org.apache.commons.math3.stat.regression
Class RegressionResults

java.lang.Object
  extended by org.apache.commons.math3.stat.regression.RegressionResults
All Implemented Interfaces:
Serializable

public class RegressionResults
extends Object
implements Serializable

Results of a Multiple Linear Regression model fit.

Since:
3.0
Version:
$Id: RegressionResults.java 1392342 2012-10-01 14:08:52Z psteitz $
See Also:
Serialized Form

Constructor Summary
RegressionResults(double[] parameters, double[][] varcov, boolean isSymmetricCompressed, long nobs, int rank, double sumy, double sumysq, double sse, boolean containsConstant, boolean copyData)
          Constructor for Regression Results.
 
Method Summary
 double getAdjustedRSquared()
          Returns the adjusted R-squared statistic, defined by the formula
 double getCovarianceOfParameters(int i, int j)
          Returns the covariance between regression parameters i and j.
 double getErrorSumSquares()
          Returns the sum of squared errors (SSE) associated with the regression model.
 double getMeanSquareError()
          Returns the sum of squared errors divided by the degrees of freedom, usually abbreviated MSE.
 long getN()
          Returns the number of observations added to the regression model.
 int getNumberOfParameters()
          Returns the number of parameters estimated in the model.
 double getParameterEstimate(int index)
          Returns the parameter estimate for the regressor at the given index.
 double[] getParameterEstimates()
          Returns a copy of the regression parameters estimates.
 double getRegressionSumSquares()
          Returns the sum of squared deviations of the predicted y values about their mean (which equals the mean of y).
 double getRSquared()
          Returns the coefficient of multiple determination, usually denoted r-square.
 double getStdErrorOfEstimate(int index)
          Returns the standard error of the parameter estimate at index, usually denoted s(bindex).
 double[] getStdErrorOfEstimates()
          Returns the standard error of the parameter estimates, usually denoted s(bi).
 double getTotalSumSquares()
          Returns the sum of squared deviations of the y values about their mean.
 boolean hasIntercept()
          Returns true if the regression model has been computed including an intercept.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

RegressionResults

public RegressionResults(double[] parameters,
                         double[][] varcov,
                         boolean isSymmetricCompressed,
                         long nobs,
                         int rank,
                         double sumy,
                         double sumysq,
                         double sse,
                         boolean containsConstant,
                         boolean copyData)
Constructor for Regression Results.

Parameters:
parameters - a double array with the regression slope estimates
varcov - the variance covariance matrix, stored either in a square matrix or as a compressed
isSymmetricCompressed - a flag which denotes that the variance covariance matrix is in symmetric compressed format
nobs - the number of observations of the regression estimation
rank - the number of independent variables in the regression
sumy - the sum of the independent variable
sumysq - the sum of the squared independent variable
sse - sum of squared errors
containsConstant - true model has constant, false model does not have constant
copyData - if true a deep copy of all input data is made, if false only references are copied and the RegressionResults become mutable
Method Detail

getParameterEstimate

public double getParameterEstimate(int index)
                            throws OutOfRangeException

Returns the parameter estimate for the regressor at the given index.

A redundant regressor will have its redundancy flag set, as well as a parameters estimated equal to Double.NaN

Parameters:
index - Index.
Returns:
the parameters estimated for regressor at index.
Throws:
OutOfRangeException - if index is not in the interval [0, number of parameters).

getParameterEstimates

public double[] getParameterEstimates()

Returns a copy of the regression parameters estimates.

The parameter estimates are returned in the natural order of the data.

A redundant regressor will have its redundancy flag set, as will a parameter estimate equal to Double.NaN.

Returns:
array of parameter estimates, null if no estimation occurred

getStdErrorOfEstimate

public double getStdErrorOfEstimate(int index)
                             throws OutOfRangeException
Returns the standard error of the parameter estimate at index, usually denoted s(bindex).

Parameters:
index - Index.
Returns:
the standard errors associated with parameters estimated at index.
Throws:
OutOfRangeException - if index is not in the interval [0, number of parameters).

getStdErrorOfEstimates

public double[] getStdErrorOfEstimates()

Returns the standard error of the parameter estimates, usually denoted s(bi).

If there are problems with an ill conditioned design matrix then the regressor which is redundant will be assigned Double.NaN.

Returns:
an array standard errors associated with parameters estimates, null if no estimation occurred

getCovarianceOfParameters

public double getCovarianceOfParameters(int i,
                                        int j)
                                 throws OutOfRangeException

Returns the covariance between regression parameters i and j.

If there are problems with an ill conditioned design matrix then the covariance which involves redundant columns will be assigned Double.NaN.

Parameters:
i - ith regression parameter.
j - jth regression parameter.
Returns:
the covariance of the parameter estimates.
Throws:
OutOfRangeException - if i or j is not in the interval [0, number of parameters).

getNumberOfParameters

public int getNumberOfParameters()

Returns the number of parameters estimated in the model.

This is the maximum number of regressors, some techniques may drop redundant parameters

Returns:
number of regressors, -1 if not estimated

getN

public long getN()
Returns the number of observations added to the regression model.

Returns:
Number of observations, -1 if an error condition prevents estimation

getTotalSumSquares

public double getTotalSumSquares()

Returns the sum of squared deviations of the y values about their mean.

This is defined as SSTO here.

If n < 2, this returns Double.NaN.

Returns:
sum of squared deviations of y values

getRegressionSumSquares

public double getRegressionSumSquares()

Returns the sum of squared deviations of the predicted y values about their mean (which equals the mean of y).

This is usually abbreviated SSR or SSM. It is defined as SSM here

Preconditions:

Returns:
sum of squared deviations of predicted y values

getErrorSumSquares

public double getErrorSumSquares()

Returns the sum of squared errors (SSE) associated with the regression model.

The return value is constrained to be non-negative - i.e., if due to rounding errors the computational formula returns a negative result, 0 is returned.

Preconditions:

Returns:
sum of squared errors associated with the regression model

getMeanSquareError

public double getMeanSquareError()

Returns the sum of squared errors divided by the degrees of freedom, usually abbreviated MSE.

If there are fewer than numberOfParameters + 1 data pairs in the model, or if there is no variation in x, this returns Double.NaN.

Returns:
sum of squared deviations of y values

getRSquared

public double getRSquared()

Returns the coefficient of multiple determination, usually denoted r-square.

Preconditions:

Returns:
r-square, a double in the interval [0, 1]

getAdjustedRSquared

public double getAdjustedRSquared()

Returns the adjusted R-squared statistic, defined by the formula

 R2adj = 1 - [SSR (n - 1)] / [SSTO (n - p)]
 
where SSR is the sum of squared residuals}, SSTO is the total sum of squares}, n is the number of observations and p is the number of parameters estimated (including the intercept).

If the regression is estimated without an intercept term, what is returned is

  1 - (1 - getRSquared() ) * (n / (n - p)) 
 

Returns:
adjusted R-Squared statistic

hasIntercept

public boolean hasIntercept()
Returns true if the regression model has been computed including an intercept. In this case, the coefficient of the intercept is the first element of the parameter estimates.

Returns:
true if the model has an intercept term


Copyright © 2003-2012 The Apache Software Foundation. All Rights Reserved.