>> algo.getK()
10
"""
return self._set(k=value)
@since("2.0.0")
[docs] def getK(self):
"""
Gets the value of :py:attr:`k` or its default value.
"""
return self.getOrDefault(self.k)
@since("2.0.0")
[docs] def setOptimizer(self, value):
"""
Sets the value of :py:attr:`optimizer`.
Currenlty only support 'em' and 'online'.
>>> algo = LDA().setOptimizer("em")
>>> algo.getOptimizer()
'em'
"""
return self._set(optimizer=value)
@since("2.0.0")
[docs] def getOptimizer(self):
"""
Gets the value of :py:attr:`optimizer` or its default value.
"""
return self.getOrDefault(self.optimizer)
@since("2.0.0")
[docs] def setLearningOffset(self, value):
"""
Sets the value of :py:attr:`learningOffset`.
>>> algo = LDA().setLearningOffset(100)
>>> algo.getLearningOffset()
100.0
"""
return self._set(learningOffset=value)
@since("2.0.0")
[docs] def getLearningOffset(self):
"""
Gets the value of :py:attr:`learningOffset` or its default value.
"""
return self.getOrDefault(self.learningOffset)
@since("2.0.0")
[docs] def setLearningDecay(self, value):
"""
Sets the value of :py:attr:`learningDecay`.
>>> algo = LDA().setLearningDecay(0.1)
>>> algo.getLearningDecay()
0.1...
"""
return self._set(learningDecay=value)
@since("2.0.0")
[docs] def getLearningDecay(self):
"""
Gets the value of :py:attr:`learningDecay` or its default value.
"""
return self.getOrDefault(self.learningDecay)
@since("2.0.0")
[docs] def setSubsamplingRate(self, value):
"""
Sets the value of :py:attr:`subsamplingRate`.
>>> algo = LDA().setSubsamplingRate(0.1)
>>> algo.getSubsamplingRate()
0.1...
"""
return self._set(subsamplingRate=value)
@since("2.0.0")
[docs] def getSubsamplingRate(self):
"""
Gets the value of :py:attr:`subsamplingRate` or its default value.
"""
return self.getOrDefault(self.subsamplingRate)
@since("2.0.0")
[docs] def setOptimizeDocConcentration(self, value):
"""
Sets the value of :py:attr:`optimizeDocConcentration`.
>>> algo = LDA().setOptimizeDocConcentration(True)
>>> algo.getOptimizeDocConcentration()
True
"""
return self._set(optimizeDocConcentration=value)
@since("2.0.0")
[docs] def getOptimizeDocConcentration(self):
"""
Gets the value of :py:attr:`optimizeDocConcentration` or its default value.
"""
return self.getOrDefault(self.optimizeDocConcentration)
@since("2.0.0")
[docs] def setDocConcentration(self, value):
"""
Sets the value of :py:attr:`docConcentration`.
>>> algo = LDA().setDocConcentration([0.1, 0.2])
>>> algo.getDocConcentration()
[0.1..., 0.2...]
"""
return self._set(docConcentration=value)
@since("2.0.0")
[docs] def getDocConcentration(self):
"""
Gets the value of :py:attr:`docConcentration` or its default value.
"""
return self.getOrDefault(self.docConcentration)
@since("2.0.0")
[docs] def setTopicConcentration(self, value):
"""
Sets the value of :py:attr:`topicConcentration`.
>>> algo = LDA().setTopicConcentration(0.5)
>>> algo.getTopicConcentration()
0.5...
"""
return self._set(topicConcentration=value)
@since("2.0.0")
[docs] def getTopicConcentration(self):
"""
Gets the value of :py:attr:`topicConcentration` or its default value.
"""
return self.getOrDefault(self.topicConcentration)
@since("2.0.0")
[docs] def setTopicDistributionCol(self, value):
"""
Sets the value of :py:attr:`topicDistributionCol`.
>>> algo = LDA().setTopicDistributionCol("topicDistributionCol")
>>> algo.getTopicDistributionCol()
'topicDistributionCol'
"""
return self._set(topicDistributionCol=value)
@since("2.0.0")
[docs] def getTopicDistributionCol(self):
"""
Gets the value of :py:attr:`topicDistributionCol` or its default value.
"""
return self.getOrDefault(self.topicDistributionCol)
@since("2.0.0")
[docs] def setKeepLastCheckpoint(self, value):
"""
Sets the value of :py:attr:`keepLastCheckpoint`.
>>> algo = LDA().setKeepLastCheckpoint(False)
>>> algo.getKeepLastCheckpoint()
False
"""
return self._set(keepLastCheckpoint=value)
@since("2.0.0")
[docs] def getKeepLastCheckpoint(self):
"""
Gets the value of :py:attr:`keepLastCheckpoint` or its default value.
"""
return self.getOrDefault(self.keepLastCheckpoint)
if __name__ == "__main__":
import doctest
import pyspark.ml.clustering
from pyspark.context import SparkContext
from pyspark.sql import SQLContext
globs = pyspark.ml.clustering.__dict__.copy()
# The small batch size here ensures that we see multiple batches,
# even in these small test examples:
sc = SparkContext("local[2]", "ml.clustering tests")
sqlContext = SQLContext(sc)
globs['sc'] = sc
globs['sqlContext'] = sqlContext
import tempfile
temp_path = tempfile.mkdtemp()
globs['temp_path'] = temp_path
try:
(failure_count, test_count) = doctest.testmod(globs=globs, optionflags=doctest.ELLIPSIS)
sc.stop()
finally:
from shutil import rmtree
try:
rmtree(temp_path)
except OSError:
pass
if failure_count:
exit(-1)