org.apache.spark.ml.regression

Class GeneralizedLinearRegression

Object

org.apache.spark.ml.PipelineStage

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

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

org.apache.spark.ml.regression.Regressor<Vector,GeneralizedLinearRegression,GeneralizedLinearRegressionModel>

org.apache.spark.ml.regression.GeneralizedLinearRegression

All Implemented Interfaces:

java.io.Serializable, org.apache.spark.internal.Logging, Params, HasAggregationDepth, HasFeaturesCol, HasFitIntercept, HasLabelCol, HasMaxIter, HasPredictionCol, HasRegParam, HasSolver, HasTol, HasWeightCol, PredictorParams, GeneralizedLinearRegressionBase, DefaultParamsWritable, Identifiable, MLWritable

public class GeneralizedLinearRegression
extends Regressor<Vector,GeneralizedLinearRegression,GeneralizedLinearRegressionModel>
implements GeneralizedLinearRegressionBase, DefaultParamsWritable, org.apache.spark.internal.Logging

Fit a Generalized Linear Model (see Generalized linear model (Wikipedia)) specified by giving a symbolic description of the linear predictor (link function) and a description of the error distribution (family). It supports "gaussian", "binomial", "poisson", "gamma" and "tweedie" as family. Valid link functions for each family is listed below. The first link function of each family is the default one. - "gaussian" : "identity", "log", "inverse" - "binomial" : "logit", "probit", "cloglog" - "poisson" : "log", "identity", "sqrt" - "gamma" : "inverse", "identity", "log" - "tweedie" : power link function specified through "linkPower". The default link power in the tweedie family is 1 - variancePower.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	GeneralizedLinearRegression.Binomial$ Binomial exponential family distribution.
static class	GeneralizedLinearRegression.CLogLog$
static class	GeneralizedLinearRegression.Family$
static class	GeneralizedLinearRegression.FamilyAndLink$
static class	GeneralizedLinearRegression.Gamma$ Gamma exponential family distribution.
static class	GeneralizedLinearRegression.Gaussian$ Gaussian exponential family distribution.
static class	GeneralizedLinearRegression.Identity$
static class	GeneralizedLinearRegression.Inverse$
static class	GeneralizedLinearRegression.Link$
static class	GeneralizedLinearRegression.Log$
static class	GeneralizedLinearRegression.Logit$
static class	GeneralizedLinearRegression.Poisson$ Poisson exponential family distribution.
static class	GeneralizedLinearRegression.Probit$
static class	GeneralizedLinearRegression.Sqrt$
static class	GeneralizedLinearRegression.Tweedie$

Constructor Summary

Constructors

Constructor and Description
GeneralizedLinearRegression()
GeneralizedLinearRegression(String uid)

Method Summary

Modifier and Type	Method and Description
IntParam	aggregationDepth() Param for suggested depth for treeAggregate (>= 2).
GeneralizedLinearRegression	copy(ParamMap extra) Creates a copy of this instance with the same UID and some extra params.
Param<String>	family() Param for the name of family which is a description of the error distribution to be used in the model.
BooleanParam	fitIntercept() Param for whether to fit an intercept term.
Param<String>	link() Param for the name of link function which provides the relationship between the linear predictor and the mean of the distribution function.
DoubleParam	linkPower() Param for the index in the power link function.
Param<String>	linkPredictionCol() Param for link prediction (linear predictor) column name.
static GeneralizedLinearRegression	load(String path)
IntParam	maxIter() Param for maximum number of iterations (&gt;= 0).
Param<String>	offsetCol() Param for offset column name.
static MLReader<T>	read()
DoubleParam	regParam() Param for regularization parameter (&gt;= 0).
GeneralizedLinearRegression	setAggregationDepth(int value)
GeneralizedLinearRegression	setFamily(String value) Sets the value of param family.
GeneralizedLinearRegression	setFitIntercept(boolean value) Sets if we should fit the intercept.
GeneralizedLinearRegression	setLink(String value) Sets the value of param link.
GeneralizedLinearRegression	setLinkPower(double value) Sets the value of param linkPower.
GeneralizedLinearRegression	setLinkPredictionCol(String value) Sets the link prediction (linear predictor) column name.
GeneralizedLinearRegression	setMaxIter(int value) Sets the maximum number of iterations (applicable for solver "irls").
GeneralizedLinearRegression	setOffsetCol(String value) Sets the value of param offsetCol.
GeneralizedLinearRegression	setRegParam(double value) Sets the regularization parameter for L2 regularization.
GeneralizedLinearRegression	setSolver(String value) Sets the solver algorithm used for optimization.
GeneralizedLinearRegression	setTol(double value) Sets the convergence tolerance of iterations.
GeneralizedLinearRegression	setVariancePower(double value) Sets the value of param variancePower.
GeneralizedLinearRegression	setWeightCol(String value) Sets the value of param weightCol.
Param<String>	solver() The solver algorithm for optimization.
DoubleParam	tol() Param for the convergence tolerance for iterative algorithms (&gt;= 0).
String	uid() An immutable unique ID for the object and its derivatives.
DoubleParam	variancePower() Param for the power in the variance function of the Tweedie distribution which provides the relationship between the variance and mean of the distribution.
Param<String>	weightCol() Param for weight column name.

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

featuresCol, fit, labelCol, predictionCol, 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

params

Methods inherited from class Object

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

Methods inherited from interface org.apache.spark.ml.regression.GeneralizedLinearRegressionBase

getFamily, getLink, getLinkPower, getLinkPredictionCol, getOffsetCol, getVariancePower, hasLinkPredictionCol, hasOffsetCol, hasWeightCol, validateAndTransformSchema

Methods inherited from interface org.apache.spark.ml.PredictorParams

extractInstances, extractInstances

Methods inherited from interface org.apache.spark.ml.param.shared.HasLabelCol

getLabelCol, labelCol

Methods inherited from interface org.apache.spark.ml.param.shared.HasFeaturesCol

featuresCol, getFeaturesCol

Methods inherited from interface org.apache.spark.ml.param.shared.HasPredictionCol

getPredictionCol, predictionCol

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.ml.param.shared.HasFitIntercept

getFitIntercept

Methods inherited from interface org.apache.spark.ml.param.shared.HasMaxIter

getMaxIter

Methods inherited from interface org.apache.spark.ml.param.shared.HasTol

getTol

Methods inherited from interface org.apache.spark.ml.param.shared.HasRegParam

getRegParam

Methods inherited from interface org.apache.spark.ml.param.shared.HasWeightCol

getWeightCol

Methods inherited from interface org.apache.spark.ml.param.shared.HasSolver

getSolver

Methods inherited from interface org.apache.spark.ml.param.shared.HasAggregationDepth

getAggregationDepth

Methods inherited from interface org.apache.spark.internal.Logging

$init$, initializeForcefully, initializeLogIfNecessary, initializeLogIfNecessary, initializeLogIfNecessary$default$2, initLock, isTraceEnabled, log, logDebug, logDebug, logError, logError, logInfo, logInfo, logName, logTrace, logTrace, logWarning, logWarning, org$apache$spark$internal$Logging$$log__$eq, org$apache$spark$internal$Logging$$log_, uninitialize

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

write

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

save

Constructor Detail

GeneralizedLinearRegression

public GeneralizedLinearRegression(String uid)

GeneralizedLinearRegression

public GeneralizedLinearRegression()

Method Detail

load

public static GeneralizedLinearRegression load(String path)

read

public static MLReader<T> read()

family

public final Param<String> family()

Description copied from interface: GeneralizedLinearRegressionBase

Param for the name of family which is a description of the error distribution to be used in the model. Supported options: "gaussian", "binomial", "poisson", "gamma" and "tweedie". Default is "gaussian".

Specified by:

family in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

variancePower

public final DoubleParam variancePower()

Description copied from interface: GeneralizedLinearRegressionBase

Param for the power in the variance function of the Tweedie distribution which provides the relationship between the variance and mean of the distribution. Only applicable to the Tweedie family. (see Tweedie Distribution (Wikipedia)) Supported values: 0 and [1, Inf). Note that variance power 0, 1, or 2 corresponds to the Gaussian, Poisson or Gamma family, respectively.

Specified by:

variancePower in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

link

public final Param<String> link()

Description copied from interface: GeneralizedLinearRegressionBase

Param for the name of link function which provides the relationship between the linear predictor and the mean of the distribution function. Supported options: "identity", "log", "inverse", "logit", "probit", "cloglog" and "sqrt". This is used only when family is not "tweedie". The link function for the "tweedie" family must be specified through linkPower.

Specified by:

link in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

linkPower

public final DoubleParam linkPower()

Description copied from interface: GeneralizedLinearRegressionBase

Param for the index in the power link function. Only applicable to the Tweedie family. Note that link power 0, 1, -1 or 0.5 corresponds to the Log, Identity, Inverse or Sqrt link, respectively. When not set, this value defaults to 1 - variancePower, which matches the R "statmod" package.

Specified by:

linkPower in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

linkPredictionCol

public final Param<String> linkPredictionCol()

Description copied from interface: GeneralizedLinearRegressionBase

Param for link prediction (linear predictor) column name. Default is not set, which means we do not output link prediction.

Specified by:

linkPredictionCol in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

offsetCol

public final Param<String> offsetCol()

Description copied from interface: GeneralizedLinearRegressionBase

Param for offset column name. If this is not set or empty, we treat all instance offsets as 0.0. The feature specified as offset has a constant coefficient of 1.0.

Specified by:

offsetCol in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

solver

public final Param<String> solver()

Description copied from interface: GeneralizedLinearRegressionBase

The solver algorithm for optimization. Supported options: "irls" (iteratively reweighted least squares). Default: "irls"

Specified by:

solver in interface HasSolver

Specified by:

solver in interface GeneralizedLinearRegressionBase

Returns:

(undocumented)

aggregationDepth

public final IntParam aggregationDepth()

Description copied from interface: HasAggregationDepth

Param for suggested depth for treeAggregate (&gt;= 2).

Specified by:

aggregationDepth in interface HasAggregationDepth

Returns:

(undocumented)

weightCol

public final Param<String> weightCol()

Description copied from interface: HasWeightCol

Param for weight column name. If this is not set or empty, we treat all instance weights as 1.0.

Specified by:

weightCol in interface HasWeightCol

Returns:

(undocumented)

regParam

public final DoubleParam regParam()

Description copied from interface: HasRegParam

Param for regularization parameter (&gt;= 0).

Specified by:

regParam in interface HasRegParam

Returns:

(undocumented)

tol

public final DoubleParam tol()

Description copied from interface: HasTol

Param for the convergence tolerance for iterative algorithms (&gt;= 0).

Specified by:

tol in interface HasTol

Returns:

(undocumented)

maxIter

public final IntParam maxIter()

Description copied from interface: HasMaxIter

Param for maximum number of iterations (&gt;= 0).

Specified by:

maxIter in interface HasMaxIter

Returns:

(undocumented)

fitIntercept

public final BooleanParam fitIntercept()

Description copied from interface: HasFitIntercept

Param for whether to fit an intercept term.

Specified by:

fitIntercept in interface HasFitIntercept

Returns:

(undocumented)

uid

public 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)

setFamily

public GeneralizedLinearRegression setFamily(String value)

Sets the value of param family. Default is "gaussian".

Parameters:

value - (undocumented)

Returns:

(undocumented)

setVariancePower

public GeneralizedLinearRegression setVariancePower(double value)

Sets the value of param variancePower. Used only when family is "tweedie". Default is 0.0, which corresponds to the "gaussian" family.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setLinkPower

public GeneralizedLinearRegression setLinkPower(double value)

Sets the value of param linkPower. Used only when family is "tweedie".

Parameters:

value - (undocumented)

Returns:

(undocumented)

setLink

public GeneralizedLinearRegression setLink(String value)

Sets the value of param link. Used only when family is not "tweedie".

Parameters:

value - (undocumented)

Returns:

(undocumented)

setFitIntercept

public GeneralizedLinearRegression setFitIntercept(boolean value)

Sets if we should fit the intercept. Default is true.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setMaxIter

public GeneralizedLinearRegression setMaxIter(int value)

Sets the maximum number of iterations (applicable for solver "irls"). Default is 25.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setTol

public GeneralizedLinearRegression setTol(double value)

Sets 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)

setRegParam

public GeneralizedLinearRegression setRegParam(double value)

Sets the regularization parameter for L2 regularization. The regularization term is

$$ 0.5 * regParam * L2norm(coefficients)^2 $$

Default is 0.0.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setWeightCol

public GeneralizedLinearRegression setWeightCol(String value)

Sets the value of param weightCol. If this is not set or empty, we treat all instance weights as 1.0. Default is not set, so all instances have weight one. In the Binomial family, weights correspond to number of trials and should be integer. Non-integer weights are rounded to integer in AIC calculation.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setOffsetCol

public GeneralizedLinearRegression setOffsetCol(String value)

Sets the value of param offsetCol. If this is not set or empty, we treat all instance offsets as 0.0. Default is not set, so all instances have offset 0.0.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setSolver

public GeneralizedLinearRegression setSolver(String value)

Sets the solver algorithm used for optimization. Currently only supports "irls" which is also the default solver.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setLinkPredictionCol

public GeneralizedLinearRegression setLinkPredictionCol(String value)

Sets the link prediction (linear predictor) column name.

Parameters:

value - (undocumented)

Returns:

(undocumented)

setAggregationDepth

public GeneralizedLinearRegression setAggregationDepth(int value)

copy

public GeneralizedLinearRegression 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. See defaultCopy().

Specified by:

copy in interface Params

Specified by:

copy in class Predictor<Vector,GeneralizedLinearRegression,GeneralizedLinearRegressionModel>

Parameters:

extra - (undocumented)

Returns:

(undocumented)