org.apache.spark.ml.classification

Class LogisticRegression

java.lang.Object

org.apache.spark.ml.PipelineStage

org.apache.spark.ml.Estimator<M>

org.apache.spark.ml.Predictor<FeaturesType,E,M>

org.apache.spark.ml.classification.Classifier<FeaturesType,E,M>

org.apache.spark.ml.classification.ProbabilisticClassifier<Vector,LogisticRegression,LogisticRegressionModel>

org.apache.spark.ml.classification.LogisticRegression

All Implemented Interfaces:

java.io.Serializable, Logging, Params, Identifiable

public class LogisticRegression
extends ProbabilisticClassifier<Vector,LogisticRegression,LogisticRegressionModel>
implements Logging

:: Experimental :: Logistic regression. Currently, this class only supports binary classification. It will support multiclass in the future.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
LogisticRegression()
LogisticRegression(java.lang.String uid)

Method Summary

Methods

Modifier and Type	Method and Description
void	checkThresholdConsistency() If threshold and thresholds are both set, ensures they are consistent.
LogisticRegression	copy(ParamMap extra) Creates a copy of this instance with the same UID and some extra params.
Param<java.lang.String>	featuresCol() Param for features column name.
java.lang.String	getFeaturesCol()
java.lang.String	getLabelCol()
java.lang.String	getPredictionCol()
java.lang.String	getRawPredictionCol()
double	getThreshold() Get threshold for binary classification.
double[]	getThresholds() Get thresholds for binary or multiclass classification.
Param<java.lang.String>	labelCol() Param for label column name.
static LogisticRegression	load(java.lang.String path)
Param<java.lang.String>	predictionCol() Param for prediction column name.
Param<java.lang.String>	rawPredictionCol() Param for raw prediction (a.k.a.
LogisticRegression	setElasticNetParam(double value) Set the ElasticNet mixing parameter.
LogisticRegression	setFitIntercept(boolean value) Whether to fit an intercept term.
LogisticRegression	setMaxIter(int value) Set the maximum number of iterations.
LogisticRegression	setRegParam(double value) Set the regularization parameter.
LogisticRegression	setStandardization(boolean value) Whether to standardize the training features before fitting the model.
LogisticRegression	setThreshold(double value) Set threshold in binary classification, in range [0, 1].
LogisticRegression	setThresholds(double[] value) Set thresholds in multiclass (or binary) classification to adjust the probability of predicting each class.
LogisticRegression	setTol(double value) Set the convergence tolerance of iterations.
LogisticRegression	setWeightCol(java.lang.String value) Whether to over-/under-sample training instances according to the given weights in weightCol.
protected LogisticRegressionModel	train(DataFrame dataset) Train a model using the given dataset and parameters.
java.lang.String	uid() An immutable unique ID for the object and its derivatives.
StructType	validateAndTransformSchema(StructType schema, boolean fitting, DataType featuresDataType)
StructType	validateAndTransformSchema(StructType schema, boolean fitting, DataType featuresDataType) Validates and transforms the input schema with the provided param map.
void	validateParams()

Methods inherited from class org.apache.spark.ml.classification.ProbabilisticClassifier

setProbabilityCol

Methods inherited from class org.apache.spark.ml.classification.Classifier

setRawPredictionCol

Methods inherited from class org.apache.spark.ml.Predictor

extractLabeledPoints, fit, setFeaturesCol, setLabelCol, setPredictionCol, transformSchema

Methods inherited from class org.apache.spark.ml.Estimator

fit, fit, fit, fit

Methods inherited from class org.apache.spark.ml.PipelineStage

transformSchema

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface org.apache.spark.ml.param.Params

clear, copyValues, defaultCopy, defaultParamMap, explainParam, explainParams, extractParamMap, extractParamMap, get, getDefault, getOrDefault, getParam, hasDefault, hasParam, isDefined, isSet, paramMap, params, set, set, set, setDefault, setDefault, shouldOwn

Methods inherited from interface org.apache.spark.ml.util.Identifiable

toString

Methods inherited from interface org.apache.spark.Logging

initializeIfNecessary, initializeLogging, isTraceEnabled, log_, log, logDebug, logDebug, logError, logError, logInfo, logInfo, logName, logTrace, logTrace, logWarning, logWarning

Constructor Detail

LogisticRegression

public LogisticRegression(java.lang.String uid)

LogisticRegression

public LogisticRegression()

Method Detail

load

public static LogisticRegression load(java.lang.String path)

uid

public java.lang.String uid()

Description copied from interface: Identifiable

An immutable unique ID for the object and its derivatives.

Specified by:

uid in interface Identifiable

Returns:

(undocumented)

setRegParam

public LogisticRegression setRegParam(double value)

Set the regularization parameter. Default is 0.0.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setElasticNetParam

public LogisticRegression setElasticNetParam(double value)

Set the ElasticNet mixing parameter. For alpha = 0, the penalty is an L2 penalty. For alpha = 1, it is an L1 penalty. For 0 < alpha < 1, the penalty is a combination of L1 and L2. Default is 0.0 which is an L2 penalty.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setMaxIter

public LogisticRegression setMaxIter(int value)

Set the maximum number of iterations. Default is 100.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setTol

public LogisticRegression setTol(double value)

Set the convergence tolerance of iterations. Smaller value will lead to higher accuracy with the cost of more iterations. Default is 1E-6.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setFitIntercept

public LogisticRegression setFitIntercept(boolean value)

Whether to fit an intercept term. Default is true.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setStandardization

public LogisticRegression setStandardization(boolean value)

Whether to standardize the training features before fitting the model. The coefficients of models will be always returned on the original scale, so it will be transparent for users. Note that with/without standardization, the models should be always converged to the same solution when no regularization is applied. In R's GLMNET package, the default behavior is true as well. Default is true.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setThreshold

public LogisticRegression setThreshold(double value)

Set threshold in binary classification, in range [0, 1].

If the estimated probability of class label 1 is > threshold, then predict 1, else 0. A high threshold encourages the model to predict 0 more often; a low threshold encourages the model to predict 1 more often.

Note: Calling this with threshold p is equivalent to calling setThresholds(Array(1-p, p)). When setThreshold() is called, any user-set value for thresholds will be cleared. If both threshold and thresholds are set in a ParamMap, then they must be equivalent.

Default is 0.5.

Parameters:

value - (undocumented)

Returns:

(undocumented)

getThreshold

public double getThreshold()

Get threshold for binary classification.

If threshold is set, returns that value. Otherwise, if thresholds is set with length 2 (i.e., binary classification), this returns the equivalent threshold:

1 / (1 + thresholds(0) / thresholds(1))

. Otherwise, returns {@link threshold} default value. @group getParam @throws IllegalArgumentException if {@link thresholds} is set to an array of length other than 2.

Returns:

(undocumented)

setWeightCol

public LogisticRegression setWeightCol(java.lang.String value)

Whether to over-/under-sample training instances according to the given weights in weightCol. If empty, all instances are treated equally (weight 1.0). Default is empty, so all instances have weight one.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setThresholds

public LogisticRegression setThresholds(double[] value)

Set thresholds in multiclass (or binary) classification to adjust the probability of predicting each class. Array must have length equal to the number of classes, with values >= 0. The class with largest value p/t is predicted, where p is the original probability of that class and t is the class' threshold.

Note: When setThresholds() is called, any user-set value for threshold will be cleared. If both threshold and thresholds are set in a ParamMap, then they must be equivalent.

Overrides:

setThresholds in class ProbabilisticClassifier<Vector,LogisticRegression,LogisticRegressionModel>

Parameters:

value - (undocumented)

Returns:

(undocumented)

getThresholds

public double[] getThresholds()

Get thresholds for binary or multiclass classification.

If thresholds is set, return its value. Otherwise, if threshold is set, return the equivalent thresholds for binary classification: (1-threshold, threshold). If neither are set, throw an exception.

Returns:

(undocumented)

train

protected LogisticRegressionModel train(DataFrame dataset)

Description copied from class: Predictor

Train a model using the given dataset and parameters. Developers can implement this instead of fit() to avoid dealing with schema validation and copying parameters into the model.

Specified by:

train in class Predictor<Vector,LogisticRegression,LogisticRegressionModel>

Parameters:

dataset - Training dataset

Returns:

Fitted model

copy

public LogisticRegression copy(ParamMap extra)

Description copied from interface: Params

Creates a copy of this instance with the same UID and some extra params. Subclasses should implement this method and set the return type properly.

Specified by:

copy in interface Params

Specified by:

copy in class Predictor<Vector,LogisticRegression,LogisticRegressionModel>

Parameters:

extra - (undocumented)

Returns:

(undocumented)

See Also:

defaultCopy()

checkThresholdConsistency

public void checkThresholdConsistency()

If threshold and thresholds are both set, ensures they are consistent.

Throws:

java.lang.IllegalArgumentException - if threshold and thresholds are not equivalent

validateParams

public void validateParams()

validateAndTransformSchema

public StructType validateAndTransformSchema(StructType schema,
                                    boolean fitting,
                                    DataType featuresDataType)

rawPredictionCol

public Param<java.lang.String> rawPredictionCol()

Param for raw prediction (a.k.a. confidence) column name.

Returns:

(undocumented)

getRawPredictionCol

public java.lang.String getRawPredictionCol()

validateAndTransformSchema

public StructType validateAndTransformSchema(StructType schema,
                                    boolean fitting,
                                    DataType featuresDataType)

Validates and transforms the input schema with the provided param map.

Parameters:

schema - input schema

fitting - whether this is in fitting

featuresDataType - SQL DataType for FeaturesType. E.g., VectorUDT for vector features.

Returns:

output schema

labelCol

public Param<java.lang.String> labelCol()

Param for label column name.

Returns:

(undocumented)

getLabelCol

public java.lang.String getLabelCol()

featuresCol

public Param<java.lang.String> featuresCol()

Param for features column name.

Returns:

(undocumented)

getFeaturesCol

public java.lang.String getFeaturesCol()

predictionCol

public Param<java.lang.String> predictionCol()

Param for prediction column name.

Returns:

(undocumented)

getPredictionCol

public java.lang.String getPredictionCol()