magpie.models.regression

Class LASSORegression

java.lang.Object

magpie.models.BaseModel

magpie.models.regression.BaseRegression

magpie.models.regression.LASSORegression

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, AbstractRegressionModel, Citable, Commandable, Options, Printable, Savable

public class LASSORegression
extends BaseRegression

This class enables regression using the LASSO algorithm, as described by this page. It is basically linear regression with some additional constraints to make the solution vector have a low L₁ norm.

For now, this uses Forward Stepwise Regression.

Usage: -maxterms <terms>
terms: Maximum number of terms allowed in expression

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected java.util.List<java.lang.Double>	Coefficients Coefficients for linear model
protected double	Intercept Intercept of the linear model
protected int	MaxNumberTerms Maximum number of features allowed in model (-1 is unlimited)
protected java.util.List<java.lang.String>	TermNames Names of terms used in the model
protected java.util.List<java.lang.Integer>	Terms Terms used in the model

Fields inherited from class magpie.models.BaseModel

AttributeSelector, trained, TrainingStats, validated, ValidationStats

Constructor Summary

Constructors

Constructor and Description
LASSORegression()

Method Summary

Modifier and Type	Method and Description
BaseRegression	clone()
protected int	findMaxCorrelation(double[][] features, double[] objective, boolean[] isSearchable) Finds the feature with the maximum correlation out of all possible candidates
static double	getMAE(double[] residuals) Get Mean Absolute Error (MAE) given residuals
int	getNFittingParameters() Number of fitting parameters in a model.
double[]	getResidual(double[][] x, double[] y) Determines the residuals from a model
static double[]	linearFit(double[] x, double[] y, boolean intercept) Find the best linear fit
protected java.lang.String	printModel_protected() Internal method that handles printing the model as a string.
java.util.List<java.lang.String>	printModelDescriptionDetails(boolean htmlFormat) Print details of the model.
java.lang.String	printUsage() Print out required format for options.
void	run_protected(Dataset TrainData) Run a model without checking if stuff is trained (use carefully)
double[]	runModel(double[][] x) Runs the model contained within object
void	setMaxNumberTerms(int MaxNumberTerms)
void	setOptions(java.util.List OptionsObj) Set any options for this object.
protected void	train_protected(Dataset TrainData) Train a model without evaluating performance

Methods inherited from class magpie.models.regression.BaseRegression

doRobustRegression, getRobustRegressionQ, robustTraining, runCommand, setRobustRegressionQ

Methods inherited from class magpie.models.BaseModel

about, crossValidate, crossValidate, crossValidate, done, externallyValidate, getAttributeSelector, getCitations, getFilter, getTrainTime, getValidationMethod, handleSetCommand, isTrained, isValidated, loadState, printCommand, printDescription, printModel, resetModel, run, saveCommand, saveState, setAttributeSelector, setComponent, setFilter, train, train

Methods inherited from class java.lang.Object

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

Field Detail

MaxNumberTerms

protected int MaxNumberTerms

Maximum number of features allowed in model (-1 is unlimited)

Terms

protected java.util.List<java.lang.Integer> Terms

Terms used in the model

TermNames

protected java.util.List<java.lang.String> TermNames

Names of terms used in the model

Coefficients

protected java.util.List<java.lang.Double> Coefficients

Coefficients for linear model

Intercept

protected double Intercept

Intercept of the linear model

Constructor Detail

LASSORegression

public LASSORegression()

Method Detail

clone

public BaseRegression clone()

Overrides:

clone in class BaseRegression

setOptions

public void setOptions(java.util.List OptionsObj)
                throws java.lang.Exception

Description copied from interface: Options

Set any options for this object. If parsing the options fails, return an Exception with the usage as its message.

Parameters:

OptionsObj - Array of options as Objects - can be null

Throws:

java.lang.Exception - if problem with inputs

printUsage

public java.lang.String printUsage()

Description copied from interface: Options

Print out required format for options. For consistency, should start with "Usage: ".

Returns:

Usage for this function

setMaxNumberTerms

public void setMaxNumberTerms(int MaxNumberTerms)

train_protected

protected void train_protected(Dataset TrainData)

Description copied from class: BaseModel

Train a model without evaluating performance

Specified by:

train_protected in class BaseModel

Parameters:

TrainData - Training data

getNFittingParameters

public int getNFittingParameters()

Description copied from interface: AbstractRegressionModel

Number of fitting parameters in a model. Only pertinent to nonlinear regression. Otherwise, this should just return 0.

Returns:

Number of fitting parameters

run_protected

public void run_protected(Dataset TrainData)

Description copied from class: BaseModel

Run a model without checking if stuff is trained (use carefully)

Specified by:

run_protected in class BaseModel

Parameters:

TrainData - Training data

findMaxCorrelation

protected int findMaxCorrelation(double[][] features,
                                 double[] objective,
                                 boolean[] isSearchable)

Finds the feature with the maximum correlation out of all possible candidates

Parameters:

features - Features for each measurement

objective - Objective function for each measurement

isSearchable - List of which features are searchable

Returns:

Index of feature that is searchable and has the highest correlation coefficient. Return -1 if no features are left to add.

linearFit

public static double[] linearFit(double[] x,
                                 double[] y,
                                 boolean intercept)

Find the best linear fit

Parameters:

x - Independent variable

y - Dependent variable

intercept - Whether to fit an intercept

Returns:

Optimal fit: [0] is the intercept, [1] is the slope

getResidual

public double[] getResidual(double[][] x,
                            double[] y)

Determines the residuals from a model

Parameters:

x - Observation matrix

y - Data to compare against model

Returns:

y - runModel(x)

runModel

public double[] runModel(double[][] x)

Runs the model contained within object

Parameters:

x - Observation matrix

Returns:

getMAE

public static double getMAE(double[] residuals)

Get Mean Absolute Error (MAE) given residuals

Parameters:

residuals - Error between model and reality for each entry

Returns:

Mean absolute error

printModel_protected

protected java.lang.String printModel_protected()

Description copied from class: BaseModel

Internal method that handles printing the model as a string. Should end with a "\n"

Specified by:

printModel_protected in class BaseModel

Returns:

String representation of model

printModelDescriptionDetails

public java.util.List<java.lang.String> printModelDescriptionDetails(boolean htmlFormat)

Description copied from class: BaseModel

Print details of the model. Used by BaseModel.printDescription(boolean).

Implementation note: No not add indentation for details. That is handled by BaseModel.printDescription(boolean). You should also call the super operation to get the Normalizer and Attribute selector settings

Overrides:

printModelDescriptionDetails in class BaseModel

Parameters:

htmlFormat - Whether to use HTML format

Returns:

List describing model details. Each entry is a different line of the description (i.e., in place of newline characters).